From ed4e8313e0f3a46e432a259841e6d58f971f2004 Mon Sep 17 00:00:00 2001 From: dapan1121 Date: Thu, 16 Feb 2023 08:55:13 +0800 Subject: [PATCH 01/31] fix: add force stop task and rename conflict structer name --- include/util/tworker.h | 6 +- source/client/src/clientHb.c | 2 +- source/dnode/mnode/impl/inc/mndInt.h | 2 +- source/dnode/qnode/inc/qndInt.h | 2 +- source/dnode/vnode/src/inc/vnodeInt.h | 2 +- source/libs/qworker/inc/qwInt.h | 2 + source/libs/qworker/src/qwDbg.c | 12 +++- source/libs/qworker/src/qworker.c | 90 ++++++++++++++++----------- source/util/src/tworker.c | 22 +++---- 9 files changed, 82 insertions(+), 58 deletions(-) diff --git a/include/util/tworker.h b/include/util/tworker.h index 0636f16dbb..8508adf052 100644 --- a/include/util/tworker.h +++ b/include/util/tworker.h @@ -26,12 +26,12 @@ extern "C" { typedef struct SQWorkerPool SQWorkerPool; typedef struct SWWorkerPool SWWorkerPool; -typedef struct SQWorker { +typedef struct SQueueWorker { int32_t id; // worker id int64_t pid; // thread pid TdThread thread; // thread id void *pool; -} SQWorker; +} SQueueWorker; typedef struct SQWorkerPool { int32_t max; // max number of workers @@ -39,7 +39,7 @@ typedef struct SQWorkerPool { int32_t num; // current number of workers STaosQset *qset; const char *name; - SQWorker *workers; + SQueueWorker *workers; TdThreadMutex mutex; } SQWorkerPool; diff --git a/source/client/src/clientHb.c b/source/client/src/clientHb.c index 47ed2cf035..b01a871702 100644 --- a/source/client/src/clientHb.c +++ b/source/client/src/clientHb.c @@ -347,7 +347,7 @@ int32_t hbBuildQueryDesc(SQueryHbReqBasic *hbBasic, STscObj *pObj) { continue; } - if (pRequest->killed) { + if (pRequest->killed || 0 == pRequest->body.queryJob) { releaseRequest(*rid); pIter = taosHashIterate(pObj->pRequests, pIter); continue; diff --git a/source/dnode/mnode/impl/inc/mndInt.h b/source/dnode/mnode/impl/inc/mndInt.h index 785ecc2bf5..d851578a74 100644 --- a/source/dnode/mnode/impl/inc/mndInt.h +++ b/source/dnode/mnode/impl/inc/mndInt.h @@ -58,7 +58,7 @@ typedef int32_t (*MndInitFp)(SMnode *pMnode); typedef void (*MndCleanupFp)(SMnode *pMnode); typedef int32_t (*ShowRetrieveFp)(SRpcMsg *pMsg, SShowObj *pShow, SSDataBlock *pBlock, int32_t rows); typedef void (*ShowFreeIterFp)(SMnode *pMnode, void *pIter); -typedef struct SQWorker SQHandle; +typedef struct SQueueWorker SQHandle; typedef struct { const char *name; diff --git a/source/dnode/qnode/inc/qndInt.h b/source/dnode/qnode/inc/qndInt.h index e8ccb75040..86deda52ad 100644 --- a/source/dnode/qnode/inc/qndInt.h +++ b/source/dnode/qnode/inc/qndInt.h @@ -29,7 +29,7 @@ extern "C" { #endif -typedef struct SQWorker SQHandle; +typedef struct SQueueWorker SQHandle; typedef struct SQnode { int32_t qndId; diff --git a/source/dnode/vnode/src/inc/vnodeInt.h b/source/dnode/vnode/src/inc/vnodeInt.h index ec925087d0..adec027284 100644 --- a/source/dnode/vnode/src/inc/vnodeInt.h +++ b/source/dnode/vnode/src/inc/vnodeInt.h @@ -58,7 +58,7 @@ typedef struct STQ STQ; typedef struct SVState SVState; typedef struct SVStatis SVStatis; typedef struct SVBufPool SVBufPool; -typedef struct SQWorker SQHandle; +typedef struct SQueueWorker SQHandle; typedef struct STsdbKeepCfg STsdbKeepCfg; typedef struct SMetaSnapReader SMetaSnapReader; typedef struct SMetaSnapWriter SMetaSnapWriter; diff --git a/source/libs/qworker/inc/qwInt.h b/source/libs/qworker/inc/qwInt.h index bde05d4116..f656a22742 100644 --- a/source/libs/qworker/inc/qwInt.h +++ b/source/libs/qworker/inc/qwInt.h @@ -76,6 +76,7 @@ typedef struct SQWDebug { bool lockEnable; bool statusEnable; bool dumpEnable; + bool forceStop; bool sleepSimulate; bool deadSimulate; bool redirectSimulate; @@ -246,6 +247,7 @@ typedef struct SQWorkerMgmt { #define QW_QUERY_RUNNING(ctx) (QW_GET_PHASE(ctx) == QW_PHASE_PRE_QUERY || QW_GET_PHASE(ctx) == QW_PHASE_PRE_CQUERY) #define QW_FETCH_RUNNING(ctx) ((ctx)->inFetch) +#define QW_QUERY_NOT_STARTED(ctx) (QW_GET_PHASE(ctx) == -1) #define QW_SET_QTID(id, qId, tId, eId) \ do { \ diff --git a/source/libs/qworker/src/qwDbg.c b/source/libs/qworker/src/qwDbg.c index 7a755cd36f..0ab501ddd5 100644 --- a/source/libs/qworker/src/qwDbg.c +++ b/source/libs/qworker/src/qwDbg.c @@ -9,11 +9,13 @@ #include "tmsg.h" #include "tname.h" -SQWDebug gQWDebug = {.statusEnable = true, +SQWDebug gQWDebug = {.lockEnable = false, + .statusEnable = true, .dumpEnable = false, .redirectSimulate = false, .deadSimulate = false, - .sleepSimulate = false}; + .sleepSimulate = false, + .forceStop = false}; int32_t qwDbgValidateStatus(QW_FPARAMS_DEF, int8_t oriStatus, int8_t newStatus, bool *ignore) { if (!gQWDebug.statusEnable) { @@ -306,6 +308,12 @@ int32_t qwDbgEnableDebug(char *option) { return TSDB_CODE_SUCCESS; } + if (0 == strcasecmp(option, "forceStop")) { + gQWDebug.forceStop = true; + qError("qw forceStop debug enabled"); + return TSDB_CODE_SUCCESS; + } + qError("invalid qw debug option:%s", option); return TSDB_CODE_APP_ERROR; diff --git a/source/libs/qworker/src/qworker.c b/source/libs/qworker/src/qworker.c index 2df0109d19..e2db0b8319 100644 --- a/source/libs/qworker/src/qworker.c +++ b/source/libs/qworker/src/qworker.c @@ -18,6 +18,51 @@ SQWorkerMgmt gQwMgmt = { .qwNum = 0, }; + +int32_t qwStopAllTasks(SQWorker *mgmt) { + uint64_t qId, tId, sId; + int32_t eId; + int64_t rId = 0; + + void *pIter = taosHashIterate(mgmt->ctxHash, NULL); + while (pIter) { + SQWTaskCtx *ctx = (SQWTaskCtx *)pIter; + void *key = taosHashGetKey(pIter, NULL); + QW_GET_QTID(key, qId, tId, eId); + + QW_LOCK(QW_WRITE, &ctx->lock); + + sId = ctx->sId; + + QW_TASK_DLOG_E("start to force stop task"); + + if (QW_EVENT_RECEIVED(ctx, QW_EVENT_DROP) || QW_EVENT_PROCESSED(ctx, QW_EVENT_DROP)) { + QW_TASK_WLOG_E("task already dropping"); + QW_UNLOCK(QW_WRITE, &ctx->lock); + + pIter = taosHashIterate(mgmt->ctxHash, pIter); + continue; + } + + if (QW_QUERY_RUNNING(ctx)) { + qwKillTaskHandle(ctx, TSDB_CODE_VND_STOPPED); + QW_TASK_DLOG_E("task running, async killed"); + } else if (QW_FETCH_RUNNING(ctx)) { + QW_UPDATE_RSP_CODE(ctx, TSDB_CODE_VND_STOPPED); + QW_SET_EVENT_RECEIVED(ctx, QW_EVENT_DROP); + QW_TASK_DLOG_E("task fetching, update drop received"); + } else { + qwDropTask(QW_FPARAMS()); + } + + QW_UNLOCK(QW_WRITE, &ctx->lock); + + pIter = taosHashIterate(mgmt->ctxHash, pIter); + } + + return TSDB_CODE_SUCCESS; +} + int32_t qwProcessHbLinkBroken(SQWorker *mgmt, SQWMsg *qwMsg, SSchedulerHbReq *req) { int32_t code = 0; SSchedulerHbRsp rsp = {0}; @@ -973,6 +1018,10 @@ void qwProcessHbTimerEvent(void *param, void *tmrId) { qwDbgDumpMgmtInfo(mgmt); + if (gQWDebug.forceStop) { + (void)qwStopAllTasks(mgmt); + } + QW_LOCK(QW_READ, &mgmt->schLock); int32_t schNum = taosHashGetSize(mgmt->schHash); @@ -1087,6 +1136,7 @@ _return: QW_RET(TSDB_CODE_SUCCESS); } + int32_t qWorkerInit(int8_t nodeType, int32_t nodeId, void **qWorkerMgmt, const SMsgCb *pMsgCb) { if (NULL == qWorkerMgmt || (pMsgCb && pMsgCb->mgmt == NULL)) { qError("invalid param to init qworker"); @@ -1185,46 +1235,10 @@ void qWorkerStopAllTasks(void *qWorkerMgmt) { SQWorker *mgmt = (SQWorker *)qWorkerMgmt; QW_DLOG("start to stop all tasks, taskNum:%d", taosHashGetSize(mgmt->ctxHash)); - - uint64_t qId, tId, sId; - int32_t eId; - int64_t rId = 0; - + atomic_store_8(&mgmt->nodeStopped, 1); - void *pIter = taosHashIterate(mgmt->ctxHash, NULL); - while (pIter) { - SQWTaskCtx *ctx = (SQWTaskCtx *)pIter; - void *key = taosHashGetKey(pIter, NULL); - QW_GET_QTID(key, qId, tId, eId); - - QW_LOCK(QW_WRITE, &ctx->lock); - - sId = ctx->sId; - - QW_TASK_DLOG_E("start to force stop task"); - - if (QW_EVENT_RECEIVED(ctx, QW_EVENT_DROP) || QW_EVENT_PROCESSED(ctx, QW_EVENT_DROP)) { - QW_TASK_WLOG_E("task already dropping"); - QW_UNLOCK(QW_WRITE, &ctx->lock); - - pIter = taosHashIterate(mgmt->ctxHash, pIter); - continue; - } - - if (QW_QUERY_RUNNING(ctx)) { - qwKillTaskHandle(ctx, TSDB_CODE_VND_STOPPED); - } else if (QW_FETCH_RUNNING(ctx)) { - QW_UPDATE_RSP_CODE(ctx, TSDB_CODE_VND_STOPPED); - QW_SET_EVENT_RECEIVED(ctx, QW_EVENT_DROP); - } else { - qwDropTask(QW_FPARAMS()); - } - - QW_UNLOCK(QW_WRITE, &ctx->lock); - - pIter = taosHashIterate(mgmt->ctxHash, pIter); - } + (void)qwStopAllTasks(mgmt); } void qWorkerDestroy(void **qWorkerMgmt) { diff --git a/source/util/src/tworker.c b/source/util/src/tworker.c index 5581931178..631bcb443e 100644 --- a/source/util/src/tworker.c +++ b/source/util/src/tworker.c @@ -22,7 +22,7 @@ typedef void *(*ThreadFp)(void *param); int32_t tQWorkerInit(SQWorkerPool *pool) { pool->qset = taosOpenQset(); - pool->workers = taosMemoryCalloc(pool->max, sizeof(SQWorker)); + pool->workers = taosMemoryCalloc(pool->max, sizeof(SQueueWorker)); if (pool->workers == NULL) { terrno = TSDB_CODE_OUT_OF_MEMORY; return -1; @@ -31,7 +31,7 @@ int32_t tQWorkerInit(SQWorkerPool *pool) { (void)taosThreadMutexInit(&pool->mutex, NULL); for (int32_t i = 0; i < pool->max; ++i) { - SQWorker *worker = pool->workers + i; + SQueueWorker *worker = pool->workers + i; worker->id = i; worker->pool = pool; } @@ -42,14 +42,14 @@ int32_t tQWorkerInit(SQWorkerPool *pool) { void tQWorkerCleanup(SQWorkerPool *pool) { for (int32_t i = 0; i < pool->max; ++i) { - SQWorker *worker = pool->workers + i; + SQueueWorker *worker = pool->workers + i; if (taosCheckPthreadValid(worker->thread)) { taosQsetThreadResume(pool->qset); } } for (int32_t i = 0; i < pool->max; ++i) { - SQWorker *worker = pool->workers + i; + SQueueWorker *worker = pool->workers + i; if (taosCheckPthreadValid(worker->thread)) { uInfo("worker:%s:%d is stopping", pool->name, worker->id); taosThreadJoin(worker->thread, NULL); @@ -65,7 +65,7 @@ void tQWorkerCleanup(SQWorkerPool *pool) { uInfo("worker:%s is closed", pool->name); } -static void *tQWorkerThreadFp(SQWorker *worker) { +static void *tQWorkerThreadFp(SQueueWorker *worker) { SQWorkerPool *pool = worker->pool; SQueueInfo qinfo = {0}; void *msg = NULL; @@ -106,7 +106,7 @@ STaosQueue *tQWorkerAllocQueue(SQWorkerPool *pool, void *ahandle, FItem fp) { // spawn a thread to process queue if (pool->num < pool->max) { do { - SQWorker *worker = pool->workers + pool->num; + SQueueWorker *worker = pool->workers + pool->num; TdThreadAttr thAttr; taosThreadAttrInit(&thAttr); @@ -138,7 +138,7 @@ void tQWorkerFreeQueue(SQWorkerPool *pool, STaosQueue *queue) { int32_t tAutoQWorkerInit(SAutoQWorkerPool *pool) { pool->qset = taosOpenQset(); - pool->workers = taosArrayInit(2, sizeof(SQWorker *)); + pool->workers = taosArrayInit(2, sizeof(SQueueWorker *)); if (pool->workers == NULL) { terrno = TSDB_CODE_OUT_OF_MEMORY; return -1; @@ -153,14 +153,14 @@ int32_t tAutoQWorkerInit(SAutoQWorkerPool *pool) { void tAutoQWorkerCleanup(SAutoQWorkerPool *pool) { int32_t size = taosArrayGetSize(pool->workers); for (int32_t i = 0; i < size; ++i) { - SQWorker *worker = taosArrayGetP(pool->workers, i); + SQueueWorker *worker = taosArrayGetP(pool->workers, i); if (taosCheckPthreadValid(worker->thread)) { taosQsetThreadResume(pool->qset); } } for (int32_t i = 0; i < size; ++i) { - SQWorker *worker = taosArrayGetP(pool->workers, i); + SQueueWorker *worker = taosArrayGetP(pool->workers, i); if (taosCheckPthreadValid(worker->thread)) { uInfo("worker:%s:%d is stopping", pool->name, worker->id); taosThreadJoin(worker->thread, NULL); @@ -177,7 +177,7 @@ void tAutoQWorkerCleanup(SAutoQWorkerPool *pool) { uInfo("worker:%s is closed", pool->name); } -static void *tAutoQWorkerThreadFp(SQWorker *worker) { +static void *tAutoQWorkerThreadFp(SQueueWorker *worker) { SAutoQWorkerPool *pool = worker->pool; SQueueInfo qinfo = {0}; void *msg = NULL; @@ -222,7 +222,7 @@ STaosQueue *tAutoQWorkerAllocQueue(SAutoQWorkerPool *pool, void *ahandle, FItem // spawn a thread to process queue while (curWorkerNum < dstWorkerNum) { - SQWorker *worker = taosMemoryCalloc(1, sizeof(SQWorker)); + SQueueWorker *worker = taosMemoryCalloc(1, sizeof(SQueueWorker)); if (worker == NULL || taosArrayPush(pool->workers, &worker) == NULL) { uError("worker:%s:%d failed to create", pool->name, curWorkerNum); taosMemoryFree(worker); From fd8cad28442cc0c0616404555157c1738fc84d78 Mon Sep 17 00:00:00 2001 From: Haojun Liao Date: Mon, 20 Feb 2023 10:22:41 +0800 Subject: [PATCH 02/31] refactor: remove xxhash. --- include/util/xxhash.h | 328 --------- source/client/test/clientTests.cpp | 2 +- source/util/src/xxhash.c | 1030 ---------------------------- 3 files changed, 1 insertion(+), 1359 deletions(-) delete mode 100644 include/util/xxhash.h delete mode 100644 source/util/src/xxhash.c diff --git a/include/util/xxhash.h b/include/util/xxhash.h deleted file mode 100644 index d6bad94335..0000000000 --- a/include/util/xxhash.h +++ /dev/null @@ -1,328 +0,0 @@ -/* - xxHash - Extremely Fast Hash algorithm - Header File - Copyright (C) 2012-2016, Yann Collet. - - BSD 2-Clause License (http://www.opensource.org/licenses/bsd-license.php) - - Redistribution and use in source and binary forms, with or without - modification, are permitted provided that the following conditions are - met: - - * Redistributions of source code must retain the above copyright - notice, this list of conditions and the following disclaimer. - * Redistributions in binary form must reproduce the above - copyright notice, this list of conditions and the following disclaimer - in the documentation and/or other materials provided with the - distribution. - - THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS - "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT - LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR - A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT - OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, - SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT - LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, - DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY - THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT - (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE - OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. - - You can contact the author at : - - xxHash source repository : https://github.com/Cyan4973/xxHash -*/ - -/* Notice extracted from xxHash homepage : - -xxHash is an extremely fast Hash algorithm, running at RAM speed limits. -It also successfully passes all tests from the SMHasher suite. - -Comparison (single thread, Windows Seven 32 bits, using SMHasher on a Core 2 Duo @3GHz) - -Name Speed Q.Score Author -xxHash 5.4 GB/s 10 -CrapWow 3.2 GB/s 2 Andrew -MumurHash 3a 2.7 GB/s 10 Austin Appleby -SpookyHash 2.0 GB/s 10 Bob Jenkins -SBox 1.4 GB/s 9 Bret Mulvey -Lookup3 1.2 GB/s 9 Bob Jenkins -SuperFastHash 1.2 GB/s 1 Paul Hsieh -CityHash64 1.05 GB/s 10 Pike & Alakuijala -FNV 0.55 GB/s 5 Fowler, Noll, Vo -CRC32 0.43 GB/s 9 -MD5-32 0.33 GB/s 10 Ronald L. Rivest -SHA1-32 0.28 GB/s 10 - -Q.Score is a measure of quality of the hash function. -It depends on successfully passing SMHasher test set. -10 is a perfect score. - -A 64-bit version, named XXH64, is available since r35. -It offers much better speed, but for 64-bit applications only. -Name Speed on 64 bits Speed on 32 bits -XXH64 13.8 GB/s 1.9 GB/s -XXH32 6.8 GB/s 6.0 GB/s -*/ - -#ifndef XXHASH_H_5627135585666179 -#define XXHASH_H_5627135585666179 1 - -#if defined (__cplusplus) -extern "C" { -#endif - - -/* **************************** -* Definitions -******************************/ -#include /* size_t */ -typedef enum { XXH_OK=0, XXH_ERROR } XXH_errorcode; - - -/* **************************** - * API modifier - ******************************/ -/** XXH_INLINE_ALL (and XXH_PRIVATE_API) - * This is useful to include xxhash functions in `static` mode - * in order to inline them, and remove their symbol from the public list. - * Inlining can offer dramatic performance improvement on small keys. - * Methodology : - * #define XXH_INLINE_ALL - * #include "xxhash.h" - * `xxhash.c` is automatically included. - * It's not useful to compile and link it as a separate module. - */ -#if defined(XXH_INLINE_ALL) || defined(XXH_PRIVATE_API) -# ifndef XXH_STATIC_LINKING_ONLY -# define XXH_STATIC_LINKING_ONLY -# endif -# if defined(__GNUC__) -# define XXH_PUBLIC_API static __inline __attribute__((unused)) -# elif defined (__cplusplus) || (defined (__STDC_VERSION__) && (__STDC_VERSION__ >= 199901L) /* C99 */) -# define XXH_PUBLIC_API static inline -# elif defined(_MSC_VER) -# define XXH_PUBLIC_API static __inline -# else - /* this version may generate warnings for unused static functions */ -# define XXH_PUBLIC_API static -# endif -#else -# define XXH_PUBLIC_API /* do nothing */ -#endif /* XXH_INLINE_ALL || XXH_PRIVATE_API */ - -/*! XXH_NAMESPACE, aka Namespace Emulation : - * - * If you want to include _and expose_ xxHash functions from within your own library, - * but also want to avoid symbol collisions with other libraries which may also include xxHash, - * - * you can use XXH_NAMESPACE, to automatically prefix any public symbol from xxhash library - * with the value of XXH_NAMESPACE (therefore, avoid NULL and numeric values). - * - * Note that no change is required within the calling program as long as it includes `xxhash.h` : - * regular symbol name will be automatically translated by this header. - */ -#ifdef XXH_NAMESPACE -# define XXH_CAT(A,B) A##B -# define XXH_NAME2(A,B) XXH_CAT(A,B) -# define XXH_versionNumber XXH_NAME2(XXH_NAMESPACE, XXH_versionNumber) -# define XXH32 XXH_NAME2(XXH_NAMESPACE, XXH32) -# define XXH32_createState XXH_NAME2(XXH_NAMESPACE, XXH32_createState) -# define XXH32_freeState XXH_NAME2(XXH_NAMESPACE, XXH32_freeState) -# define XXH32_reset XXH_NAME2(XXH_NAMESPACE, XXH32_reset) -# define XXH32_update XXH_NAME2(XXH_NAMESPACE, XXH32_update) -# define XXH32_digest XXH_NAME2(XXH_NAMESPACE, XXH32_digest) -# define XXH32_copyState XXH_NAME2(XXH_NAMESPACE, XXH32_copyState) -# define XXH32_canonicalFromHash XXH_NAME2(XXH_NAMESPACE, XXH32_canonicalFromHash) -# define XXH32_hashFromCanonical XXH_NAME2(XXH_NAMESPACE, XXH32_hashFromCanonical) -# define XXH64 XXH_NAME2(XXH_NAMESPACE, XXH64) -# define XXH64_createState XXH_NAME2(XXH_NAMESPACE, XXH64_createState) -# define XXH64_freeState XXH_NAME2(XXH_NAMESPACE, XXH64_freeState) -# define XXH64_reset XXH_NAME2(XXH_NAMESPACE, XXH64_reset) -# define XXH64_update XXH_NAME2(XXH_NAMESPACE, XXH64_update) -# define XXH64_digest XXH_NAME2(XXH_NAMESPACE, XXH64_digest) -# define XXH64_copyState XXH_NAME2(XXH_NAMESPACE, XXH64_copyState) -# define XXH64_canonicalFromHash XXH_NAME2(XXH_NAMESPACE, XXH64_canonicalFromHash) -# define XXH64_hashFromCanonical XXH_NAME2(XXH_NAMESPACE, XXH64_hashFromCanonical) -#endif - - -/* ************************************* -* Version -***************************************/ -#define XXH_VERSION_MAJOR 0 -#define XXH_VERSION_MINOR 6 -#define XXH_VERSION_RELEASE 5 -#define XXH_VERSION_NUMBER (XXH_VERSION_MAJOR *100*100 + XXH_VERSION_MINOR *100 + XXH_VERSION_RELEASE) -XXH_PUBLIC_API unsigned XXH_versionNumber (void); - - -/*-********************************************************************** -* 32-bit hash -************************************************************************/ -typedef unsigned int XXH32_hash_t; - -/*! XXH32() : - Calculate the 32-bit hash of sequence "length" bytes stored at memory address "input". - The memory between input & input+length must be valid (allocated and read-accessible). - "seed" can be used to alter the result predictably. - Speed on Core 2 Duo @ 3 GHz (single thread, SMHasher benchmark) : 5.4 GB/s */ -XXH_PUBLIC_API XXH32_hash_t XXH32 (const void* input, size_t length, unsigned int seed); - -/*====== Streaming ======*/ -typedef struct XXH32_state_s XXH32_state_t; /* incomplete type */ -XXH_PUBLIC_API XXH32_state_t* XXH32_createState(void); -XXH_PUBLIC_API XXH_errorcode XXH32_freeState(XXH32_state_t* statePtr); -XXH_PUBLIC_API void XXH32_copyState(XXH32_state_t* dst_state, const XXH32_state_t* src_state); - -XXH_PUBLIC_API XXH_errorcode XXH32_reset (XXH32_state_t* statePtr, unsigned int seed); -XXH_PUBLIC_API XXH_errorcode XXH32_update (XXH32_state_t* statePtr, const void* input, size_t length); -XXH_PUBLIC_API XXH32_hash_t XXH32_digest (const XXH32_state_t* statePtr); - -/* - * Streaming functions generate the xxHash of an input provided in multiple segments. - * Note that, for small input, they are slower than single-call functions, due to state management. - * For small inputs, prefer `XXH32()` and `XXH64()`, which are better optimized. - * - * XXH state must first be allocated, using XXH*_createState() . - * - * Start a new hash by initializing state with a seed, using XXH*_reset(). - * - * Then, feed the hash state by calling XXH*_update() as many times as necessary. - * The function returns an error code, with 0 meaning OK, and any other value meaning there is an error. - * - * Finally, a hash value can be produced anytime, by using XXH*_digest(). - * This function returns the nn-bits hash as an int or long long. - * - * It's still possible to continue inserting input into the hash state after a digest, - * and generate some new hashes later on, by calling again XXH*_digest(). - * - * When done, free XXH state space if it was allocated dynamically. - */ - -/*====== Canonical representation ======*/ - -typedef struct { unsigned char digest[4]; } XXH32_canonical_t; -XXH_PUBLIC_API void XXH32_canonicalFromHash(XXH32_canonical_t* dst, XXH32_hash_t hash); -XXH_PUBLIC_API XXH32_hash_t XXH32_hashFromCanonical(const XXH32_canonical_t* src); - -/* Default result type for XXH functions are primitive unsigned 32 and 64 bits. - * The canonical representation uses human-readable write convention, aka big-endian (large digits first). - * These functions allow transformation of hash result into and from its canonical format. - * This way, hash values can be written into a file / memory, and remain comparable on different systems and programs. - */ - - -#ifndef XXH_NO_LONG_LONG -/*-********************************************************************** -* 64-bit hash -************************************************************************/ -typedef unsigned long long XXH64_hash_t; - -/*! XXH64() : - Calculate the 64-bit hash of sequence of length "len" stored at memory address "input". - "seed" can be used to alter the result predictably. - This function runs faster on 64-bit systems, but slower on 32-bit systems (see benchmark). -*/ -XXH_PUBLIC_API XXH64_hash_t XXH64 (const void* input, size_t length, unsigned long long seed); - -/*====== Streaming ======*/ -typedef struct XXH64_state_s XXH64_state_t; /* incomplete type */ -XXH_PUBLIC_API XXH64_state_t* XXH64_createState(void); -XXH_PUBLIC_API XXH_errorcode XXH64_freeState(XXH64_state_t* statePtr); -XXH_PUBLIC_API void XXH64_copyState(XXH64_state_t* dst_state, const XXH64_state_t* src_state); - -XXH_PUBLIC_API XXH_errorcode XXH64_reset (XXH64_state_t* statePtr, unsigned long long seed); -XXH_PUBLIC_API XXH_errorcode XXH64_update (XXH64_state_t* statePtr, const void* input, size_t length); -XXH_PUBLIC_API XXH64_hash_t XXH64_digest (const XXH64_state_t* statePtr); - -/*====== Canonical representation ======*/ -typedef struct { unsigned char digest[8]; } XXH64_canonical_t; -XXH_PUBLIC_API void XXH64_canonicalFromHash(XXH64_canonical_t* dst, XXH64_hash_t hash); -XXH_PUBLIC_API XXH64_hash_t XXH64_hashFromCanonical(const XXH64_canonical_t* src); -#endif /* XXH_NO_LONG_LONG */ - - - -#ifdef XXH_STATIC_LINKING_ONLY - -/* ================================================================================================ - This section contains declarations which are not guaranteed to remain stable. - They may change in future versions, becoming incompatible with a different version of the library. - These declarations should only be used with static linking. - Never use them in association with dynamic linking ! -=================================================================================================== */ - -/* These definitions are only present to allow - * static allocation of XXH state, on stack or in a struct for example. - * Never **ever** use members directly. */ - -#if !defined (__VMS) \ - && (defined (__cplusplus) \ - || (defined (__STDC_VERSION__) && (__STDC_VERSION__ >= 199901L) /* C99 */) ) -# include - -struct XXH32_state_s { - uint32_t total_len_32; - uint32_t large_len; - uint32_t v1; - uint32_t v2; - uint32_t v3; - uint32_t v4; - uint32_t mem32[4]; - uint32_t memsize; - uint32_t reserved; /* never read nor write, might be removed in a future version */ -}; /* typedef'd to XXH32_state_t */ - -struct XXH64_state_s { - uint64_t total_len; - uint64_t v1; - uint64_t v2; - uint64_t v3; - uint64_t v4; - uint64_t mem64[4]; - uint32_t memsize; - uint32_t reserved[2]; /* never read nor write, might be removed in a future version */ -}; /* typedef'd to XXH64_state_t */ - -# else - -struct XXH32_state_s { - unsigned total_len_32; - unsigned large_len; - unsigned v1; - unsigned v2; - unsigned v3; - unsigned v4; - unsigned mem32[4]; - unsigned memsize; - unsigned reserved; /* never read nor write, might be removed in a future version */ -}; /* typedef'd to XXH32_state_t */ - -# ifndef XXH_NO_LONG_LONG /* remove 64-bit support */ -struct XXH64_state_s { - unsigned long long total_len; - unsigned long long v1; - unsigned long long v2; - unsigned long long v3; - unsigned long long v4; - unsigned long long mem64[4]; - unsigned memsize; - unsigned reserved[2]; /* never read nor write, might be removed in a future version */ -}; /* typedef'd to XXH64_state_t */ -# endif - -# endif - - -#if defined(XXH_INLINE_ALL) || defined(XXH_PRIVATE_API) -# include "xxhash.c" /* include xxhash function bodies as `static`, for inlining */ -#endif - -#endif /* XXH_STATIC_LINKING_ONLY */ - - -#if defined (__cplusplus) -} -#endif - -#endif /* XXHASH_H_5627135585666179 */ diff --git a/source/client/test/clientTests.cpp b/source/client/test/clientTests.cpp index cb3c2f8c68..a3b1c4f779 100644 --- a/source/client/test/clientTests.cpp +++ b/source/client/test/clientTests.cpp @@ -912,7 +912,7 @@ TEST(clientCase, subscription_test) { tmq_conf_t* conf = tmq_conf_new(); tmq_conf_set(conf, "enable.auto.commit", "true"); tmq_conf_set(conf, "auto.commit.interval.ms", "1000"); - tmq_conf_set(conf, "group.id", "newabcdefgjhijlm__"); + tmq_conf_set(conf, "group.id", "consumer_group"); tmq_conf_set(conf, "td.connect.user", "root"); tmq_conf_set(conf, "td.connect.pass", "taosdata"); tmq_conf_set(conf, "auto.offset.reset", "earliest"); diff --git a/source/util/src/xxhash.c b/source/util/src/xxhash.c deleted file mode 100644 index ff28749e31..0000000000 --- a/source/util/src/xxhash.c +++ /dev/null @@ -1,1030 +0,0 @@ -/* -* xxHash - Fast Hash algorithm -* Copyright (C) 2012-2016, Yann Collet -* -* BSD 2-Clause License (http://www.opensource.org/licenses/bsd-license.php) -* -* Redistribution and use in source and binary forms, with or without -* modification, are permitted provided that the following conditions are -* met: -* -* * Redistributions of source code must retain the above copyright -* notice, this list of conditions and the following disclaimer. -* * Redistributions in binary form must reproduce the above -* copyright notice, this list of conditions and the following disclaimer -* in the documentation and/or other materials provided with the -* distribution. -* -* THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS -* "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT -* LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR -* A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT -* OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, -* SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT -* LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, -* DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY -* THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT -* (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE -* OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. -* -* You can contact the author at : -* - xxHash homepage: http://www.xxhash.com -* - xxHash source repository : https://github.com/Cyan4973/xxHash -*/ - - -/* ************************************* -* Tuning parameters -***************************************/ -/*!XXH_FORCE_MEMORY_ACCESS : - * By default, access to unaligned memory is controlled by `memcpy()`, which is safe and portable. - * Unfortunately, on some target/compiler combinations, the generated assembly is sub-optimal. - * The below switch allow to select different access method for improved performance. - * Method 0 (default) : use `memcpy()`. Safe and portable. - * Method 1 : `__packed` statement. It depends on compiler extension (ie, not portable). - * This method is safe if your compiler supports it, and *generally* as fast or faster than `memcpy`. - * Method 2 : direct access. This method doesn't depend on compiler but violate C standard. - * It can generate buggy code on targets which do not support unaligned memory accesses. - * But in some circumstances, it's the only known way to get the most performance (ie GCC + ARMv6) - * See http://stackoverflow.com/a/32095106/646947 for details. - * Prefer these methods in priority order (0 > 1 > 2) - */ -#ifndef XXH_FORCE_MEMORY_ACCESS /* can be defined externally, on command line for example */ -# if defined(__GNUC__) && ( defined(__ARM_ARCH_6__) || defined(__ARM_ARCH_6J__) \ - || defined(__ARM_ARCH_6K__) || defined(__ARM_ARCH_6Z__) \ - || defined(__ARM_ARCH_6ZK__) || defined(__ARM_ARCH_6T2__) ) -# define XXH_FORCE_MEMORY_ACCESS 2 -# elif (defined(__INTEL_COMPILER) && !defined(_WIN32)) || \ - (defined(__GNUC__) && ( defined(__ARM_ARCH_7__) || defined(__ARM_ARCH_7A__) \ - || defined(__ARM_ARCH_7R__) || defined(__ARM_ARCH_7M__) \ - || defined(__ARM_ARCH_7S__) )) -# define XXH_FORCE_MEMORY_ACCESS 1 -# endif -#endif - -/*!XXH_ACCEPT_NULL_INPUT_POINTER : - * If input pointer is NULL, xxHash default behavior is to dereference it, triggering a segfault. - * When this macro is enabled, xxHash actively checks input for null pointer. - * It it is, result for null input pointers is the same as a null-length input. - */ -#ifndef XXH_ACCEPT_NULL_INPUT_POINTER /* can be defined externally */ -# define XXH_ACCEPT_NULL_INPUT_POINTER 0 -#endif - -/*!XXH_FORCE_NATIVE_FORMAT : - * By default, xxHash library provides endian-independent Hash values, based on little-endian convention. - * Results are therefore identical for little-endian and big-endian CPU. - * This comes at a performance cost for big-endian CPU, since some swapping is required to emulate little-endian format. - * Should endian-independence be of no importance for your application, you may set the #define below to 1, - * to improve speed for Big-endian CPU. - * This option has no impact on Little_Endian CPU. - */ -#ifndef XXH_FORCE_NATIVE_FORMAT /* can be defined externally */ -# define XXH_FORCE_NATIVE_FORMAT 0 -#endif - -/*!XXH_FORCE_ALIGN_CHECK : - * This is a minor performance trick, only useful with lots of very small keys. - * It means : check for aligned/unaligned input. - * The check costs one initial branch per hash; - * set it to 0 when the input is guaranteed to be aligned, - * or when alignment doesn't matter for performance. - */ -#ifndef XXH_FORCE_ALIGN_CHECK /* can be defined externally */ -# if defined(__i386) || defined(_M_IX86) || defined(__x86_64__) || defined(_M_X64) -# define XXH_FORCE_ALIGN_CHECK 0 -# else -# define XXH_FORCE_ALIGN_CHECK 1 -# endif -#endif - - -/* ************************************* -* Includes & Memory related functions -***************************************/ -/*! Modify the local functions below should you wish to use some other memory routines -* for malloc(), free() */ -#include -static void* XXH_malloc(size_t s) { return malloc(s); } -static void XXH_free (void* p) { free(p); } -/*! and for memcpy() */ -#include -static void* XXH_memcpy(void* dest, const void* src, size_t size) { return memcpy(dest,src,size); } - -#include /* assert */ - -#define XXH_STATIC_LINKING_ONLY -#include "xxhash.h" - - -/* ************************************* -* Compiler Specific Options -***************************************/ -#ifdef _MSC_VER /* Visual Studio */ -# pragma warning(disable : 4127) /* disable: C4127: conditional expression is constant */ -# define FORCE_INLINE static __forceinline -#else -# if defined (__cplusplus) || defined (__STDC_VERSION__) && __STDC_VERSION__ >= 199901L /* C99 */ -# ifdef __GNUC__ -# define FORCE_INLINE static inline __attribute__((always_inline)) -# else -# define FORCE_INLINE static inline -# endif -# else -# define FORCE_INLINE static -# endif /* __STDC_VERSION__ */ -#endif - - -/* ************************************* -* Basic Types -***************************************/ -#ifndef MEM_MODULE -# if !defined (__VMS) \ - && (defined (__cplusplus) \ - || (defined (__STDC_VERSION__) && (__STDC_VERSION__ >= 199901L) /* C99 */) ) -# include - typedef uint8_t BYTE; - typedef uint16_t U16; - typedef uint32_t U32; -# else - typedef unsigned char BYTE; - typedef unsigned short U16; - typedef unsigned int U32; -# endif -#endif - -#if (defined(XXH_FORCE_MEMORY_ACCESS) && (XXH_FORCE_MEMORY_ACCESS==2)) - -/* Force direct memory access. Only works on CPU which support unaligned memory access in hardware */ -static U32 XXH_read32(const void* memPtr) { return *(const U32*) memPtr; } - -#elif (defined(XXH_FORCE_MEMORY_ACCESS) && (XXH_FORCE_MEMORY_ACCESS==1)) - -/* __pack instructions are safer, but compiler specific, hence potentially problematic for some compilers */ -/* currently only defined for gcc and icc */ -typedef union { U32 u32; } __attribute__((packed)) unalign; -static U32 XXH_read32(const void* ptr) { return ((const unalign*)ptr)->u32; } - -#else - -/* portable and safe solution. Generally efficient. - * see : http://stackoverflow.com/a/32095106/646947 - */ -static U32 XXH_read32(const void* memPtr) -{ - U32 val; - memcpy(&val, memPtr, sizeof(val)); - return val; -} - -#endif /* XXH_FORCE_DIRECT_MEMORY_ACCESS */ - - -/* **************************************** -* Compiler-specific Functions and Macros -******************************************/ -#define XXH_GCC_VERSION (__GNUC__ * 100 + __GNUC_MINOR__) - -/* Note : although _rotl exists for minGW (GCC under windows), performance seems poor */ -#if defined(_MSC_VER) -# define XXH_rotl32(x,r) _rotl(x,r) -# define XXH_rotl64(x,r) _rotl64(x,r) -#else -# define XXH_rotl32(x,r) ((x << r) | (x >> (32 - r))) -# define XXH_rotl64(x,r) ((x << r) | (x >> (64 - r))) -#endif - -#if defined(_MSC_VER) /* Visual Studio */ -# define XXH_swap32 _byteswap_ulong -#elif XXH_GCC_VERSION >= 403 -# define XXH_swap32 __builtin_bswap32 -#else -static U32 XXH_swap32 (U32 x) -{ - return ((x << 24) & 0xff000000 ) | - ((x << 8) & 0x00ff0000 ) | - ((x >> 8) & 0x0000ff00 ) | - ((x >> 24) & 0x000000ff ); -} -#endif - - -/* ************************************* -* Architecture Macros -***************************************/ -typedef enum { XXH_bigEndian=0, XXH_littleEndian=1 } XXH_endianess; - -/* XXH_CPU_LITTLE_ENDIAN can be defined externally, for example on the compiler command line */ -#ifndef XXH_CPU_LITTLE_ENDIAN -static int XXH_isLittleEndian(void) -{ - const union { U32 u; BYTE c[4]; } one = { 1 }; /* don't use static : performance detrimental */ - return one.c[0]; -} -# define XXH_CPU_LITTLE_ENDIAN XXH_isLittleEndian() -#endif - - -/* *************************** -* Memory reads -*****************************/ -typedef enum { XXH_aligned, XXH_unaligned } XXH_alignment; - -FORCE_INLINE U32 XXH_readLE32_align(const void* ptr, XXH_endianess endian, XXH_alignment align) -{ - if (align==XXH_unaligned) - return endian==XXH_littleEndian ? XXH_read32(ptr) : XXH_swap32(XXH_read32(ptr)); - else - return endian==XXH_littleEndian ? *(const U32*)ptr : XXH_swap32(*(const U32*)ptr); -} - -FORCE_INLINE U32 XXH_readLE32(const void* ptr, XXH_endianess endian) -{ - return XXH_readLE32_align(ptr, endian, XXH_unaligned); -} - -static U32 XXH_readBE32(const void* ptr) -{ - return XXH_CPU_LITTLE_ENDIAN ? XXH_swap32(XXH_read32(ptr)) : XXH_read32(ptr); -} - - -/* ************************************* -* Macros -***************************************/ -#define XXH_STATIC_ASSERT(c) { enum { XXH_sa = 1/(int)(!!(c)) }; } /* use after variable declarations */ -XXH_PUBLIC_API unsigned XXH_versionNumber (void) { return XXH_VERSION_NUMBER; } - - -/* ******************************************************************* -* 32-bit hash functions -*********************************************************************/ -static const U32 PRIME32_1 = 2654435761U; -static const U32 PRIME32_2 = 2246822519U; -static const U32 PRIME32_3 = 3266489917U; -static const U32 PRIME32_4 = 668265263U; -static const U32 PRIME32_5 = 374761393U; - -static U32 XXH32_round(U32 seed, U32 input) -{ - seed += input * PRIME32_2; - seed = XXH_rotl32(seed, 13); - seed *= PRIME32_1; - return seed; -} - -/* mix all bits */ -static U32 XXH32_avalanche(U32 h32) -{ - h32 ^= h32 >> 15; - h32 *= PRIME32_2; - h32 ^= h32 >> 13; - h32 *= PRIME32_3; - h32 ^= h32 >> 16; - return(h32); -} - -#define XXH_get32bits(p) XXH_readLE32_align(p, endian, align) - -static U32 -XXH32_finalize(U32 h32, const void* ptr, size_t len, - XXH_endianess endian, XXH_alignment align) - -{ - const BYTE* p = (const BYTE*)ptr; - -#define PROCESS1 \ - h32 += (*p++) * PRIME32_5; \ - h32 = XXH_rotl32(h32, 11) * PRIME32_1 ; - -#define PROCESS4 \ - h32 += XXH_get32bits(p) * PRIME32_3; \ - p+=4; \ - h32 = XXH_rotl32(h32, 17) * PRIME32_4 ; - - switch(len&15) /* or switch(bEnd - p) */ - { - case 12: PROCESS4; - /* fallthrough */ - case 8: PROCESS4; - /* fallthrough */ - case 4: PROCESS4; - return XXH32_avalanche(h32); - - case 13: PROCESS4; - /* fallthrough */ - case 9: PROCESS4; - /* fallthrough */ - case 5: PROCESS4; - PROCESS1; - return XXH32_avalanche(h32); - - case 14: PROCESS4; - /* fallthrough */ - case 10: PROCESS4; - /* fallthrough */ - case 6: PROCESS4; - PROCESS1; - PROCESS1; - return XXH32_avalanche(h32); - - case 15: PROCESS4; - /* fallthrough */ - case 11: PROCESS4; - /* fallthrough */ - case 7: PROCESS4; - /* fallthrough */ - case 3: PROCESS1; - /* fallthrough */ - case 2: PROCESS1; - /* fallthrough */ - case 1: PROCESS1; - /* fallthrough */ - case 0: return XXH32_avalanche(h32); - } - assert(0); - return h32; /* reaching this point is deemed impossible */ -} - - -FORCE_INLINE U32 -XXH32_endian_align(const void* input, size_t len, U32 seed, - XXH_endianess endian, XXH_alignment align) -{ - const BYTE* p = (const BYTE*)input; - const BYTE* bEnd = p + len; - U32 h32; - -#if defined(XXH_ACCEPT_NULL_INPUT_POINTER) && (XXH_ACCEPT_NULL_INPUT_POINTER>=1) - if (p==NULL) { - len=0; - bEnd=p=(const BYTE*)(size_t)16; - } -#endif - - if (len>=16) { - const BYTE* const limit = bEnd - 15; - U32 v1 = seed + PRIME32_1 + PRIME32_2; - U32 v2 = seed + PRIME32_2; - U32 v3 = seed + 0; - U32 v4 = seed - PRIME32_1; - - do { - v1 = XXH32_round(v1, XXH_get32bits(p)); p+=4; - v2 = XXH32_round(v2, XXH_get32bits(p)); p+=4; - v3 = XXH32_round(v3, XXH_get32bits(p)); p+=4; - v4 = XXH32_round(v4, XXH_get32bits(p)); p+=4; - } while (p < limit); - - h32 = XXH_rotl32(v1, 1) + XXH_rotl32(v2, 7) - + XXH_rotl32(v3, 12) + XXH_rotl32(v4, 18); - } else { - h32 = seed + PRIME32_5; - } - - h32 += (U32)len; - - return XXH32_finalize(h32, p, len&15, endian, align); -} - - -XXH_PUBLIC_API unsigned int XXH32 (const void* input, size_t len, unsigned int seed) -{ -#if 0 - /* Simple version, good for code maintenance, but unfortunately slow for small inputs */ - XXH32_state_t state; - XXH32_reset(&state, seed); - XXH32_update(&state, input, len); - return XXH32_digest(&state); -#else - XXH_endianess endian_detected = (XXH_endianess)XXH_CPU_LITTLE_ENDIAN; - - if (XXH_FORCE_ALIGN_CHECK) { - if ((((size_t)input) & 3) == 0) { /* Input is 4-bytes aligned, leverage the speed benefit */ - if ((endian_detected==XXH_littleEndian) || XXH_FORCE_NATIVE_FORMAT) - return XXH32_endian_align(input, len, seed, XXH_littleEndian, XXH_aligned); - else - return XXH32_endian_align(input, len, seed, XXH_bigEndian, XXH_aligned); - } } - - if ((endian_detected==XXH_littleEndian) || XXH_FORCE_NATIVE_FORMAT) - return XXH32_endian_align(input, len, seed, XXH_littleEndian, XXH_unaligned); - else - return XXH32_endian_align(input, len, seed, XXH_bigEndian, XXH_unaligned); -#endif -} - - - -/*====== Hash streaming ======*/ - -XXH_PUBLIC_API XXH32_state_t* XXH32_createState(void) -{ - return (XXH32_state_t*)XXH_malloc(sizeof(XXH32_state_t)); -} -XXH_PUBLIC_API XXH_errorcode XXH32_freeState(XXH32_state_t* statePtr) -{ - XXH_free(statePtr); - return XXH_OK; -} - -XXH_PUBLIC_API void XXH32_copyState(XXH32_state_t* dstState, const XXH32_state_t* srcState) -{ - memcpy(dstState, srcState, sizeof(*dstState)); -} - -XXH_PUBLIC_API XXH_errorcode XXH32_reset(XXH32_state_t* statePtr, unsigned int seed) -{ - XXH32_state_t state; /* using a local state to memcpy() in order to avoid strict-aliasing warnings */ - memset(&state, 0, sizeof(state)); - state.v1 = seed + PRIME32_1 + PRIME32_2; - state.v2 = seed + PRIME32_2; - state.v3 = seed + 0; - state.v4 = seed - PRIME32_1; - /* do not write into reserved, planned to be removed in a future version */ - memcpy(statePtr, &state, sizeof(state) - sizeof(state.reserved)); - return XXH_OK; -} - - -FORCE_INLINE XXH_errorcode -XXH32_update_endian(XXH32_state_t* state, const void* input, size_t len, XXH_endianess endian) -{ - if (input==NULL) -#if defined(XXH_ACCEPT_NULL_INPUT_POINTER) && (XXH_ACCEPT_NULL_INPUT_POINTER>=1) - return XXH_OK; -#else - return XXH_ERROR; -#endif - - { const BYTE* p = (const BYTE*)input; - const BYTE* const bEnd = p + len; - - state->total_len_32 += (unsigned)len; - state->large_len |= (len>=16) | (state->total_len_32>=16); - - if (state->memsize + len < 16) { /* fill in tmp buffer */ - XXH_memcpy((BYTE*)(state->mem32) + state->memsize, input, len); - state->memsize += (unsigned)len; - return XXH_OK; - } - - if (state->memsize) { /* some data left from previous update */ - XXH_memcpy((BYTE*)(state->mem32) + state->memsize, input, 16-state->memsize); - { const U32* p32 = state->mem32; - state->v1 = XXH32_round(state->v1, XXH_readLE32(p32, endian)); p32++; - state->v2 = XXH32_round(state->v2, XXH_readLE32(p32, endian)); p32++; - state->v3 = XXH32_round(state->v3, XXH_readLE32(p32, endian)); p32++; - state->v4 = XXH32_round(state->v4, XXH_readLE32(p32, endian)); - } - p += 16-state->memsize; - state->memsize = 0; - } - - if (p <= bEnd-16) { - const BYTE* const limit = bEnd - 16; - U32 v1 = state->v1; - U32 v2 = state->v2; - U32 v3 = state->v3; - U32 v4 = state->v4; - - do { - v1 = XXH32_round(v1, XXH_readLE32(p, endian)); p+=4; - v2 = XXH32_round(v2, XXH_readLE32(p, endian)); p+=4; - v3 = XXH32_round(v3, XXH_readLE32(p, endian)); p+=4; - v4 = XXH32_round(v4, XXH_readLE32(p, endian)); p+=4; - } while (p<=limit); - - state->v1 = v1; - state->v2 = v2; - state->v3 = v3; - state->v4 = v4; - } - - if (p < bEnd) { - XXH_memcpy(state->mem32, p, (size_t)(bEnd-p)); - state->memsize = (unsigned)(bEnd-p); - } - } - - return XXH_OK; -} - - -XXH_PUBLIC_API XXH_errorcode XXH32_update (XXH32_state_t* state_in, const void* input, size_t len) -{ - XXH_endianess endian_detected = (XXH_endianess)XXH_CPU_LITTLE_ENDIAN; - - if ((endian_detected==XXH_littleEndian) || XXH_FORCE_NATIVE_FORMAT) - return XXH32_update_endian(state_in, input, len, XXH_littleEndian); - else - return XXH32_update_endian(state_in, input, len, XXH_bigEndian); -} - - -FORCE_INLINE U32 -XXH32_digest_endian (const XXH32_state_t* state, XXH_endianess endian) -{ - U32 h32; - - if (state->large_len) { - h32 = XXH_rotl32(state->v1, 1) - + XXH_rotl32(state->v2, 7) - + XXH_rotl32(state->v3, 12) - + XXH_rotl32(state->v4, 18); - } else { - h32 = state->v3 /* == seed */ + PRIME32_5; - } - - h32 += state->total_len_32; - - return XXH32_finalize(h32, state->mem32, state->memsize, endian, XXH_aligned); -} - - -XXH_PUBLIC_API unsigned int XXH32_digest (const XXH32_state_t* state_in) -{ - XXH_endianess endian_detected = (XXH_endianess)XXH_CPU_LITTLE_ENDIAN; - - if ((endian_detected==XXH_littleEndian) || XXH_FORCE_NATIVE_FORMAT) - return XXH32_digest_endian(state_in, XXH_littleEndian); - else - return XXH32_digest_endian(state_in, XXH_bigEndian); -} - - -/*====== Canonical representation ======*/ - -/*! Default XXH result types are basic unsigned 32 and 64 bits. -* The canonical representation follows human-readable write convention, aka big-endian (large digits first). -* These functions allow transformation of hash result into and from its canonical format. -* This way, hash values can be written into a file or buffer, remaining comparable across different systems. -*/ - -XXH_PUBLIC_API void XXH32_canonicalFromHash(XXH32_canonical_t* dst, XXH32_hash_t hash) -{ - XXH_STATIC_ASSERT(sizeof(XXH32_canonical_t) == sizeof(XXH32_hash_t)); - if (XXH_CPU_LITTLE_ENDIAN) hash = XXH_swap32(hash); - memcpy(dst, &hash, sizeof(*dst)); -} - -XXH_PUBLIC_API XXH32_hash_t XXH32_hashFromCanonical(const XXH32_canonical_t* src) -{ - return XXH_readBE32(src); -} - - -#ifndef XXH_NO_LONG_LONG - -/* ******************************************************************* -* 64-bit hash functions -*********************************************************************/ - -/*====== Memory access ======*/ - -#ifndef MEM_MODULE -# define MEM_MODULE -# if !defined (__VMS) \ - && (defined (__cplusplus) \ - || (defined (__STDC_VERSION__) && (__STDC_VERSION__ >= 199901L) /* C99 */) ) -# include - typedef uint64_t U64; -# else - /* if compiler doesn't support unsigned long long, replace by another 64-bit type */ - typedef unsigned long long U64; -# endif -#endif - - -#if (defined(XXH_FORCE_MEMORY_ACCESS) && (XXH_FORCE_MEMORY_ACCESS==2)) - -/* Force direct memory access. Only works on CPU which support unaligned memory access in hardware */ -static U64 XXH_read64(const void* memPtr) { return *(const U64*) memPtr; } - -#elif (defined(XXH_FORCE_MEMORY_ACCESS) && (XXH_FORCE_MEMORY_ACCESS==1)) - -/* __pack instructions are safer, but compiler specific, hence potentially problematic for some compilers */ -/* currently only defined for gcc and icc */ -typedef union { U32 u32; U64 u64; } __attribute__((packed)) unalign64; -static U64 XXH_read64(const void* ptr) { return ((const unalign64*)ptr)->u64; } - -#else - -/* portable and safe solution. Generally efficient. - * see : http://stackoverflow.com/a/32095106/646947 - */ - -static U64 XXH_read64(const void* memPtr) -{ - U64 val; - memcpy(&val, memPtr, sizeof(val)); - return val; -} - -#endif /* XXH_FORCE_DIRECT_MEMORY_ACCESS */ - -#if defined(_MSC_VER) /* Visual Studio */ -# define XXH_swap64 _byteswap_uint64 -#elif XXH_GCC_VERSION >= 403 -# define XXH_swap64 __builtin_bswap64 -#else -static U64 XXH_swap64 (U64 x) -{ - return ((x << 56) & 0xff00000000000000ULL) | - ((x << 40) & 0x00ff000000000000ULL) | - ((x << 24) & 0x0000ff0000000000ULL) | - ((x << 8) & 0x000000ff00000000ULL) | - ((x >> 8) & 0x00000000ff000000ULL) | - ((x >> 24) & 0x0000000000ff0000ULL) | - ((x >> 40) & 0x000000000000ff00ULL) | - ((x >> 56) & 0x00000000000000ffULL); -} -#endif - -FORCE_INLINE U64 XXH_readLE64_align(const void* ptr, XXH_endianess endian, XXH_alignment align) -{ - if (align==XXH_unaligned) - return endian==XXH_littleEndian ? XXH_read64(ptr) : XXH_swap64(XXH_read64(ptr)); - else - return endian==XXH_littleEndian ? *(const U64*)ptr : XXH_swap64(*(const U64*)ptr); -} - -FORCE_INLINE U64 XXH_readLE64(const void* ptr, XXH_endianess endian) -{ - return XXH_readLE64_align(ptr, endian, XXH_unaligned); -} - -static U64 XXH_readBE64(const void* ptr) -{ - return XXH_CPU_LITTLE_ENDIAN ? XXH_swap64(XXH_read64(ptr)) : XXH_read64(ptr); -} - - -/*====== xxh64 ======*/ - -static const U64 PRIME64_1 = 11400714785074694791ULL; -static const U64 PRIME64_2 = 14029467366897019727ULL; -static const U64 PRIME64_3 = 1609587929392839161ULL; -static const U64 PRIME64_4 = 9650029242287828579ULL; -static const U64 PRIME64_5 = 2870177450012600261ULL; - -static U64 XXH64_round(U64 acc, U64 input) -{ - acc += input * PRIME64_2; - acc = XXH_rotl64(acc, 31); - acc *= PRIME64_1; - return acc; -} - -static U64 XXH64_mergeRound(U64 acc, U64 val) -{ - val = XXH64_round(0, val); - acc ^= val; - acc = acc * PRIME64_1 + PRIME64_4; - return acc; -} - -static U64 XXH64_avalanche(U64 h64) -{ - h64 ^= h64 >> 33; - h64 *= PRIME64_2; - h64 ^= h64 >> 29; - h64 *= PRIME64_3; - h64 ^= h64 >> 32; - return h64; -} - - -#define XXH_get64bits(p) XXH_readLE64_align(p, endian, align) - -static U64 -XXH64_finalize(U64 h64, const void* ptr, size_t len, - XXH_endianess endian, XXH_alignment align) -{ - const BYTE* p = (const BYTE*)ptr; - -#define PROCESS1_64 \ - h64 ^= (*p++) * PRIME64_5; \ - h64 = XXH_rotl64(h64, 11) * PRIME64_1; - -#define PROCESS4_64 \ - h64 ^= (U64)(XXH_get32bits(p)) * PRIME64_1; \ - p+=4; \ - h64 = XXH_rotl64(h64, 23) * PRIME64_2 + PRIME64_3; - -#define PROCESS8_64 { \ - U64 const k1 = XXH64_round(0, XXH_get64bits(p)); \ - p+=8; \ - h64 ^= k1; \ - h64 = XXH_rotl64(h64,27) * PRIME64_1 + PRIME64_4; \ -} - - switch(len&31) { - case 24: PROCESS8_64; - /* fallthrough */ - case 16: PROCESS8_64; - /* fallthrough */ - case 8: PROCESS8_64; - return XXH64_avalanche(h64); - - case 28: PROCESS8_64; - /* fallthrough */ - case 20: PROCESS8_64; - /* fallthrough */ - case 12: PROCESS8_64; - /* fallthrough */ - case 4: PROCESS4_64; - return XXH64_avalanche(h64); - - case 25: PROCESS8_64; - /* fallthrough */ - case 17: PROCESS8_64; - /* fallthrough */ - case 9: PROCESS8_64; - PROCESS1_64; - return XXH64_avalanche(h64); - - case 29: PROCESS8_64; - /* fallthrough */ - case 21: PROCESS8_64; - /* fallthrough */ - case 13: PROCESS8_64; - /* fallthrough */ - case 5: PROCESS4_64; - PROCESS1_64; - return XXH64_avalanche(h64); - - case 26: PROCESS8_64; - /* fallthrough */ - case 18: PROCESS8_64; - /* fallthrough */ - case 10: PROCESS8_64; - PROCESS1_64; - PROCESS1_64; - return XXH64_avalanche(h64); - - case 30: PROCESS8_64; - /* fallthrough */ - case 22: PROCESS8_64; - /* fallthrough */ - case 14: PROCESS8_64; - /* fallthrough */ - case 6: PROCESS4_64; - PROCESS1_64; - PROCESS1_64; - return XXH64_avalanche(h64); - - case 27: PROCESS8_64; - /* fallthrough */ - case 19: PROCESS8_64; - /* fallthrough */ - case 11: PROCESS8_64; - PROCESS1_64; - PROCESS1_64; - PROCESS1_64; - return XXH64_avalanche(h64); - - case 31: PROCESS8_64; - /* fallthrough */ - case 23: PROCESS8_64; - /* fallthrough */ - case 15: PROCESS8_64; - /* fallthrough */ - case 7: PROCESS4_64; - /* fallthrough */ - case 3: PROCESS1_64; - /* fallthrough */ - case 2: PROCESS1_64; - /* fallthrough */ - case 1: PROCESS1_64; - /* fallthrough */ - case 0: return XXH64_avalanche(h64); - } - - /* impossible to reach */ - assert(0); - return 0; /* unreachable, but some compilers complain without it */ -} - -FORCE_INLINE U64 -XXH64_endian_align(const void* input, size_t len, U64 seed, - XXH_endianess endian, XXH_alignment align) -{ - const BYTE* p = (const BYTE*)input; - const BYTE* bEnd = p + len; - U64 h64; - -#if defined(XXH_ACCEPT_NULL_INPUT_POINTER) && (XXH_ACCEPT_NULL_INPUT_POINTER>=1) - if (p==NULL) { - len=0; - bEnd=p=(const BYTE*)(size_t)32; - } -#endif - - if (len>=32) { - const BYTE* const limit = bEnd - 32; - U64 v1 = seed + PRIME64_1 + PRIME64_2; - U64 v2 = seed + PRIME64_2; - U64 v3 = seed + 0; - U64 v4 = seed - PRIME64_1; - - do { - v1 = XXH64_round(v1, XXH_get64bits(p)); p+=8; - v2 = XXH64_round(v2, XXH_get64bits(p)); p+=8; - v3 = XXH64_round(v3, XXH_get64bits(p)); p+=8; - v4 = XXH64_round(v4, XXH_get64bits(p)); p+=8; - } while (p<=limit); - - h64 = XXH_rotl64(v1, 1) + XXH_rotl64(v2, 7) + XXH_rotl64(v3, 12) + XXH_rotl64(v4, 18); - h64 = XXH64_mergeRound(h64, v1); - h64 = XXH64_mergeRound(h64, v2); - h64 = XXH64_mergeRound(h64, v3); - h64 = XXH64_mergeRound(h64, v4); - - } else { - h64 = seed + PRIME64_5; - } - - h64 += (U64) len; - - return XXH64_finalize(h64, p, len, endian, align); -} - - -XXH_PUBLIC_API unsigned long long XXH64 (const void* input, size_t len, unsigned long long seed) -{ -#if 0 - /* Simple version, good for code maintenance, but unfortunately slow for small inputs */ - XXH64_state_t state; - XXH64_reset(&state, seed); - XXH64_update(&state, input, len); - return XXH64_digest(&state); -#else - XXH_endianess endian_detected = (XXH_endianess)XXH_CPU_LITTLE_ENDIAN; - - if (XXH_FORCE_ALIGN_CHECK) { - if ((((size_t)input) & 7)==0) { /* Input is aligned, let's leverage the speed advantage */ - if ((endian_detected==XXH_littleEndian) || XXH_FORCE_NATIVE_FORMAT) - return XXH64_endian_align(input, len, seed, XXH_littleEndian, XXH_aligned); - else - return XXH64_endian_align(input, len, seed, XXH_bigEndian, XXH_aligned); - } } - - if ((endian_detected==XXH_littleEndian) || XXH_FORCE_NATIVE_FORMAT) - return XXH64_endian_align(input, len, seed, XXH_littleEndian, XXH_unaligned); - else - return XXH64_endian_align(input, len, seed, XXH_bigEndian, XXH_unaligned); -#endif -} - -/*====== Hash Streaming ======*/ - -XXH_PUBLIC_API XXH64_state_t* XXH64_createState(void) -{ - return (XXH64_state_t*)XXH_malloc(sizeof(XXH64_state_t)); -} -XXH_PUBLIC_API XXH_errorcode XXH64_freeState(XXH64_state_t* statePtr) -{ - XXH_free(statePtr); - return XXH_OK; -} - -XXH_PUBLIC_API void XXH64_copyState(XXH64_state_t* dstState, const XXH64_state_t* srcState) -{ - memcpy(dstState, srcState, sizeof(*dstState)); -} - -XXH_PUBLIC_API XXH_errorcode XXH64_reset(XXH64_state_t* statePtr, unsigned long long seed) -{ - XXH64_state_t state; /* using a local state to memcpy() in order to avoid strict-aliasing warnings */ - memset(&state, 0, sizeof(state)); - state.v1 = seed + PRIME64_1 + PRIME64_2; - state.v2 = seed + PRIME64_2; - state.v3 = seed + 0; - state.v4 = seed - PRIME64_1; - /* do not write into reserved, planned to be removed in a future version */ - memcpy(statePtr, &state, sizeof(state) - sizeof(state.reserved)); - return XXH_OK; -} - -FORCE_INLINE XXH_errorcode -XXH64_update_endian (XXH64_state_t* state, const void* input, size_t len, XXH_endianess endian) -{ - if (input==NULL) -#if defined(XXH_ACCEPT_NULL_INPUT_POINTER) && (XXH_ACCEPT_NULL_INPUT_POINTER>=1) - return XXH_OK; -#else - return XXH_ERROR; -#endif - - { const BYTE* p = (const BYTE*)input; - const BYTE* const bEnd = p + len; - - state->total_len += len; - - if (state->memsize + len < 32) { /* fill in tmp buffer */ - XXH_memcpy(((BYTE*)state->mem64) + state->memsize, input, len); - state->memsize += (U32)len; - return XXH_OK; - } - - if (state->memsize) { /* tmp buffer is full */ - XXH_memcpy(((BYTE*)state->mem64) + state->memsize, input, 32-state->memsize); - state->v1 = XXH64_round(state->v1, XXH_readLE64(state->mem64+0, endian)); - state->v2 = XXH64_round(state->v2, XXH_readLE64(state->mem64+1, endian)); - state->v3 = XXH64_round(state->v3, XXH_readLE64(state->mem64+2, endian)); - state->v4 = XXH64_round(state->v4, XXH_readLE64(state->mem64+3, endian)); - p += 32-state->memsize; - state->memsize = 0; - } - - if (p+32 <= bEnd) { - const BYTE* const limit = bEnd - 32; - U64 v1 = state->v1; - U64 v2 = state->v2; - U64 v3 = state->v3; - U64 v4 = state->v4; - - do { - v1 = XXH64_round(v1, XXH_readLE64(p, endian)); p+=8; - v2 = XXH64_round(v2, XXH_readLE64(p, endian)); p+=8; - v3 = XXH64_round(v3, XXH_readLE64(p, endian)); p+=8; - v4 = XXH64_round(v4, XXH_readLE64(p, endian)); p+=8; - } while (p<=limit); - - state->v1 = v1; - state->v2 = v2; - state->v3 = v3; - state->v4 = v4; - } - - if (p < bEnd) { - XXH_memcpy(state->mem64, p, (size_t)(bEnd-p)); - state->memsize = (unsigned)(bEnd-p); - } - } - - return XXH_OK; -} - -XXH_PUBLIC_API XXH_errorcode XXH64_update (XXH64_state_t* state_in, const void* input, size_t len) -{ - XXH_endianess endian_detected = (XXH_endianess)XXH_CPU_LITTLE_ENDIAN; - - if ((endian_detected==XXH_littleEndian) || XXH_FORCE_NATIVE_FORMAT) - return XXH64_update_endian(state_in, input, len, XXH_littleEndian); - else - return XXH64_update_endian(state_in, input, len, XXH_bigEndian); -} - -FORCE_INLINE U64 XXH64_digest_endian (const XXH64_state_t* state, XXH_endianess endian) -{ - U64 h64; - - if (state->total_len >= 32) { - U64 const v1 = state->v1; - U64 const v2 = state->v2; - U64 const v3 = state->v3; - U64 const v4 = state->v4; - - h64 = XXH_rotl64(v1, 1) + XXH_rotl64(v2, 7) + XXH_rotl64(v3, 12) + XXH_rotl64(v4, 18); - h64 = XXH64_mergeRound(h64, v1); - h64 = XXH64_mergeRound(h64, v2); - h64 = XXH64_mergeRound(h64, v3); - h64 = XXH64_mergeRound(h64, v4); - } else { - h64 = state->v3 /*seed*/ + PRIME64_5; - } - - h64 += (U64) state->total_len; - - return XXH64_finalize(h64, state->mem64, (size_t)state->total_len, endian, XXH_aligned); -} - -XXH_PUBLIC_API unsigned long long XXH64_digest (const XXH64_state_t* state_in) -{ - XXH_endianess endian_detected = (XXH_endianess)XXH_CPU_LITTLE_ENDIAN; - - if ((endian_detected==XXH_littleEndian) || XXH_FORCE_NATIVE_FORMAT) - return XXH64_digest_endian(state_in, XXH_littleEndian); - else - return XXH64_digest_endian(state_in, XXH_bigEndian); -} - - -/*====== Canonical representation ======*/ - -XXH_PUBLIC_API void XXH64_canonicalFromHash(XXH64_canonical_t* dst, XXH64_hash_t hash) -{ - XXH_STATIC_ASSERT(sizeof(XXH64_canonical_t) == sizeof(XXH64_hash_t)); - if (XXH_CPU_LITTLE_ENDIAN) hash = XXH_swap64(hash); - memcpy(dst, &hash, sizeof(*dst)); -} - -XXH_PUBLIC_API XXH64_hash_t XXH64_hashFromCanonical(const XXH64_canonical_t* src) -{ - return XXH_readBE64(src); -} - -#endif /* XXH_NO_LONG_LONG */ From e10ec8415439147b7f16e3637a3a4b3b809810d3 Mon Sep 17 00:00:00 2001 From: Haojun Liao Date: Mon, 20 Feb 2023 14:52:16 +0800 Subject: [PATCH 03/31] refactor: do some internal refactor. --- source/client/src/clientTmq.c | 26 +++++++++++++++++--------- 1 file changed, 17 insertions(+), 9 deletions(-) diff --git a/source/client/src/clientTmq.c b/source/client/src/clientTmq.c index b9b66e1222..fb8cdcfde3 100644 --- a/source/client/src/clientTmq.c +++ b/source/client/src/clientTmq.c @@ -458,6 +458,7 @@ static int32_t tmqSendCommitReq(tmq_t* tmq, SMqClientVg* pVg, SMqClientTopic* pT terrno = TSDB_CODE_OUT_OF_MEMORY; return -1; } + pOffset->val = pVg->currentOffset; int32_t groupLen = strlen(tmq->groupId); @@ -471,11 +472,13 @@ static int32_t tmqSendCommitReq(tmq_t* tmq, SMqClientVg* pVg, SMqClientTopic* pT if (code < 0) { return -1; } + void* buf = taosMemoryCalloc(1, sizeof(SMsgHead) + len); if (buf == NULL) { taosMemoryFree(pOffset); return -1; } + ((SMsgHead*)buf)->vgId = htonl(pVg->vgId); void* abuf = POINTER_SHIFT(buf, sizeof(SMsgHead)); @@ -492,6 +495,7 @@ static int32_t tmqSendCommitReq(tmq_t* tmq, SMqClientVg* pVg, SMqClientTopic* pT taosMemoryFree(buf); return -1; } + pParam->params = pParamSet; pParam->pOffset = pOffset; @@ -503,14 +507,16 @@ static int32_t tmqSendCommitReq(tmq_t* tmq, SMqClientVg* pVg, SMqClientTopic* pT taosMemoryFree(pParam); return -1; } + pMsgSendInfo->msgInfo = (SDataBuf){ .pData = buf, .len = sizeof(SMsgHead) + len, .handle = NULL, }; - tscDebug("consumer:0x%" PRIx64 " topic:%s on vgId:%d offset:%" PRId64, tmq->consumerId, pOffset->subKey, - pVg->vgId, pOffset->val.version); + SEp* pEp = &pVg->epSet.eps[pVg->epSet.inUse]; + tscDebug("consumer:0x%" PRIx64 " topic:%s on vgId:%d offset:%" PRId64" prev:%"PRId64", ep:%s:%d", tmq->consumerId, pOffset->subKey, + pVg->vgId, pOffset->val.version, pVg->committedOffset.version, pEp->fqdn, pEp->port); // TODO: put into cb pVg->committedOffset = pVg->currentOffset; @@ -637,15 +643,16 @@ static int32_t tmqCommitConsumerImpl(tmq_t* tmq, int8_t automatic, int8_t async, // init as 1 to prevent concurrency issue pParamSet->waitingRspNum = 1; - for (int32_t i = 0; i < taosArrayGetSize(tmq->clientTopics); i++) { + int32_t numOfTopics = taosArrayGetSize(tmq->clientTopics); + tscDebug("consumer:0x%"PRIx64" start to commit offset for %d topics", tmq->consumerId, numOfTopics); + + for (int32_t i = 0; i < numOfTopics; i++) { SMqClientTopic* pTopic = taosArrayGet(tmq->clientTopics, i); int32_t numOfVgroups = taosArrayGetSize(pTopic->vgs); for (int32_t j = 0; j < numOfVgroups; j++) { SMqClientVg* pVg = taosArrayGet(pTopic->vgs, j); if (pVg->currentOffset.type > 0 && !tOffsetEqual(&pVg->currentOffset, &pVg->committedOffset)) { - tscDebug("consumer:0x%" PRIx64 " topic:%s vgId:%d, current %" PRId64 ", committed %" PRId64, tmq->consumerId, - pTopic->topicName, pVg->vgId, pVg->currentOffset.version, pVg->committedOffset.version); if (tmqSendCommitReq(tmq, pVg, pTopic, pParamSet) < 0) { continue; } @@ -1085,7 +1092,7 @@ int32_t tmq_subscribe(tmq_t* tmq, const tmq_list_t* topic_list) { } tNameExtractFullName(&name, topicFName); - tscDebug("consumer:0x%"PRIx64", subscribe topic: %s", tmq->consumerId, topicFName); + tscDebug("consumer:0x%"PRIx64" subscribe topic: %s", tmq->consumerId, topicFName); taosArrayPush(req.topicNames, &topicFName); } @@ -1398,7 +1405,7 @@ bool tmqUpdateEp(tmq_t* tmq, int32_t epoch, const SMqAskEpRsp* pRsp) { } atomic_store_32(&tmq->epoch, epoch); - tscDebug("consumer:0x%" PRIx64 ", update topic info completed", tmq->consumerId); + tscDebug("consumer:0x%" PRIx64 " update topic info completed", tmq->consumerId); return set; } @@ -1548,7 +1555,7 @@ int32_t tmqAskEp(tmq_t* tmq, bool async) { sendInfo->msgType = TDMT_MND_TMQ_ASK_EP; SEpSet epSet = getEpSet_s(&tmq->pTscObj->pAppInfo->mgmtEp); - tscDebug("consumer:0x%" PRIx64 " ask ep from mnode, async:%d", tmq->consumerId, async); + tscDebug("consumer:0x%" PRIx64 " ask ep from mnode, async:%d, reqId:0x%"PRIx64, tmq->consumerId, async, tmq->consumerId); int64_t transporterId = 0; asyncSendMsgToServer(tmq->pTscObj->pAppInfo->pTransporter, &epSet, &transporterId, sendInfo); @@ -1759,6 +1766,7 @@ void* tmqHandleAllRsp(tmq_t* tmq, int64_t timeout, bool pollIfReset) { while (1) { SMqRspWrapper* rspWrapper = NULL; taosGetQitem(tmq->qall, (void**)&rspWrapper); + if (rspWrapper == NULL) { taosReadAllQitems(tmq->mqueue, tmq->qall); taosGetQitem(tmq->qall, (void**)&rspWrapper); @@ -1881,7 +1889,7 @@ TAOS_RES* tmq_consumer_poll(tmq_t* tmq, int64_t timeout) { // in no topic status, delayed task also need to be processed if (atomic_load_8(&tmq->status) == TMQ_CONSUMER_STATUS__INIT) { - tscDebug("consumer:0x%" PRIx64 ", poll return since consumer status is init", tmq->consumerId); + tscDebug("consumer:0x%" PRIx64 ", poll return since consumer is init", tmq->consumerId); return NULL; } From f570ac22cffb28b97b312bba4dce1f7a514b7c63 Mon Sep 17 00:00:00 2001 From: yihaoDeng Date: Mon, 20 Feb 2023 18:48:12 +0800 Subject: [PATCH 04/31] fix: fix invalid read-write --- source/libs/transport/src/transCli.c | 20 ++++++++++++++------ 1 file changed, 14 insertions(+), 6 deletions(-) diff --git a/source/libs/transport/src/transCli.c b/source/libs/transport/src/transCli.c index 7e1aeafaad..e470d989a7 100644 --- a/source/libs/transport/src/transCli.c +++ b/source/libs/transport/src/transCli.c @@ -727,7 +727,7 @@ static SCliConn* cliCreateConn(SCliThrd* pThrd) { QUEUE_INIT(&conn->q); conn->hostThrd = pThrd; conn->status = ConnNormal; - conn->broken = 0; + conn->broken = false; transRefCliHandle(conn); atomic_add_fetch_32(&pThrd->connCount, 1); @@ -997,6 +997,11 @@ static void cliDestroyBatch(SCliBatch* pBatch) { taosMemoryFree(pBatch); } static void cliHandleBatchReq(SCliBatch* pBatch, SCliThrd* pThrd) { + if (pThrd->quit == true) { + cliDestroyBatch(pBatch); + return; + } + if (pBatch == NULL || pBatch->wLen == 0 || QUEUE_IS_EMPTY(&pBatch->wq)) { return; } @@ -1087,12 +1092,15 @@ static void cliSendBatchCb(uv_write_t* req, int status) { } else { tDebug("%s conn %p succ to send batch msg, batch size:%d, msgLen:%d", CONN_GET_INST_LABEL(conn), conn, p->wLen, p->batchSize); - - if (nxtBatch != NULL) { - conn->pBatch = nxtBatch; - cliSendBatch(conn); + if (conn->broken != true) { + if (nxtBatch != NULL) { + conn->pBatch = nxtBatch; + cliSendBatch(conn); + } else { + addConnToPool(thrd->pool, conn); + } } else { - addConnToPool(thrd->pool, conn); + // release by other callback } } From 328cde55d5e8161d049de705305ba509eef4dc78 Mon Sep 17 00:00:00 2001 From: yihaoDeng Date: Mon, 20 Feb 2023 18:57:53 +0800 Subject: [PATCH 05/31] fix: fix invalid read-write --- source/libs/transport/src/transCli.c | 7 +++++-- 1 file changed, 5 insertions(+), 2 deletions(-) diff --git a/source/libs/transport/src/transCli.c b/source/libs/transport/src/transCli.c index e470d989a7..0ede6f9b2d 100644 --- a/source/libs/transport/src/transCli.c +++ b/source/libs/transport/src/transCli.c @@ -1087,7 +1087,9 @@ static void cliSendBatchCb(uv_write_t* req, int status) { if (status != 0) { tDebug("%s conn %p failed to send batch msg, batch size:%d, msgLen:%d, reason:%s", CONN_GET_INST_LABEL(conn), conn, p->wLen, p->batchSize, uv_err_name(status)); - cliHandleExcept(conn); + + if (!uv_is_closing((uv_handle_t*)&conn->stream)) cliHandleExcept(conn); + cliHandleBatchReq(nxtBatch, thrd); } else { tDebug("%s conn %p succ to send batch msg, batch size:%d, msgLen:%d", CONN_GET_INST_LABEL(conn), conn, p->wLen, @@ -1100,7 +1102,8 @@ static void cliSendBatchCb(uv_write_t* req, int status) { addConnToPool(thrd->pool, conn); } } else { - // release by other callback + cliDestroyBatch(nxtBatch); + // conn release by other callback } } From 2bd38f453cd78c3117f16248066c4d1496933dca Mon Sep 17 00:00:00 2001 From: yihaoDeng Date: Mon, 20 Feb 2023 19:13:29 +0800 Subject: [PATCH 06/31] fix: fix invalid read-write --- source/libs/transport/src/transCli.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/source/libs/transport/src/transCli.c b/source/libs/transport/src/transCli.c index 0ede6f9b2d..65e16eb618 100644 --- a/source/libs/transport/src/transCli.c +++ b/source/libs/transport/src/transCli.c @@ -1094,7 +1094,7 @@ static void cliSendBatchCb(uv_write_t* req, int status) { } else { tDebug("%s conn %p succ to send batch msg, batch size:%d, msgLen:%d", CONN_GET_INST_LABEL(conn), conn, p->wLen, p->batchSize); - if (conn->broken != true) { + if (!uv_is_closing((uv_handle_t*)&conn->stream)) { if (nxtBatch != NULL) { conn->pBatch = nxtBatch; cliSendBatch(conn); From 409e9a9ac8f6d350be10705f57582407aef7698a Mon Sep 17 00:00:00 2001 From: yihaoDeng Date: Mon, 20 Feb 2023 20:45:16 +0800 Subject: [PATCH 07/31] fix: fix invalid read-write --- source/libs/transport/src/transCli.c | 11 +++++++++++ 1 file changed, 11 insertions(+) diff --git a/source/libs/transport/src/transCli.c b/source/libs/transport/src/transCli.c index 65e16eb618..9c3ca20387 100644 --- a/source/libs/transport/src/transCli.c +++ b/source/libs/transport/src/transCli.c @@ -1465,6 +1465,11 @@ static void cliNoBatchDealReq(queue* wq, SCliThrd* pThrd) { QUEUE_REMOVE(h); SCliMsg* pMsg = QUEUE_DATA(h, SCliMsg, q); + + if (pMsg->type == Quit) { + pThrd->stopMsg = pMsg; + continue; + } (*cliAsyncHandle[pMsg->type])(pMsg, pThrd); count++; @@ -1496,6 +1501,12 @@ static void cliBatchDealReq(queue* wq, SCliThrd* pThrd) { QUEUE_REMOVE(h); SCliMsg* pMsg = QUEUE_DATA(h, SCliMsg, q); + + if (pMsg->type == Quit) { + pThrd->stopMsg = pMsg; + continue; + } + if (pMsg->type == Normal && REQUEST_NO_RESP(&pMsg->msg)) { STransConnCtx* pCtx = pMsg->ctx; From 8fa55d38446e3d7dd6bc4dd35c7884fad7450495 Mon Sep 17 00:00:00 2001 From: yihaoDeng Date: Mon, 20 Feb 2023 21:40:24 +0800 Subject: [PATCH 08/31] fix: fix invalid read-write --- source/libs/transport/src/transCli.c | 1 - 1 file changed, 1 deletion(-) diff --git a/source/libs/transport/src/transCli.c b/source/libs/transport/src/transCli.c index 9c3ca20387..b3b8e8b230 100644 --- a/source/libs/transport/src/transCli.c +++ b/source/libs/transport/src/transCli.c @@ -1604,7 +1604,6 @@ static void cliAsyncCb(uv_async_t* handle) { SCliThrd* pThrd = item->pThrd; STrans* pTransInst = pThrd->pTransInst; - SCliMsg* pMsg = NULL; // batch process to avoid to lock/unlock frequently queue wq; taosThreadMutexLock(&item->mtx); From 487ca893b1cdd6dbaf3064873a3d001f4be04c77 Mon Sep 17 00:00:00 2001 From: yihaoDeng Date: Mon, 20 Feb 2023 22:41:49 +0800 Subject: [PATCH 09/31] fix: fix invalid read-write --- source/libs/transport/src/transCli.c | 14 -------------- 1 file changed, 14 deletions(-) diff --git a/source/libs/transport/src/transCli.c b/source/libs/transport/src/transCli.c index b3b8e8b230..9e74825a1f 100644 --- a/source/libs/transport/src/transCli.c +++ b/source/libs/transport/src/transCli.c @@ -2306,22 +2306,8 @@ int transSendRequest(void* shandle, const SEpSet* pEpSet, STransMsg* pReq, STran transReleaseExHandle(transGetInstMgt(), (int64_t)shandle); return TSDB_CODE_RPC_BROKEN_LINK; } - /*if (pTransInst->connLimitNum > 0 && REQUEST_NO_RESP(pReq)) { - char key[TSDB_FQDN_LEN + 64] = {0}; - char* ip = EPSET_GET_INUSE_IP((SEpSet*)pEpSet); - uint16_t port = EPSET_GET_INUSE_PORT((SEpSet*)pEpSet); - CONN_CONSTRUCT_HASH_KEY(key, ip, port); - - int32_t* val = taosHashGet(pThrd->connLimitCache, key, strlen(key)); - if (val != NULL && *val >= pTransInst->connLimitNum) { - transFreeMsg(pReq->pCont); - transReleaseExHandle(transGetInstMgt(), (int64_t)shandle); - return TSDB_CODE_RPC_MAX_SESSIONS; - } - }*/ TRACE_SET_MSGID(&pReq->info.traceId, tGenIdPI64()); - STransConnCtx* pCtx = taosMemoryCalloc(1, sizeof(STransConnCtx)); pCtx->epSet = *pEpSet; pCtx->ahandle = pReq->info.ahandle; From ab3fb17a74a3a41f6b503f80477f93507fbebe6f Mon Sep 17 00:00:00 2001 From: Haojun Liao Date: Mon, 20 Feb 2023 23:43:30 +0800 Subject: [PATCH 10/31] fix(query): set correct shift bits. --- source/util/src/tcompression.c | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-) diff --git a/source/util/src/tcompression.c b/source/util/src/tcompression.c index 695a83abb1..6d3260e0cd 100644 --- a/source/util/src/tcompression.c +++ b/source/util/src/tcompression.c @@ -315,8 +315,8 @@ int32_t tsDecompressINTImp(const char *const input, const int32_t nelements, cha __m256i inc = _mm256_set1_epi64x(bit << 2); for (int32_t i = 0; i < batch; ++i) { - __m256i after = _mm256_srlv_epi64(base, shiftBits); - __m256i zigzagVal = _mm256_and_si256(after, maskVal); + base = _mm256_srlv_epi64(base, shiftBits); + __m256i zigzagVal = _mm256_and_si256(base, maskVal); // ZIGZAG_DECODE(T, v) (((v) >> 1) ^ -((T)((v)&1))) __m256i signmask = _mm256_and_si256(_mm256_set1_epi64x(1), zigzagVal); @@ -327,8 +327,8 @@ int32_t tsDecompressINTImp(const char *const input, const int32_t nelements, cha // calculate the cumulative sum (prefix sum) for each number // decode[0] = prev_value + final[0] // decode[1] = decode[0] + final[1] -----> prev_value + final[0] + final[1] - // decode[2] = decode[1] + final[1] -----> prev_value + final[0] + final[1] + final[2] - // decode[3] = decode[2] + final[1] -----> prev_value + final[0] + final[1] + final[2] + final[3] + // decode[2] = decode[1] + final[2] -----> prev_value + final[0] + final[1] + final[2] + // decode[3] = decode[2] + final[3] -----> prev_value + final[0] + final[1] + final[2] + final[3] // 1, 2, 3, 4 //+ 0, 1, 2, 3 From da2787ee56092e0b46a10ed2eaf427de17f3ef3f Mon Sep 17 00:00:00 2001 From: Haojun Liao Date: Tue, 21 Feb 2023 00:50:36 +0800 Subject: [PATCH 11/31] fix(query): set correct shift bits. --- source/util/src/tcompression.c | 19 +++++++++++-------- 1 file changed, 11 insertions(+), 8 deletions(-) diff --git a/source/util/src/tcompression.c b/source/util/src/tcompression.c index 6d3260e0cd..df3a0f7d6b 100644 --- a/source/util/src/tcompression.c +++ b/source/util/src/tcompression.c @@ -315,8 +315,8 @@ int32_t tsDecompressINTImp(const char *const input, const int32_t nelements, cha __m256i inc = _mm256_set1_epi64x(bit << 2); for (int32_t i = 0; i < batch; ++i) { - base = _mm256_srlv_epi64(base, shiftBits); - __m256i zigzagVal = _mm256_and_si256(base, maskVal); + __m256i after = _mm256_srlv_epi64(base, shiftBits); + __m256i zigzagVal = _mm256_and_si256(after, maskVal); // ZIGZAG_DECODE(T, v) (((v) >> 1) ^ -((T)((v)&1))) __m256i signmask = _mm256_and_si256(_mm256_set1_epi64x(1), zigzagVal); @@ -331,19 +331,22 @@ int32_t tsDecompressINTImp(const char *const input, const int32_t nelements, cha // decode[3] = decode[2] + final[3] -----> prev_value + final[0] + final[1] + final[2] + final[3] // 1, 2, 3, 4 - //+ 0, 1, 2, 3 - // 1, 3, 5, 7 + //+ 0, 1, 0, 3 + // 1, 3, 3, 7 // shift and add for the first round __m128i prev = _mm_set1_epi64x(prev_value); - delta = _mm256_add_epi64(delta, _mm256_slli_si256(delta, 8)); + __m256i x = _mm256_slli_si256(delta, 8); + + delta = _mm256_add_epi64(delta, x); _mm256_storeu_si256((__m256i *)&p[_pos], delta); - // 1, 3, 5, 7 - //+ 0, 0, 1, 3 + // 1, 3, 3, 7 + //+ 0, 0, 3, 3 // 1, 3, 6, 10 // shift and add operation for the second round __m128i firstPart = _mm_loadu_si128((__m128i *)&p[_pos]); - __m128i secPart = _mm_add_epi64(_mm_loadu_si128((__m128i *)&p[_pos + 2]), firstPart); + __m128i secondItem = _mm_set1_epi64x(p[_pos + 1]); + __m128i secPart = _mm_add_epi64(_mm_loadu_si128((__m128i *)&p[_pos + 2]), secondItem); firstPart = _mm_add_epi64(firstPart, prev); secPart = _mm_add_epi64(secPart, prev); From ec7549608c2feb32fdd8d177b8ef640ee2e9bd9b Mon Sep 17 00:00:00 2001 From: Haojun Liao Date: Tue, 21 Feb 2023 10:23:02 +0800 Subject: [PATCH 12/31] fix(query): fix error in simd. --- source/util/src/tcompression.c | 8 +++++++- 1 file changed, 7 insertions(+), 1 deletion(-) diff --git a/source/util/src/tcompression.c b/source/util/src/tcompression.c index df3a0f7d6b..bef22ef95a 100644 --- a/source/util/src/tcompression.c +++ b/source/util/src/tcompression.c @@ -356,15 +356,19 @@ int32_t tsDecompressINTImp(const char *const input, const int32_t nelements, cha shiftBits = _mm256_add_epi64(shiftBits, inc); prev_value = p[_pos + 3]; + +// uDebug("_pos:%d %"PRId64", %"PRId64", %"PRId64", %"PRId64, _pos, p[_pos], p[_pos+1], p[_pos+2], p[_pos+3]); _pos += 4; } // handle the remain value for (int32_t i = 0; i < remain; i++) { - zigzag_value = ((w >> (v + (batch * bit))) & mask); + zigzag_value = ((w >> (v + (batch * bit * 4))) & mask); prev_value += ZIGZAG_DECODE(int64_t, zigzag_value); p[_pos++] = prev_value; +// uDebug("_pos:%d %"PRId64, _pos-1, p[_pos-1]); + v += bit; } } else { @@ -373,6 +377,8 @@ int32_t tsDecompressINTImp(const char *const input, const int32_t nelements, cha prev_value += ZIGZAG_DECODE(int64_t, zigzag_value); p[_pos++] = prev_value; +// uDebug("_pos:%d %"PRId64, _pos-1, p[_pos-1]); + v += bit; } } From 466d5fe7161264d3edb5a3c8fc69339c9f682683 Mon Sep 17 00:00:00 2001 From: Haojun Liao Date: Tue, 21 Feb 2023 10:31:30 +0800 Subject: [PATCH 13/31] fix(query): fix error in simd. --- source/util/src/tcompression.c | 1 - 1 file changed, 1 deletion(-) diff --git a/source/util/src/tcompression.c b/source/util/src/tcompression.c index bef22ef95a..7631da1b56 100644 --- a/source/util/src/tcompression.c +++ b/source/util/src/tcompression.c @@ -356,7 +356,6 @@ int32_t tsDecompressINTImp(const char *const input, const int32_t nelements, cha shiftBits = _mm256_add_epi64(shiftBits, inc); prev_value = p[_pos + 3]; - // uDebug("_pos:%d %"PRId64", %"PRId64", %"PRId64", %"PRId64, _pos, p[_pos], p[_pos+1], p[_pos+2], p[_pos+3]); _pos += 4; } From 2d85d0191c1f619d131a32262417fa42202fdcd6 Mon Sep 17 00:00:00 2001 From: yihaoDeng Date: Tue, 21 Feb 2023 14:40:10 +0800 Subject: [PATCH 14/31] fix: add vgroup count --- include/common/tcommon.h | 19 +++--- source/libs/executor/src/sysscanoperator.c | 72 +++++++++++----------- source/libs/function/src/builtinsimpl.c | 51 ++++++++------- 3 files changed, 75 insertions(+), 67 deletions(-) diff --git a/include/common/tcommon.h b/include/common/tcommon.h index d88747eb92..72d969a0d5 100644 --- a/include/common/tcommon.h +++ b/include/common/tcommon.h @@ -195,7 +195,7 @@ typedef struct SDataBlockInfo { uint32_t capacity; SBlockID id; int16_t hasVarCol; - int16_t dataLoad; // denote if the data is loaded or not + int16_t dataLoad; // denote if the data is loaded or not // TODO: optimize and remove following int64_t version; // used for stream, and need serialization @@ -204,8 +204,8 @@ typedef struct SDataBlockInfo { STimeWindow calWin; // used for stream, do not serialize TSKEY watermark; // used for stream - char parTbName[TSDB_TABLE_NAME_LEN]; // used for stream partition - STag* pTag; // used for stream partition + char parTbName[TSDB_TABLE_NAME_LEN]; // used for stream partition + STag* pTag; // used for stream partition } SDataBlockInfo; typedef struct SSDataBlock { @@ -239,22 +239,22 @@ typedef struct SVarColAttr { // pBlockAgg->numOfNull == info.rows, all data are null // pBlockAgg->numOfNull == 0, no data are null. typedef struct SColumnInfoData { - char* pData; // the corresponding block data in memory + char* pData; // the corresponding block data in memory union { char* nullbitmap; // bitmap, one bit for each item in the list SVarColAttr varmeta; }; - SColumnInfo info; // column info - bool hasNull; // if current column data has null value. + SColumnInfo info; // column info + bool hasNull; // if current column data has null value. } SColumnInfoData; typedef struct SQueryTableDataCond { uint64_t suid; - int32_t order; // desc|asc order to iterate the data block + int32_t order; // desc|asc order to iterate the data block int32_t numOfCols; SColumnInfo* colList; - int32_t* pSlotList; // the column output destation slot, and it may be null - int32_t type; // data block load type: + int32_t* pSlotList; // the column output destation slot, and it may be null + int32_t type; // data block load type: STimeWindow twindows; int64_t startVersion; int64_t endVersion; @@ -300,6 +300,7 @@ typedef struct STableBlockDistInfo { int32_t firstSeekTimeUs; uint32_t numOfInmemRows; uint32_t numOfSmallBlocks; + uint32_t numOfVgroups; int32_t blockRowsHisto[20]; } STableBlockDistInfo; diff --git a/source/libs/executor/src/sysscanoperator.c b/source/libs/executor/src/sysscanoperator.c index 88d9fb1a1f..6516c2c920 100644 --- a/source/libs/executor/src/sysscanoperator.c +++ b/source/libs/executor/src/sysscanoperator.c @@ -81,10 +81,10 @@ typedef struct MergeIndex { } MergeIndex; typedef struct SBlockDistInfo { - SSDataBlock* pResBlock; - STsdbReader* pHandle; - SReadHandle readHandle; - uint64_t uid; // table uid + SSDataBlock* pResBlock; + STsdbReader* pHandle; + SReadHandle readHandle; + uint64_t uid; // table uid } SBlockDistInfo; static int32_t sysChkFilter__Comm(SNode* pNode); @@ -129,10 +129,10 @@ static char* SYSTABLE_IDX_COLUMN[] = {"table_name", "db_name", "create_time" static char* SYSTABLE_SPECIAL_COL[] = {"db_name", "vgroup_id"}; -static int32_t buildSysDbTableInfo(const SSysTableScanInfo* pInfo, int32_t capacity); -static SSDataBlock* buildInfoSchemaTableMetaBlock(char* tableName); -static void destroySysScanOperator(void* param); -static int32_t loadSysTableCallback(void* param, SDataBuf* pMsg, int32_t code); +static int32_t buildSysDbTableInfo(const SSysTableScanInfo* pInfo, int32_t capacity); +static SSDataBlock* buildInfoSchemaTableMetaBlock(char* tableName); +static void destroySysScanOperator(void* param); +static int32_t loadSysTableCallback(void* param, SDataBuf* pMsg, int32_t code); static __optSysFilter optSysGetFilterFunc(int32_t ctype, bool* reverse); static int32_t sysTableUserTagsFillOneTableTags(const SSysTableScanInfo* pInfo, SMetaReader* smrSuperTable, @@ -199,11 +199,11 @@ int32_t sysFilte__TableName(void* arg, SNode* pNode, SArray* result) { if (func == NULL) return -1; SMetaFltParam param = {.suid = 0, - .cid = 0, - .type = TSDB_DATA_TYPE_VARCHAR, - .val = pVal->datum.p, - .reverse = reverse, - .filterFunc = func}; + .cid = 0, + .type = TSDB_DATA_TYPE_VARCHAR, + .val = pVal->datum.p, + .reverse = reverse, + .filterFunc = func}; return -1; } @@ -218,11 +218,11 @@ int32_t sysFilte__CreateTime(void* arg, SNode* pNode, SArray* result) { if (func == NULL) return -1; SMetaFltParam param = {.suid = 0, - .cid = 0, - .type = TSDB_DATA_TYPE_BIGINT, - .val = &pVal->datum.i, - .reverse = reverse, - .filterFunc = func}; + .cid = 0, + .type = TSDB_DATA_TYPE_BIGINT, + .val = &pVal->datum.i, + .reverse = reverse, + .filterFunc = func}; int32_t ret = metaFilterCreateTime(pMeta, ¶m, result); return ret; @@ -350,9 +350,9 @@ static int32_t optSysMergeRslt(SArray* mRslt, SArray* rslt); static SSDataBlock* sysTableScanFromMNode(SOperatorInfo* pOperator, SSysTableScanInfo* pInfo, const char* name, SExecTaskInfo* pTaskInfo); void extractTbnameSlotId(SSysTableScanInfo* pInfo, const SScanPhysiNode* pScanNode); -static void sysTableScanFillTbName(SOperatorInfo* pOperator, const SSysTableScanInfo* pInfo, - const char* name, SSDataBlock* pBlock); -__optSysFilter optSysGetFilterFunc(int32_t ctype, bool* reverse) { +static void sysTableScanFillTbName(SOperatorInfo* pOperator, const SSysTableScanInfo* pInfo, const char* name, + SSDataBlock* pBlock); +__optSysFilter optSysGetFilterFunc(int32_t ctype, bool* reverse) { if (ctype == OP_TYPE_LOWER_EQUAL || ctype == OP_TYPE_LOWER_THAN) { *reverse = true; } @@ -516,9 +516,10 @@ static SSDataBlock* sysTableScanUserTags(SOperatorInfo* pOperator) { metaTbCursorPrev(pInfo->pCur); blockFull = true; } else { - sysTableUserTagsFillOneTableTags(pInfo, &smrSuperTable, &pInfo->pCur->mr, dbname, tableName, &numOfRows, dataBlock); + sysTableUserTagsFillOneTableTags(pInfo, &smrSuperTable, &pInfo->pCur->mr, dbname, tableName, &numOfRows, + dataBlock); } - + metaReaderClear(&smrSuperTable); if (blockFull || numOfRows >= pOperator->resultInfo.capacity) { @@ -1343,7 +1344,7 @@ static SSDataBlock* doSysTableScan(SOperatorInfo* pOperator) { setOperatorCompleted(pOperator); } - return pBlock->info.rows > 0? pBlock:NULL; + return pBlock->info.rows > 0 ? pBlock : NULL; } else { return NULL; } @@ -1357,7 +1358,7 @@ static void sysTableScanFillTbName(SOperatorInfo* pOperator, const SSysTableScan if (pInfo->tbnameSlotId != -1) { SColumnInfoData* pColumnInfoData = (SColumnInfoData*)taosArrayGet(pBlock->pDataBlock, pInfo->tbnameSlotId); - char varTbName[TSDB_TABLE_FNAME_LEN - 1 + VARSTR_HEADER_SIZE] = {0}; + char varTbName[TSDB_TABLE_FNAME_LEN - 1 + VARSTR_HEADER_SIZE] = {0}; memcpy(varDataVal(varTbName), name, strlen(name)); varDataSetLen(varTbName, strlen(name)); @@ -1489,10 +1490,11 @@ SOperatorInfo* createSysTableScanOperatorInfo(void* readHandle, SSystemTableScan setOperatorInfo(pOperator, "SysTableScanOperator", QUERY_NODE_PHYSICAL_PLAN_SYSTABLE_SCAN, false, OP_NOT_OPENED, pInfo, pTaskInfo); pOperator->exprSupp.numOfExprs = taosArrayGetSize(pInfo->pRes->pDataBlock); - pOperator->fpSet = createOperatorFpSet(optrDummyOpenFn, doSysTableScan, NULL, destroySysScanOperator, optrDefaultBufFn, NULL); + pOperator->fpSet = + createOperatorFpSet(optrDummyOpenFn, doSysTableScan, NULL, destroySysScanOperator, optrDefaultBufFn, NULL); return pOperator; - _error: +_error: if (pInfo != NULL) { destroySysScanOperator(pInfo); } @@ -1927,7 +1929,7 @@ static SSDataBlock* doBlockInfoScan(SOperatorInfo* pOperator) { // make the valgrind happy that all memory buffer has been initialized already. if (slotId != 0) { SColumnInfoData* p1 = taosArrayGet(pBlock->pDataBlock, 0); - int64_t v = 0; + int64_t v = 0; colDataAppendInt64(p1, 0, &v); } @@ -1937,10 +1939,10 @@ static SSDataBlock* doBlockInfoScan(SOperatorInfo* pOperator) { } static void destroyBlockDistScanOperatorInfo(void* param) { - SBlockDistInfo* pDistInfo = (SBlockDistInfo*)param; - blockDataDestroy(pDistInfo->pResBlock); - tsdbReaderClose(pDistInfo->pHandle); - taosMemoryFreeClear(param); + SBlockDistInfo* pDistInfo = (SBlockDistInfo*)param; + blockDataDestroy(pDistInfo->pResBlock); + tsdbReaderClose(pDistInfo->pHandle); + taosMemoryFreeClear(param); } static int32_t initTableblockDistQueryCond(uint64_t uid, SQueryTableDataCond* pCond) { @@ -2001,7 +2003,7 @@ SOperatorInfo* createDataBlockInfoScanOperator(SReadHandle* readHandle, SBlockDi } pInfo->readHandle = *readHandle; - pInfo->uid = (pBlockScanNode->suid != 0)? pBlockScanNode->suid:pBlockScanNode->uid; + pInfo->uid = (pBlockScanNode->suid != 0) ? pBlockScanNode->suid : pBlockScanNode->uid; int32_t numOfCols = 0; SExprInfo* pExprInfo = createExprInfo(pBlockScanNode->pScanPseudoCols, NULL, &numOfCols); @@ -2012,8 +2014,8 @@ SOperatorInfo* createDataBlockInfoScanOperator(SReadHandle* readHandle, SBlockDi setOperatorInfo(pOperator, "DataBlockDistScanOperator", QUERY_NODE_PHYSICAL_PLAN_BLOCK_DIST_SCAN, false, OP_NOT_OPENED, pInfo, pTaskInfo); - pOperator->fpSet = - createOperatorFpSet(optrDummyOpenFn, doBlockInfoScan, NULL, destroyBlockDistScanOperatorInfo, optrDefaultBufFn, NULL); + pOperator->fpSet = createOperatorFpSet(optrDummyOpenFn, doBlockInfoScan, NULL, destroyBlockDistScanOperatorInfo, + optrDefaultBufFn, NULL); return pOperator; _error: diff --git a/source/libs/function/src/builtinsimpl.c b/source/libs/function/src/builtinsimpl.c index 50ceda4605..e42812db56 100644 --- a/source/libs/function/src/builtinsimpl.c +++ b/source/libs/function/src/builtinsimpl.c @@ -775,13 +775,13 @@ int32_t maxFunction(SqlFunctionCtx* pCtx) { static int32_t setNullSelectivityValue(SqlFunctionCtx* pCtx, SSDataBlock* pBlock, int32_t rowIndex); static int32_t setSelectivityValue(SqlFunctionCtx* pCtx, SSDataBlock* pBlock, const STuplePos* pTuplePos, - int32_t rowIndex); + int32_t rowIndex); int32_t minmaxFunctionFinalize(SqlFunctionCtx* pCtx, SSDataBlock* pBlock) { int32_t code = TSDB_CODE_SUCCESS; SResultRowEntryInfo* pEntryInfo = GET_RES_INFO(pCtx); - SMinmaxResInfo* pRes = GET_ROWCELL_INTERBUF(pEntryInfo); + SMinmaxResInfo* pRes = GET_ROWCELL_INTERBUF(pEntryInfo); int32_t slotId = pCtx->pExpr->base.resSchema.slotId; int32_t currentRow = pBlock->info.rows; @@ -795,7 +795,7 @@ int32_t minmaxFunctionFinalize(SqlFunctionCtx* pCtx, SSDataBlock* pBlock) { case TSDB_DATA_TYPE_UBIGINT: case TSDB_DATA_TYPE_BIGINT: ((int64_t*)pCol->pData)[currentRow] = pRes->v; -// colDataAppendInt64(pCol, currentRow, &pRes->v); + // colDataAppendInt64(pCol, currentRow, &pRes->v); break; case TSDB_DATA_TYPE_UINT: case TSDB_DATA_TYPE_INT: @@ -1691,7 +1691,7 @@ int32_t percentileFunction(SqlFunctionCtx* pCtx) { int32_t percentileFinalize(SqlFunctionCtx* pCtx, SSDataBlock* pBlock) { SVariant* pVal = &pCtx->param[1].param; - int32_t code = 0; + int32_t code = 0; double v = 0; GET_TYPED_DATA(v, double, pVal->nType, &pVal->i); @@ -2070,7 +2070,7 @@ static void prepareBuf(SqlFunctionCtx* pCtx) { } static int32_t firstlastSaveTupleData(const SSDataBlock* pSrcBlock, int32_t rowIndex, SqlFunctionCtx* pCtx, - SFirstLastRes* pInfo) { + SFirstLastRes* pInfo) { int32_t code = TSDB_CODE_SUCCESS; if (pCtx->subsidiaries.num <= 0) { @@ -2123,7 +2123,8 @@ int32_t firstFunction(SqlFunctionCtx* pCtx) { } // All null data column, return directly. - if (pInput->colDataSMAIsSet && (pInput->pColumnDataAgg[0]->numOfNull == pInput->totalRows) && pInputCol->hasNull == true) { + if (pInput->colDataSMAIsSet && (pInput->pColumnDataAgg[0]->numOfNull == pInput->totalRows) && + pInputCol->hasNull == true) { // save selectivity value for column consisted of all null values int32_t code = firstlastSaveTupleData(pCtx->pSrcBlock, pInput->startRowIndex, pCtx, pInfo); if (code != TSDB_CODE_SUCCESS) { @@ -2239,7 +2240,8 @@ int32_t lastFunction(SqlFunctionCtx* pCtx) { } // All null data column, return directly. - if (pInput->colDataSMAIsSet && (pInput->pColumnDataAgg[0]->numOfNull == pInput->totalRows) && pInputCol->hasNull == true) { + if (pInput->colDataSMAIsSet && (pInput->pColumnDataAgg[0]->numOfNull == pInput->totalRows) && + pInputCol->hasNull == true) { // save selectivity value for column consisted of all null values int32_t code = firstlastSaveTupleData(pCtx->pSrcBlock, pInput->startRowIndex, pCtx, pInfo); if (code != TSDB_CODE_SUCCESS) { @@ -2333,7 +2335,7 @@ int32_t lastFunction(SqlFunctionCtx* pCtx) { } if (pResInfo->numOfRes == 0 || pInfo->ts < cts) { - char* data = colDataGetData(pInputCol, chosen); + char* data = colDataGetData(pInputCol, chosen); int32_t code = doSaveCurrentVal(pCtx, i, cts, type, data); if (code != TSDB_CODE_SUCCESS) { return code; @@ -2344,7 +2346,7 @@ int32_t lastFunction(SqlFunctionCtx* pCtx) { for (int32_t i = pInput->startRowIndex + round * 4; i < pInput->startRowIndex + pInput->numOfRows; ++i) { if (pResInfo->numOfRes == 0 || pInfo->ts < pts[i]) { - char* data = colDataGetData(pInputCol, i); + char* data = colDataGetData(pInputCol, i); int32_t code = doSaveCurrentVal(pCtx, i, pts[i], type, data); if (code != TSDB_CODE_SUCCESS) { return code; @@ -2361,7 +2363,7 @@ int32_t lastFunction(SqlFunctionCtx* pCtx) { numOfElems++; if (pResInfo->numOfRes == 0 || pInfo->ts < pts[i]) { - char* data = colDataGetData(pInputCol, i); + char* data = colDataGetData(pInputCol, i); int32_t code = doSaveCurrentVal(pCtx, i, pts[i], type, data); if (code != TSDB_CODE_SUCCESS) { return code; @@ -2408,7 +2410,7 @@ static int32_t firstLastTransferInfoImpl(SFirstLastRes* pInput, SFirstLastRes* p } static int32_t firstLastTransferInfo(SqlFunctionCtx* pCtx, SFirstLastRes* pInput, SFirstLastRes* pOutput, bool isFirst, - int32_t rowIndex) { + int32_t rowIndex) { if (TSDB_CODE_SUCCESS == firstLastTransferInfoImpl(pInput, pOutput, isFirst)) { int32_t code = firstlastSaveTupleData(pCtx->pSrcBlock, rowIndex, pCtx, pOutput); if (code != TSDB_CODE_SUCCESS) { @@ -2435,7 +2437,7 @@ static int32_t firstLastFunctionMergeImpl(SqlFunctionCtx* pCtx, bool isFirstQuer for (int32_t i = start; i < start + pInput->numOfRows; ++i) { char* data = colDataGetData(pCol, i); SFirstLastRes* pInputInfo = (SFirstLastRes*)varDataVal(data); - int32_t code = firstLastTransferInfo(pCtx, pInputInfo, pInfo, isFirstQuery, i); + int32_t code = firstLastTransferInfo(pCtx, pInputInfo, pInfo, isFirstQuery, i); if (code != TSDB_CODE_SUCCESS) { return code; } @@ -2667,7 +2669,7 @@ static int32_t doSetPrevVal(SDiffInfo* pDiffInfo, int32_t type, const char* pv, } static int32_t doHandleDiff(SDiffInfo* pDiffInfo, int32_t type, const char* pv, SColumnInfoData* pOutput, int32_t pos, - int32_t order, int64_t ts) { + int32_t order, int64_t ts) { int32_t factor = (order == TSDB_ORDER_ASC) ? 1 : -1; pDiffInfo->prevTs = ts; switch (type) { @@ -2875,8 +2877,8 @@ static STopBotRes* getTopBotOutputInfo(SqlFunctionCtx* pCtx) { return pRes; } -static int32_t doAddIntoResult(SqlFunctionCtx* pCtx, void* pData, int32_t rowIndex, SSDataBlock* pSrcBlock, uint16_t type, - uint64_t uid, SResultRowEntryInfo* pEntryInfo, bool isTopQuery); +static int32_t doAddIntoResult(SqlFunctionCtx* pCtx, void* pData, int32_t rowIndex, SSDataBlock* pSrcBlock, + uint16_t type, uint64_t uid, SResultRowEntryInfo* pEntryInfo, bool isTopQuery); static void addResult(SqlFunctionCtx* pCtx, STopBotResItem* pSourceItem, int16_t type, bool isTopQuery); @@ -2897,7 +2899,7 @@ int32_t topFunction(SqlFunctionCtx* pCtx) { } numOfElems++; - char* data = colDataGetData(pCol, i); + char* data = colDataGetData(pCol, i); int32_t code = doAddIntoResult(pCtx, data, i, pCtx->pSrcBlock, pRes->type, pInput->uid, pResInfo, true); if (code != TSDB_CODE_SUCCESS) { return code; @@ -2931,7 +2933,7 @@ int32_t bottomFunction(SqlFunctionCtx* pCtx) { } numOfElems++; - char* data = colDataGetData(pCol, i); + char* data = colDataGetData(pCol, i); int32_t code = doAddIntoResult(pCtx, data, i, pCtx->pSrcBlock, pRes->type, pInput->uid, pResInfo, false); if (code != TSDB_CODE_SUCCESS) { return code; @@ -2983,7 +2985,7 @@ static int32_t topBotResComparFn(const void* p1, const void* p2, const void* par } int32_t doAddIntoResult(SqlFunctionCtx* pCtx, void* pData, int32_t rowIndex, SSDataBlock* pSrcBlock, uint16_t type, - uint64_t uid, SResultRowEntryInfo* pEntryInfo, bool isTopQuery) { + uint64_t uid, SResultRowEntryInfo* pEntryInfo, bool isTopQuery) { STopBotRes* pRes = getTopBotOutputInfo(pCtx); SVariant val = {0}; @@ -3174,7 +3176,7 @@ static char* doLoadTupleData(SSerializeDataHandle* pHandle, const STuplePos* pPo if (pPage == NULL) { return NULL; } - char* p = pPage->data + pPos->offset; + char* p = pPage->data + pPos->offset; releaseBufPage(pHandle->pBuf, pPage); return p; } else { @@ -4635,7 +4637,7 @@ int32_t sampleFunction(SqlFunctionCtx* pCtx) { continue; } - char* data = colDataGetData(pInputCol, i); + char* data = colDataGetData(pInputCol, i); int32_t code = doReservoirSample(pCtx, pInfo, data, i); if (code != TSDB_CODE_SUCCESS) { return code; @@ -4655,7 +4657,7 @@ int32_t sampleFunction(SqlFunctionCtx* pCtx) { } int32_t sampleFinalize(SqlFunctionCtx* pCtx, SSDataBlock* pBlock) { - int32_t code = TSDB_CODE_SUCCESS; + int32_t code = TSDB_CODE_SUCCESS; SResultRowEntryInfo* pEntryInfo = GET_RES_INFO(pCtx); SSampleInfo* pInfo = getSampleOutputInfo(pCtx); @@ -4989,7 +4991,7 @@ int32_t modeFunction(SqlFunctionCtx* pCtx) { } numOfElems++; - char* data = colDataGetData(pInputCol, i); + char* data = colDataGetData(pInputCol, i); int32_t code = doModeAdd(pInfo, i, pCtx, data); if (code != TSDB_CODE_SUCCESS) { return code; @@ -5410,6 +5412,7 @@ int32_t blockDistFunction(SqlFunctionCtx* pCtx) { if (pDistInfo->maxRows < p1.maxRows) { pDistInfo->maxRows = p1.maxRows; } + pDistInfo->numOfVgroups += (p1.numOfTables != 0 ? 1 : 0); for (int32_t i = 0; i < tListLen(pDistInfo->blockRowsHisto); ++i) { pDistInfo->blockRowsHisto[i] += p1.blockRowsHisto[i]; @@ -5438,6 +5441,7 @@ int32_t tSerializeBlockDistInfo(void* buf, int32_t bufLen, const STableBlockDist if (tEncodeI32(&encoder, pInfo->defMinRows) < 0) return -1; if (tEncodeU32(&encoder, pInfo->numOfInmemRows) < 0) return -1; if (tEncodeU32(&encoder, pInfo->numOfSmallBlocks) < 0) return -1; + if (tEncodeU32(&encoder, pInfo->numOfVgroups) < 0) return -1; for (int32_t i = 0; i < tListLen(pInfo->blockRowsHisto); ++i) { if (tEncodeI32(&encoder, pInfo->blockRowsHisto[i]) < 0) return -1; @@ -5469,6 +5473,7 @@ int32_t tDeserializeBlockDistInfo(void* buf, int32_t bufLen, STableBlockDistInfo if (tDecodeI32(&decoder, &pInfo->defMinRows) < 0) return -1; if (tDecodeU32(&decoder, &pInfo->numOfInmemRows) < 0) return -1; if (tDecodeU32(&decoder, &pInfo->numOfSmallBlocks) < 0) return -1; + if (tDecodeU32(&decoder, &pInfo->numOfVgroups) < 0) return -1; for (int32_t i = 0; i < tListLen(pInfo->blockRowsHisto); ++i) { if (tDecodeI32(&decoder, &pInfo->blockRowsHisto[i]) < 0) return -1; @@ -5520,7 +5525,7 @@ int32_t blockDistFinalize(SqlFunctionCtx* pCtx, SSDataBlock* pBlock) { colDataAppend(pColInfo, row++, st, false); len = sprintf(st + VARSTR_HEADER_SIZE, "Total_Tables=[%d] Total_Files=[%d] Total_Vgroups=[%d]", pData->numOfTables, - pData->numOfFiles, 0); + pData->numOfFiles, pData->numOfVgroups); varDataSetLen(st, len); colDataAppend(pColInfo, row++, st, false); From 9a5b7ce2202d1beddd38d52a91c76d319894ff06 Mon Sep 17 00:00:00 2001 From: Haojun Liao Date: Tue, 21 Feb 2023 15:32:17 +0800 Subject: [PATCH 15/31] refactor: do some internal refactor. --- include/common/tmsg.h | 8 +- source/dnode/mnode/impl/src/mndConsumer.c | 59 +++++++++------ source/dnode/mnode/impl/src/mndScheduler.c | 29 ++++--- source/dnode/mnode/impl/src/mndSubscribe.c | 88 ++++++++++++---------- source/dnode/mnode/impl/src/mndTopic.c | 2 +- 5 files changed, 102 insertions(+), 84 deletions(-) diff --git a/include/common/tmsg.h b/include/common/tmsg.h index 8dcadf47b6..ba13195b97 100644 --- a/include/common/tmsg.h +++ b/include/common/tmsg.h @@ -1830,10 +1830,10 @@ typedef struct { } SMqConsumerLostMsg, SMqConsumerRecoverMsg, SMqConsumerClearMsg; typedef struct { - int64_t consumerId; - char cgroup[TSDB_CGROUP_LEN]; - char clientId[256]; - SArray* topicNames; // SArray + uint64_t consumerId; + char cgroup[TSDB_CGROUP_LEN]; + char clientId[256]; + SArray* topicNames; // SArray } SCMSubscribeReq; static FORCE_INLINE int32_t tSerializeSCMSubscribeReq(void** buf, const SCMSubscribeReq* pReq) { diff --git a/source/dnode/mnode/impl/src/mndConsumer.c b/source/dnode/mnode/impl/src/mndConsumer.c index 1aa2fa997b..40feb22906 100644 --- a/source/dnode/mnode/impl/src/mndConsumer.c +++ b/source/dnode/mnode/impl/src/mndConsumer.c @@ -238,7 +238,9 @@ static int32_t mndProcessMqTimerMsg(SRpcMsg *pMsg) { // iterate all consumers, find all modification while (1) { pIter = sdbFetch(pSdb, SDB_CONSUMER, pIter, (void **)&pConsumer); - if (pIter == NULL) break; + if (pIter == NULL) { + break; + } int32_t hbStatus = atomic_add_fetch_32(&pConsumer->hbStatus, 1); int32_t status = atomic_load_32(&pConsumer->status); @@ -335,7 +337,7 @@ static int32_t mndProcessMqHbReq(SRpcMsg *pMsg) { int64_t consumerId = req.consumerId; SMqConsumerObj *pConsumer = mndAcquireConsumer(pMnode, consumerId); if (pConsumer == NULL) { - mError("consumer %" PRId64 " not exist", consumerId); + mError("consumer:0x%"PRIx64 " not exist", consumerId); terrno = TSDB_CODE_MND_CONSUMER_NOT_EXIST; return -1; } @@ -345,7 +347,7 @@ static int32_t mndProcessMqHbReq(SRpcMsg *pMsg) { int32_t status = atomic_load_32(&pConsumer->status); if (status == MQ_CONSUMER_STATUS__LOST_REBD) { - mInfo("try to recover consumer %" PRId64 "", consumerId); + mInfo("try to recover consumer:0x%"PRIx64 "", consumerId); SMqConsumerRecoverMsg *pRecoverMsg = rpcMallocCont(sizeof(SMqConsumerRecoverMsg)); pRecoverMsg->consumerId = consumerId; @@ -390,7 +392,7 @@ static int32_t mndProcessAskEpReq(SRpcMsg *pMsg) { #if 1 if (status == MQ_CONSUMER_STATUS__LOST_REBD) { - mInfo("try to recover consumer %" PRId64 "", consumerId); + mInfo("try to recover consumer:0x%"PRIx64 "", consumerId); SMqConsumerRecoverMsg *pRecoverMsg = rpcMallocCont(sizeof(SMqConsumerRecoverMsg)); pRecoverMsg->consumerId = consumerId; @@ -404,7 +406,7 @@ static int32_t mndProcessAskEpReq(SRpcMsg *pMsg) { #endif if (status != MQ_CONSUMER_STATUS__READY) { - mInfo("consumer %" PRId64 " not ready, status: %s", consumerId, mndConsumerStatusName(status)); + mInfo("consumer:0x%"PRIx64 " not ready, status: %s", consumerId, mndConsumerStatusName(status)); terrno = TSDB_CODE_MND_CONSUMER_NOT_READY; return -1; } @@ -526,12 +528,14 @@ int32_t mndSetConsumerCommitLogs(SMnode *pMnode, STrans *pTrans, SMqConsumerObj return 0; } -static int32_t mndProcessSubscribeReq(SRpcMsg *pMsg) { - SMnode *pMnode = pMsg->info.node; - char *msgStr = pMsg->pCont; +int32_t mndProcessSubscribeReq(SRpcMsg *pMsg) { + SMnode *pMnode = pMsg->info.node; + char *msgStr = pMsg->pCont; + SCMSubscribeReq subscribe = {0}; tDeserializeSCMSubscribeReq(msgStr, &subscribe); - int64_t consumerId = subscribe.consumerId; + + uint64_t consumerId = subscribe.consumerId; char *cgroup = subscribe.cgroup; SMqConsumerObj *pConsumerOld = NULL; SMqConsumerObj *pConsumerNew = NULL; @@ -542,21 +546,23 @@ static int32_t mndProcessSubscribeReq(SRpcMsg *pMsg) { taosArrayRemoveDuplicateP(newSub, taosArrayCompareString, taosMemoryFree); int32_t newTopicNum = taosArrayGetSize(newSub); + // check topic existance STrans *pTrans = mndTransCreate(pMnode, TRN_POLICY_RETRY, TRN_CONFLICT_NOTHING, pMsg, "subscribe"); - if (pTrans == NULL) goto SUBSCRIBE_OVER; + if (pTrans == NULL) { + goto _over; + } for (int32_t i = 0; i < newTopicNum; i++) { char *topic = taosArrayGetP(newSub, i); SMqTopicObj *pTopic = mndAcquireTopic(pMnode, topic); - if (pTopic == NULL) { - terrno = TSDB_CODE_MND_TOPIC_NOT_EXIST; - goto SUBSCRIBE_OVER; + if (pTopic == NULL) { // terrno has been set by callee function + goto _over; } if (mndCheckTopicPrivilege(pMnode, pMsg->info.conn.user, MND_OPER_SUBSCRIBE, pTopic) != 0) { mndReleaseTopic(pMnode, pTopic); - goto SUBSCRIBE_OVER; + goto _over; } mndReleaseTopic(pMnode, pTopic); @@ -578,8 +584,8 @@ static int32_t mndProcessSubscribeReq(SRpcMsg *pMsg) { taosArrayPush(pConsumerNew->assignedTopics, &newTopicCopy); } - if (mndSetConsumerCommitLogs(pMnode, pTrans, pConsumerNew) != 0) goto SUBSCRIBE_OVER; - if (mndTransPrepare(pMnode, pTrans) != 0) goto SUBSCRIBE_OVER; + if (mndSetConsumerCommitLogs(pMnode, pTrans, pConsumerNew) != 0) goto _over; + if (mndTransPrepare(pMnode, pTrans) != 0) goto _over; } else { /*taosRLockLatch(&pConsumerOld->lock);*/ @@ -591,13 +597,13 @@ static int32_t mndProcessSubscribeReq(SRpcMsg *pMsg) { if (status != MQ_CONSUMER_STATUS__READY) { terrno = TSDB_CODE_MND_CONSUMER_NOT_READY; - goto SUBSCRIBE_OVER; + goto _over; } pConsumerNew = tNewSMqConsumerObj(consumerId, cgroup); if (pConsumerNew == NULL) { terrno = TSDB_CODE_OUT_OF_MEMORY; - goto SUBSCRIBE_OVER; + goto _over; } pConsumerNew->updateType = CONSUMER_UPDATE__MODIFY; @@ -650,16 +656,16 @@ static int32_t mndProcessSubscribeReq(SRpcMsg *pMsg) { /*if (taosArrayGetSize(pConsumerNew->assignedTopics) == 0) {*/ /*pConsumerNew->updateType = */ /*}*/ - goto SUBSCRIBE_OVER; + goto _over; } - if (mndSetConsumerCommitLogs(pMnode, pTrans, pConsumerNew) != 0) goto SUBSCRIBE_OVER; - if (mndTransPrepare(pMnode, pTrans) != 0) goto SUBSCRIBE_OVER; + if (mndSetConsumerCommitLogs(pMnode, pTrans, pConsumerNew) != 0) goto _over; + if (mndTransPrepare(pMnode, pTrans) != 0) goto _over; } code = TSDB_CODE_ACTION_IN_PROGRESS; -SUBSCRIBE_OVER: +_over: mndTransDrop(pTrans); if (pConsumerOld) { @@ -971,16 +977,19 @@ static int32_t mndRetrieveConsumer(SRpcMsg *pReq, SShowObj *pShow, SSDataBlock * while (numOfRows < rowsCapacity) { pShow->pIter = sdbFetch(pSdb, SDB_CONSUMER, pShow->pIter, (void **)&pConsumer); - if (pShow->pIter == NULL) break; + if (pShow->pIter == NULL) { + break; + } + if (taosArrayGetSize(pConsumer->assignedTopics) == 0) { - mDebug("showing consumer %" PRId64 " no assigned topic, skip", pConsumer->consumerId); + mDebug("showing consumer:0x%"PRIx64 " no assigned topic, skip", pConsumer->consumerId); sdbRelease(pSdb, pConsumer); continue; } taosRLockLatch(&pConsumer->lock); - mDebug("showing consumer %" PRId64, pConsumer->consumerId); + mDebug("showing consumer:0x%"PRIx64, pConsumer->consumerId); int32_t topicSz = taosArrayGetSize(pConsumer->assignedTopics); bool hasTopic = true; diff --git a/source/dnode/mnode/impl/src/mndScheduler.c b/source/dnode/mnode/impl/src/mndScheduler.c index af1a29def0..ca79b8e122 100644 --- a/source/dnode/mnode/impl/src/mndScheduler.c +++ b/source/dnode/mnode/impl/src/mndScheduler.c @@ -523,7 +523,7 @@ int32_t mndSchedInitSubEp(SMnode* pMnode, const SMqTopicObj* pTopic, SMqSubscrib SSdb* pSdb = pMnode->pSdb; SVgObj* pVgroup = NULL; SQueryPlan* pPlan = NULL; - SSubplan* plan = NULL; + SSubplan* pSubplan = NULL; if (pTopic->subType == TOPIC_SUB_TYPE__COLUMN) { pPlan = qStringToQueryPlan(pTopic->physicalPlan); @@ -539,24 +539,27 @@ int32_t mndSchedInitSubEp(SMnode* pMnode, const SMqTopicObj* pTopic, SMqSubscrib return -1; } - SNodeListNode* inner = (SNodeListNode*)nodesListGetNode(pPlan->pSubplans, 0); + SNodeListNode* pNodeListNode = (SNodeListNode*)nodesListGetNode(pPlan->pSubplans, 0); - int32_t opNum = LIST_LENGTH(inner->pNodeList); + int32_t opNum = LIST_LENGTH(pNodeListNode->pNodeList); if (opNum != 1) { qDestroyQueryPlan(pPlan); terrno = TSDB_CODE_MND_INVALID_TOPIC_QUERY; return -1; } - plan = (SSubplan*)nodesListGetNode(inner->pNodeList, 0); + + pSubplan = (SSubplan*)nodesListGetNode(pNodeListNode->pNodeList, 0); } ASSERT(pSub->unassignedVgs); - ASSERT(taosHashGetSize(pSub->consumerHash) == 0); void* pIter = NULL; while (1) { pIter = sdbFetch(pSdb, SDB_VGROUP, pIter, (void**)&pVgroup); - if (pIter == NULL) break; + if (pIter == NULL) { + break; + } + if (!mndVgroupInDb(pVgroup, pTopic->dbUid)) { sdbRelease(pSdb, pVgroup); continue; @@ -569,15 +572,15 @@ int32_t mndSchedInitSubEp(SMnode* pMnode, const SMqTopicObj* pTopic, SMqSubscrib pVgEp->vgId = pVgroup->vgId; taosArrayPush(pSub->unassignedVgs, &pVgEp); - mDebug("init subscription %s, assign vg: %d", pSub->key, pVgEp->vgId); + mDebug("init subscription %s for topic:%s assign vgId:%d", pSub->key, pTopic->name, pVgEp->vgId); if (pTopic->subType == TOPIC_SUB_TYPE__COLUMN) { int32_t msgLen; - plan->execNode.epSet = pVgEp->epSet; - plan->execNode.nodeId = pVgEp->vgId; + pSubplan->execNode.epSet = pVgEp->epSet; + pSubplan->execNode.nodeId = pVgEp->vgId; - if (qSubPlanToString(plan, &pVgEp->qmsg, &msgLen) < 0) { + if (qSubPlanToString(pSubplan, &pVgEp->qmsg, &msgLen) < 0) { sdbRelease(pSdb, pVgroup); qDestroyQueryPlan(pPlan); terrno = TSDB_CODE_QRY_INVALID_INPUT; @@ -590,11 +593,7 @@ int32_t mndSchedInitSubEp(SMnode* pMnode, const SMqTopicObj* pTopic, SMqSubscrib sdbRelease(pSdb, pVgroup); } - ASSERT(pSub->unassignedVgs->size > 0); - - ASSERT(taosHashGetSize(pSub->consumerHash) == 0); - + ASSERT(taosArrayGetSize(pSub->unassignedVgs) > 0); qDestroyQueryPlan(pPlan); - return 0; } diff --git a/source/dnode/mnode/impl/src/mndSubscribe.c b/source/dnode/mnode/impl/src/mndSubscribe.c index d127ceacf5..0d805b04fc 100644 --- a/source/dnode/mnode/impl/src/mndSubscribe.c +++ b/source/dnode/mnode/impl/src/mndSubscribe.c @@ -39,12 +39,10 @@ static SSdbRow *mndSubActionDecode(SSdbRaw *pRaw); static int32_t mndSubActionInsert(SSdb *pSdb, SMqSubscribeObj *); static int32_t mndSubActionDelete(SSdb *pSdb, SMqSubscribeObj *); static int32_t mndSubActionUpdate(SSdb *pSdb, SMqSubscribeObj *pOldSub, SMqSubscribeObj *pNewSub); - -static int32_t mndProcessRebalanceReq(SRpcMsg *pMsg); -static int32_t mndProcessDropCgroupReq(SRpcMsg *pMsg); - -static int32_t mndRetrieveSubscribe(SRpcMsg *pReq, SShowObj *pShow, SSDataBlock *pBlock, int32_t rows); -static void mndCancelGetNextSubscribe(SMnode *pMnode, void *pIter); +static int32_t mndProcessRebalanceReq(SRpcMsg *pMsg); +static int32_t mndProcessDropCgroupReq(SRpcMsg *pMsg); +static int32_t mndRetrieveSubscribe(SRpcMsg *pReq, SShowObj *pShow, SSDataBlock *pBlock, int32_t rows); +static void mndCancelGetNextSubscribe(SMnode *pMnode, void *pIter); static int32_t mndSetSubRedoLogs(SMnode *pMnode, STrans *pTrans, SMqSubscribeObj *pSub) { SSdbRaw *pRedoRaw = mndSubActionEncode(pSub); @@ -85,12 +83,13 @@ int32_t mndInitSubscribe(SMnode *pMnode) { return sdbSetTable(pMnode->pSdb, table); } -static SMqSubscribeObj *mndCreateSub(SMnode *pMnode, const SMqTopicObj *pTopic, const char *subKey) { +static SMqSubscribeObj *mndCreateSubscription(SMnode *pMnode, const SMqTopicObj *pTopic, const char *subKey) { SMqSubscribeObj *pSub = tNewSubscribeObj(subKey); if (pSub == NULL) { terrno = TSDB_CODE_OUT_OF_MEMORY; return NULL; } + pSub->dbUid = pTopic->dbUid; pSub->stbUid = pTopic->stbUid; pSub->subType = pTopic->subType; @@ -205,7 +204,7 @@ static SMqRebInfo *mndGetOrCreateRebSub(SHashObj *pHash, const char *key) { static int32_t mndDoRebalance(SMnode *pMnode, const SMqRebInputObj *pInput, SMqRebOutputObj *pOutput) { int32_t totalVgNum = pOutput->pSub->vgNum; const char *sub = pOutput->pSub->key; - mInfo("sub:%s, mq rebalance vgNum:%d", sub, pOutput->pSub->vgNum); + mInfo("sub:%s mq re-balance %d vgroups", sub, pOutput->pSub->vgNum); // 1. build temporary hash(vgId -> SMqRebOutputVg) to store modified vg SHashObj *pHash = taosHashInit(64, taosGetDefaultHashFunction(TSDB_DATA_TYPE_INT), false, HASH_NO_LOCK); @@ -214,7 +213,7 @@ static int32_t mndDoRebalance(SMnode *pMnode, const SMqRebInputObj *pInput, SMqR int32_t removedNum = taosArrayGetSize(pInput->pRebInfo->removedConsumers); int32_t actualRemoved = 0; for (int32_t i = 0; i < removedNum; i++) { - int64_t consumerId = *(int64_t *)taosArrayGet(pInput->pRebInfo->removedConsumers, i); + uint64_t consumerId = *(uint64_t *)taosArrayGet(pInput->pRebInfo->removedConsumers, i); SMqConsumerEp *pConsumerEp = taosHashGet(pOutput->pSub->consumerHash, &consumerId, sizeof(int64_t)); @@ -229,7 +228,7 @@ static int32_t mndDoRebalance(SMnode *pMnode, const SMqRebInputObj *pInput, SMqR .pVgEp = pVgEp, }; taosHashPut(pHash, &pVgEp->vgId, sizeof(int32_t), &outputVg, sizeof(SMqRebOutputVg)); - mInfo("sub:%s, mq rebalance remove vgId:%d from consumer:%" PRId64, sub, pVgEp->vgId, consumerId); + mInfo("sub:%s mq re-balance remove vgId:%d from consumer:%" PRId64, sub, pVgEp->vgId, consumerId); } taosArrayDestroy(pConsumerEp->vgs); taosHashRemove(pOutput->pSub->consumerHash, &consumerId, sizeof(int64_t)); @@ -239,7 +238,7 @@ static int32_t mndDoRebalance(SMnode *pMnode, const SMqRebInputObj *pInput, SMqR } if (removedNum != actualRemoved) { - mError("sub:%s, mq rebalance removedNum:%d not matched with actual:%d", sub, removedNum, actualRemoved); + mError("sub:%s mq re-balance removedNum:%d not matched with actual:%d", sub, removedNum, actualRemoved); } // if previously no consumer, there are vgs not assigned @@ -253,7 +252,7 @@ static int32_t mndDoRebalance(SMnode *pMnode, const SMqRebInputObj *pInput, SMqR .pVgEp = pVgEp, }; taosHashPut(pHash, &pVgEp->vgId, sizeof(int32_t), &rebOutput, sizeof(SMqRebOutputVg)); - mInfo("sub:%s, mq rebalance remove vgId:%d from unassigned", sub, pVgEp->vgId); + mInfo("sub:%s mq re-balance remove vgId:%d from unassigned", sub, pVgEp->vgId); } } @@ -267,7 +266,8 @@ static int32_t mndDoRebalance(SMnode *pMnode, const SMqRebInputObj *pInput, SMqR minVgCnt = totalVgNum / afterRebConsumerNum; imbConsumerNum = totalVgNum % afterRebConsumerNum; } - mInfo("sub:%s, mq rebalance %d consumer after rebalance, at least %d vg each, %d consumer has more vg", sub, + + mInfo("sub:%s mq re-balance %d consumers: at least %d vg each, %d consumer has more vg", sub, afterRebConsumerNum, minVgCnt, imbConsumerNum); // 4. first scan: remove consumer more than wanted, put to remove hash @@ -275,7 +275,10 @@ static int32_t mndDoRebalance(SMnode *pMnode, const SMqRebInputObj *pInput, SMqR void *pIter = NULL; while (1) { pIter = taosHashIterate(pOutput->pSub->consumerHash, pIter); - if (pIter == NULL) break; + if (pIter == NULL) { + break; + } + SMqConsumerEp *pConsumerEp = (SMqConsumerEp *)pIter; int32_t consumerVgNum = taosArrayGetSize(pConsumerEp->vgs); @@ -297,7 +300,7 @@ static int32_t mndDoRebalance(SMnode *pMnode, const SMqRebInputObj *pInput, SMqR .pVgEp = pVgEp, }; taosHashPut(pHash, &pVgEp->vgId, sizeof(int32_t), &outputVg, sizeof(SMqRebOutputVg)); - mInfo("sub:%s, mq rebalance remove vgId:%d from consumer:%" PRId64 ",(first scan)", sub, pVgEp->vgId, + mInfo("sub:%s mq rebalance remove vgId:%d from consumer:%" PRId64 ",(first scan)", sub, pVgEp->vgId, pConsumerEp->consumerId); } imbCnt++; @@ -312,7 +315,7 @@ static int32_t mndDoRebalance(SMnode *pMnode, const SMqRebInputObj *pInput, SMqR .pVgEp = pVgEp, }; taosHashPut(pHash, &pVgEp->vgId, sizeof(int32_t), &outputVg, sizeof(SMqRebOutputVg)); - mInfo("sub:%s, mq rebalance remove vgId:%d from consumer:%" PRId64 ",(first scan)", sub, pVgEp->vgId, + mInfo("sub:%s mq rebalance remove vgId:%d from consumer:%" PRId64 ",(first scan)", sub, pVgEp->vgId, pConsumerEp->consumerId); } } @@ -330,7 +333,7 @@ static int32_t mndDoRebalance(SMnode *pMnode, const SMqRebInputObj *pInput, SMqR newConsumerEp.vgs = taosArrayInit(0, sizeof(void *)); taosHashPut(pOutput->pSub->consumerHash, &consumerId, sizeof(int64_t), &newConsumerEp, sizeof(SMqConsumerEp)); taosArrayPush(pOutput->newConsumers, &consumerId); - mInfo("sub:%s, mq rebalance add new consumer:%" PRId64, sub, consumerId); + mInfo("sub:%s mq rebalance add new consumer:%" PRId64, sub, consumerId); } } @@ -349,7 +352,7 @@ static int32_t mndDoRebalance(SMnode *pMnode, const SMqRebInputObj *pInput, SMqR // iter hash and find one vg pRemovedIter = taosHashIterate(pHash, pRemovedIter); if (pRemovedIter == NULL) { - mError("sub:%s, removed iter is null", sub); + mError("sub:%s removed iter is null", sub); continue; } @@ -402,33 +405,36 @@ static int32_t mndDoRebalance(SMnode *pMnode, const SMqRebInputObj *pInput, SMqR SMqRebOutputVg *pRebOutput = NULL; while (1) { pIter = taosHashIterate(pHash, pIter); - if (pIter == NULL) break; + if (pIter == NULL) { + break; + } + pRebOutput = (SMqRebOutputVg *)pIter; taosArrayPush(pOutput->pSub->unassignedVgs, &pRebOutput->pVgEp); taosArrayPush(pOutput->rebVgs, pRebOutput); - mInfo("sub:%s, mq rebalance unassign vgId:%d (second scan)", sub, pRebOutput->pVgEp->vgId); + mInfo("sub:%s mq re-balance unassign vgId:%d (second scan)", sub, pRebOutput->pVgEp->vgId); } } // 8. generate logs - mInfo("sub:%s, mq rebalance calculation completed, rebalanced vg", sub); + mInfo("sub:%s mq re-balance calculation completed, re-balanced vg", sub); for (int32_t i = 0; i < taosArrayGetSize(pOutput->rebVgs); i++) { SMqRebOutputVg *pOutputRebVg = taosArrayGet(pOutput->rebVgs, i); - mInfo("sub:%s, mq rebalance vgId:%d, moved from consumer:%" PRId64 ", to consumer:%" PRId64, sub, + mInfo("sub:%s mq re-balance vgId:%d, moved from consumer:0x%" PRIx64 ", to consumer:0x%" PRIx64, sub, pOutputRebVg->pVgEp->vgId, pOutputRebVg->oldConsumerId, pOutputRebVg->newConsumerId); } { - void *pIter = NULL; + pIter = NULL; while (1) { pIter = taosHashIterate(pOutput->pSub->consumerHash, pIter); if (pIter == NULL) break; SMqConsumerEp *pConsumerEp = (SMqConsumerEp *)pIter; int32_t sz = taosArrayGetSize(pConsumerEp->vgs); - mInfo("sub:%s, mq rebalance final cfg: consumer %" PRId64 " has %d vg", sub, pConsumerEp->consumerId, sz); + mInfo("sub:%s mq re-balance final cfg: consumer:0x%" PRId64 " has %d vg", sub, pConsumerEp->consumerId, sz); for (int32_t i = 0; i < sz; i++) { SMqVgEp *pVgEp = taosArrayGetP(pConsumerEp->vgs, i); - mInfo("sub:%s, mq rebalance final cfg: vg %d to consumer %" PRId64 "", sub, pVgEp->vgId, + mInfo("sub:%s mq re-balance final cfg: vg %d to consumer:0x%" PRId64, sub, pVgEp->vgId, pConsumerEp->consumerId); } } @@ -552,11 +558,14 @@ static int32_t mndProcessRebalanceReq(SRpcMsg *pMsg) { SMqDoRebalanceMsg *pReq = pMsg->pCont; void *pIter = NULL; - mInfo("mq rebalance start"); + mInfo("mq re-balance start"); while (1) { pIter = taosHashIterate(pReq->rebSubHash, pIter); - if (pIter == NULL) break; + if (pIter == NULL) { + break; + } + SMqRebInputObj rebInput = {0}; SMqRebOutputObj rebOutput = {0}; @@ -577,12 +586,13 @@ static int32_t mndProcessRebalanceReq(SRpcMsg *pMsg) { mndSplitSubscribeKey(pRebInfo->key, topic, cgroup, true); SMqTopicObj *pTopic = mndAcquireTopic(pMnode, topic); if (pTopic == NULL) { - mError("mq rebalance %s failed since topic %s not exist, abort", pRebInfo->key, topic); + mError("mq re-balance %s ignored since topic %s not exist", pRebInfo->key, topic); continue; } + taosRLockLatch(&pTopic->lock); - rebOutput.pSub = mndCreateSub(pMnode, pTopic, pRebInfo->key); + rebOutput.pSub = mndCreateSubscription(pMnode, pTopic, pRebInfo->key); if (rebOutput.pSub == NULL) { mError("mq rebalance %s failed create sub since %s, abort", pRebInfo->key, terrstr()); @@ -605,15 +615,16 @@ static int32_t mndProcessRebalanceReq(SRpcMsg *pMsg) { } if (mndDoRebalance(pMnode, &rebInput, &rebOutput) < 0) { - mError("mq rebalance internal error"); + mError("mq re-balance internal error"); } // if add more consumer to balanced subscribe, // possibly no vg is changed if (mndPersistRebResult(pMnode, pMsg, &rebOutput) < 0) { - mError("mq rebalance persist rebalance output error, possibly vnode splitted or dropped"); + mError("mq re-balance persist re-balance output error, possibly vnode splitted or dropped"); } + taosArrayDestroy(pRebInfo->lostConsumers); taosArrayDestroy(pRebInfo->newConsumers); taosArrayDestroy(pRebInfo->removedConsumers); @@ -627,19 +638,18 @@ static int32_t mndProcessRebalanceReq(SRpcMsg *pMsg) { } // reset flag - mInfo("mq rebalance completed successfully"); + mInfo("mq re-balance completed successfully"); taosHashCleanup(pReq->rebSubHash); mndRebEnd(); return 0; } -static int32_t mndProcessDropCgroupReq(SRpcMsg *pReq) { - SMnode *pMnode = pReq->info.node; - SSdb *pSdb = pMnode->pSdb; +static int32_t mndProcessDropCgroupReq(SRpcMsg *pMsg) { + SMnode *pMnode = pMsg->info.node; SMDropCgroupReq dropReq = {0}; - if (tDeserializeSMDropCgroupReq(pReq->pCont, pReq->contLen, &dropReq) != 0) { + if (tDeserializeSMDropCgroupReq(pMsg->pCont, pMsg->contLen, &dropReq) != 0) { terrno = TSDB_CODE_INVALID_MSG; return -1; } @@ -663,7 +673,7 @@ static int32_t mndProcessDropCgroupReq(SRpcMsg *pReq) { return -1; } - STrans *pTrans = mndTransCreate(pMnode, TRN_POLICY_ROLLBACK, TRN_CONFLICT_NOTHING, pReq, "drop-cgroup"); + STrans *pTrans = mndTransCreate(pMnode, TRN_POLICY_ROLLBACK, TRN_CONFLICT_NOTHING, pMsg, "drop-cgroup"); if (pTrans == NULL) { mError("cgroup: %s on topic:%s, failed to drop since %s", dropReq.cgroup, dropReq.topic, terrstr()); mndReleaseSubscribe(pMnode, pSub); @@ -956,7 +966,7 @@ END: return code; } -static int32_t mndRetrieveSubscribe(SRpcMsg *pReq, SShowObj *pShow, SSDataBlock *pBlock, int32_t rowsCapacity) { +int32_t mndRetrieveSubscribe(SRpcMsg *pReq, SShowObj *pShow, SSDataBlock *pBlock, int32_t rowsCapacity) { SMnode *pMnode = pReq->info.node; SSdb *pSdb = pMnode->pSdb; int32_t numOfRows = 0; @@ -1090,7 +1100,7 @@ static int32_t mndRetrieveSubscribe(SRpcMsg *pReq, SShowObj *pShow, SSDataBlock return numOfRows; } -static void mndCancelGetNextSubscribe(SMnode *pMnode, void *pIter) { +void mndCancelGetNextSubscribe(SMnode *pMnode, void *pIter) { SSdb *pSdb = pMnode->pSdb; sdbCancelFetch(pSdb, pIter); } diff --git a/source/dnode/mnode/impl/src/mndTopic.c b/source/dnode/mnode/impl/src/mndTopic.c index 48c35f3f07..f712825b22 100644 --- a/source/dnode/mnode/impl/src/mndTopic.c +++ b/source/dnode/mnode/impl/src/mndTopic.c @@ -33,7 +33,7 @@ static int32_t mndTopicActionInsert(SSdb *pSdb, SMqTopicObj *pTopic); static int32_t mndTopicActionDelete(SSdb *pSdb, SMqTopicObj *pTopic); -static int32_t mndTopicActionUpdate(SSdb *pSdb, SMqTopicObj *pTopic, SMqTopicObj *pNewTopic); +static int32_t mndTopicActionUpdate(SSdb *pSdb, SMqTopicObj *pOldTopic, SMqTopicObj *pNewTopic); static int32_t mndProcessCreateTopicReq(SRpcMsg *pReq); static int32_t mndProcessDropTopicReq(SRpcMsg *pReq); From 7b193ed9435f06da4f2049da5a3d2f0d664c2717 Mon Sep 17 00:00:00 2001 From: dapan1121 Date: Tue, 21 Feb 2023 15:52:22 +0800 Subject: [PATCH 16/31] fix: interval start window error --- source/libs/executor/src/executil.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/source/libs/executor/src/executil.c b/source/libs/executor/src/executil.c index 040e67713d..b2d2a06953 100644 --- a/source/libs/executor/src/executil.c +++ b/source/libs/executor/src/executil.c @@ -1696,7 +1696,7 @@ static void getInitialStartTimeWindow(SInterval* pInterval, TSKEY ts, STimeWindo int64_t key = w->skey; while (key < ts) { // moving towards end key = taosTimeAdd(key, pInterval->sliding, pInterval->slidingUnit, pInterval->precision); - if (key >= ts) { + if (key > ts) { break; } From 37fea8cf3d53a83b002fff837327884951e538bc Mon Sep 17 00:00:00 2001 From: Haojun Liao Date: Tue, 21 Feb 2023 23:04:58 +0800 Subject: [PATCH 17/31] fix(tmq): update the epset when the leader of vnode changed. --- include/common/tmisce.h | 9 ++++--- include/common/tmsg.h | 2 +- source/client/src/clientTmq.c | 33 +++++++++++++++-------- source/common/src/tmisce.c | 13 +++++++++ source/dnode/mnode/impl/src/mndConsumer.c | 2 +- source/libs/transport/src/transCli.c | 1 + 6 files changed, 43 insertions(+), 17 deletions(-) diff --git a/include/common/tmisce.h b/include/common/tmisce.h index b9f5cf5b91..bc6558900c 100644 --- a/include/common/tmisce.h +++ b/include/common/tmisce.h @@ -27,13 +27,14 @@ typedef struct SCorEpSet { SEpSet epSet; } SCorEpSet; +#define GET_ACTIVE_EP(_eps) (&((_eps)->eps[(_eps)->inUse])) int32_t taosGetFqdnPortFromEp(const char* ep, SEp* pEp); void addEpIntoEpSet(SEpSet* pEpSet, const char* fqdn, uint16_t port); -bool isEpsetEqual(const SEpSet* s1, const SEpSet* s2); - -void updateEpSet_s(SCorEpSet* pEpSet, SEpSet* pNewEpSet); -SEpSet getEpSet_s(SCorEpSet* pEpSet); +bool isEpsetEqual(const SEpSet* s1, const SEpSet* s2); +void epsetAssign(SEpSet* dst, const SEpSet* pSrc); +void updateEpSet_s(SCorEpSet* pEpSet, SEpSet* pNewEpSet); +SEpSet getEpSet_s(SCorEpSet* pEpSet); #ifdef __cplusplus } diff --git a/include/common/tmsg.h b/include/common/tmsg.h index ba13195b97..335c57afd3 100644 --- a/include/common/tmsg.h +++ b/include/common/tmsg.h @@ -1830,7 +1830,7 @@ typedef struct { } SMqConsumerLostMsg, SMqConsumerRecoverMsg, SMqConsumerClearMsg; typedef struct { - uint64_t consumerId; + int64_t consumerId; char cgroup[TSDB_CGROUP_LEN]; char clientId[256]; SArray* topicNames; // SArray diff --git a/source/client/src/clientTmq.c b/source/client/src/clientTmq.c index fb8cdcfde3..1e98f62f36 100644 --- a/source/client/src/clientTmq.c +++ b/source/client/src/clientTmq.c @@ -58,15 +58,14 @@ struct tmq_list_t { }; struct tmq_conf_t { - char clientId[256]; - char groupId[TSDB_CGROUP_LEN]; - int8_t autoCommit; - int8_t resetOffset; - int8_t withTbName; - int8_t snapEnable; - int32_t snapBatchSize; - bool hbBgEnable; - + char clientId[256]; + char groupId[TSDB_CGROUP_LEN]; + int8_t autoCommit; + int8_t resetOffset; + int8_t withTbName; + int8_t snapEnable; + int32_t snapBatchSize; + bool hbBgEnable; uint16_t port; int32_t autoCommitInterval; char* ip; @@ -213,6 +212,7 @@ typedef struct { typedef struct { SMqCommitCbParamSet* params; STqOffset* pOffset; + SMqClientVg* pMqVg; /*char topicName[TSDB_TOPIC_FNAME_LEN];*/ /*int32_t vgId;*/ } SMqCommitCbParam; @@ -440,6 +440,17 @@ int32_t tmqCommitCb(void* param, SDataBuf* pBuf, int32_t code) { } #endif + // there may be race condition. fix it + if (pBuf->pEpSet != NULL && pParam->pMqVg != NULL) { + SMqClientVg* pMqVg = pParam->pMqVg; + + SEp* pEp = GET_ACTIVE_EP(pBuf->pEpSet); + SEp* pOld = GET_ACTIVE_EP(&(pMqVg->epSet)); + uDebug("subKey:%s update the epset vgId:%d, ep:%s:%d, old ep:%s:%d", pParam->pOffset->subKey, pMqVg->vgId, + pEp->fqdn, pEp->port, pOld->fqdn, pOld->port); + pParam->pMqVg->epSet = *pBuf->pEpSet; + } + taosMemoryFree(pParam->pOffset); taosMemoryFree(pBuf->pData); taosMemoryFree(pBuf->pEpSet); @@ -448,7 +459,6 @@ int32_t tmqCommitCb(void* param, SDataBuf* pBuf, int32_t code) { * pOffset->version);*/ tmqCommitRspCountDown(pParamSet); - return 0; } @@ -498,6 +508,7 @@ static int32_t tmqSendCommitReq(tmq_t* tmq, SMqClientVg* pVg, SMqClientTopic* pT pParam->params = pParamSet; pParam->pOffset = pOffset; + pParam->pMqVg = pVg; // there may be an race condition // build send info SMsgSendInfo* pMsgSendInfo = taosMemoryCalloc(1, sizeof(SMsgSendInfo)); @@ -518,7 +529,7 @@ static int32_t tmqSendCommitReq(tmq_t* tmq, SMqClientVg* pVg, SMqClientTopic* pT tscDebug("consumer:0x%" PRIx64 " topic:%s on vgId:%d offset:%" PRId64" prev:%"PRId64", ep:%s:%d", tmq->consumerId, pOffset->subKey, pVg->vgId, pOffset->val.version, pVg->committedOffset.version, pEp->fqdn, pEp->port); - // TODO: put into cb + // TODO: put into cb, the commit offset should be move to the callback function pVg->committedOffset = pVg->currentOffset; pMsgSendInfo->requestId = generateRequestId(); diff --git a/source/common/src/tmisce.c b/source/common/src/tmisce.c index dfb1917fcf..59afce1bbb 100644 --- a/source/common/src/tmisce.c +++ b/source/common/src/tmisce.c @@ -60,6 +60,19 @@ bool isEpsetEqual(const SEpSet* s1, const SEpSet* s2) { return true; } +void epsetAssign(SEpSet* pDst, const SEpSet* pSrc) { + if (pSrc == NULL || pDst == NULL) { + return; + } + + pDst->inUse = pSrc->inUse; + pDst->numOfEps = pSrc->numOfEps; + for (int32_t i = 0; i < pSrc->numOfEps; ++i) { + pDst->eps[i].port = pSrc->eps[i].port; + tstrncpy(pDst->eps[i].fqdn, pSrc->eps[i].fqdn, tListLen(pSrc->eps[i].fqdn)); + } +} + void updateEpSet_s(SCorEpSet* pEpSet, SEpSet* pNewEpSet) { taosCorBeginWrite(&pEpSet->version); pEpSet->epSet = *pNewEpSet; diff --git a/source/dnode/mnode/impl/src/mndConsumer.c b/source/dnode/mnode/impl/src/mndConsumer.c index 40feb22906..d167c9c09a 100644 --- a/source/dnode/mnode/impl/src/mndConsumer.c +++ b/source/dnode/mnode/impl/src/mndConsumer.c @@ -413,7 +413,7 @@ static int32_t mndProcessAskEpReq(SRpcMsg *pMsg) { int32_t serverEpoch = atomic_load_32(&pConsumer->epoch); - // 2. check epoch, only send ep info when epoches do not match + // 2. check epoch, only send ep info when epochs do not match if (epoch != serverEpoch) { taosRLockLatch(&pConsumer->lock); mInfo("process ask ep, consumer:%" PRId64 "(epoch %d), server epoch %d", consumerId, epoch, serverEpoch); diff --git a/source/libs/transport/src/transCli.c b/source/libs/transport/src/transCli.c index 7e1aeafaad..8faaf04108 100644 --- a/source/libs/transport/src/transCli.c +++ b/source/libs/transport/src/transCli.c @@ -2303,6 +2303,7 @@ int transSendRequest(void* shandle, const SEpSet* pEpSet, STransMsg* pReq, STran STransConnCtx* pCtx = taosMemoryCalloc(1, sizeof(STransConnCtx)); pCtx->epSet = *pEpSet; + pCtx->origEpSet = *pEpSet; pCtx->ahandle = pReq->info.ahandle; pCtx->msgType = pReq->msgType; From 4b7c88bcb1cb04ead1fcf00d1e750d4e07f96fdf Mon Sep 17 00:00:00 2001 From: Haojun Liao Date: Wed, 22 Feb 2023 08:52:20 +0800 Subject: [PATCH 18/31] refactor: do some internal refactor. --- source/client/src/clientTmq.c | 32 +++++++++++++++++--------------- 1 file changed, 17 insertions(+), 15 deletions(-) diff --git a/source/client/src/clientTmq.c b/source/client/src/clientTmq.c index 1e98f62f36..d358f4eecf 100644 --- a/source/client/src/clientTmq.c +++ b/source/client/src/clientTmq.c @@ -1080,7 +1080,7 @@ int32_t tmq_subscribe(tmq_t* tmq, const tmq_list_t* topic_list) { SCMSubscribeReq req = {0}; int32_t code = 0; - tscDebug("consumer:0x%"PRIx64" tmq subscribe start, numOfTopic %d", tmq->consumerId, sz); + tscDebug("consumer:0x%"PRIx64" subscribe %d topics", tmq->consumerId, sz); req.consumerId = tmq->consumerId; tstrncpy(req.clientId, tmq->clientId, 256); @@ -1103,7 +1103,7 @@ int32_t tmq_subscribe(tmq_t* tmq, const tmq_list_t* topic_list) { } tNameExtractFullName(&name, topicFName); - tscDebug("consumer:0x%"PRIx64" subscribe topic: %s", tmq->consumerId, topicFName); + tscDebug("consumer:0x%"PRIx64" subscribe topic:%s", tmq->consumerId, topicFName); taosArrayPush(req.topicNames, &topicFName); } @@ -1774,6 +1774,8 @@ int32_t tmqHandleNoPollRsp(tmq_t* tmq, SMqRspWrapper* rspWrapper, bool* pReset) } void* tmqHandleAllRsp(tmq_t* tmq, int64_t timeout, bool pollIfReset) { + tscDebug("consumer:0x%"PRIx64" start to handle the rsp", tmq->consumerId); + while (1) { SMqRspWrapper* rspWrapper = NULL; taosGetQitem(tmq->qall, (void**)&rspWrapper); @@ -1783,20 +1785,17 @@ void* tmqHandleAllRsp(tmq_t* tmq, int64_t timeout, bool pollIfReset) { taosGetQitem(tmq->qall, (void**)&rspWrapper); if (rspWrapper == NULL) { - /*tscDebug("consumer %" PRId64 " mqueue empty", tmq->consumerId);*/ return NULL; } } - tscDebug("consumer:0x%" PRIx64 " handle rsp %p", tmq->consumerId, rspWrapper); - if (rspWrapper->tmqRspType == TMQ_MSG_TYPE__END_RSP) { taosFreeQitem(rspWrapper); terrno = TSDB_CODE_TQ_NO_COMMITTED_OFFSET; return NULL; } else if (rspWrapper->tmqRspType == TMQ_MSG_TYPE__POLL_RSP) { SMqPollRspWrapper* pollRspWrapper = (SMqPollRspWrapper*)rspWrapper; - tscDebug("consumer:0x%" PRIx64 " actual process poll rsp", tmq->consumerId); + tscDebug("consumer:0x%" PRIx64 " process poll rsp", tmq->consumerId); /*atomic_sub_fetch_32(&tmq->readyRequest, 1);*/ int32_t consumerEpoch = atomic_load_32(&tmq->epoch); if (pollRspWrapper->dataRsp.head.epoch == consumerEpoch) { @@ -1823,6 +1822,9 @@ void* tmqHandleAllRsp(tmq_t* tmq, int64_t timeout, bool pollIfReset) { } else if (rspWrapper->tmqRspType == TMQ_MSG_TYPE__POLL_META_RSP) { SMqPollRspWrapper* pollRspWrapper = (SMqPollRspWrapper*)rspWrapper; int32_t consumerEpoch = atomic_load_32(&tmq->epoch); + + tscDebug("consumer:0x%" PRIx64 " process meta rsp", tmq->consumerId); + if (pollRspWrapper->metaRsp.head.epoch == consumerEpoch) { SMqClientVg* pVg = pollRspWrapper->vgHandle; /*printf("vgId:%d, offset %" PRId64 " up to %" PRId64 "\n", pVg->vgId, pVg->currentOffset, @@ -1887,7 +1889,7 @@ TAOS_RES* tmq_consumer_poll(tmq_t* tmq, int64_t timeout) { void* rspObj; int64_t startTime = taosGetTimestampMs(); - tscDebug("consumer:0x%" PRIx64 ", start poll at %" PRId64, tmq->consumerId, startTime); + tscDebug("consumer:0x%" PRIx64 " start to poll at %" PRId64, tmq->consumerId, startTime); #if 0 tmqHandleAllDelayedTask(tmq); @@ -1900,7 +1902,7 @@ TAOS_RES* tmq_consumer_poll(tmq_t* tmq, int64_t timeout) { // in no topic status, delayed task also need to be processed if (atomic_load_8(&tmq->status) == TMQ_CONSUMER_STATUS__INIT) { - tscDebug("consumer:0x%" PRIx64 ", poll return since consumer is init", tmq->consumerId); + tscDebug("consumer:0x%" PRIx64 " poll return since consumer is init", tmq->consumerId); return NULL; } @@ -1926,25 +1928,25 @@ TAOS_RES* tmq_consumer_poll(tmq_t* tmq, int64_t timeout) { rspObj = tmqHandleAllRsp(tmq, timeout, false); if (rspObj) { - tscDebug("consumer:0x%" PRIx64 ", return rsp %p", tmq->consumerId, rspObj); + tscDebug("consumer:0x%" PRIx64 " return rsp %p", tmq->consumerId, rspObj); return (TAOS_RES*)rspObj; } else if (terrno == TSDB_CODE_TQ_NO_COMMITTED_OFFSET) { - tscDebug("consumer:0x%" PRIx64 ", return null since no committed offset", tmq->consumerId); + tscDebug("consumer:0x%" PRIx64 " return null since no committed offset", tmq->consumerId); return NULL; } if (timeout != -1) { int64_t currentTime = taosGetTimestampMs(); - int64_t passedTime = currentTime - startTime; - if (passedTime > timeout) { - tscDebug("consumer:0x%" PRIx64 ", (epoch %d) timeout, no rsp, start time %" PRId64 ", current time %" PRId64, + int64_t elapsedTime = currentTime - startTime; + if (elapsedTime > timeout) { + tscDebug("consumer:0x%" PRIx64 " (epoch %d) timeout, no rsp, start time %" PRId64 ", current time %" PRId64, tmq->consumerId, tmq->epoch, startTime, currentTime); return NULL; } /*tscInfo("consumer:0x%" PRIx64 ", (epoch %d) wait, start time %" PRId64 ", current time %" PRId64*/ /*", left time %" PRId64,*/ - /*tmq->consumerId, tmq->epoch, startTime, currentTime, (timeout - passedTime));*/ - tsem_timewait(&tmq->rspSem, (timeout - passedTime)); + /*tmq->consumerId, tmq->epoch, startTime, currentTime, (timeout - elapsedTime));*/ + tsem_timewait(&tmq->rspSem, (timeout - elapsedTime)); } else { // use tsem_timewait instead of tsem_wait to avoid unexpected stuck tsem_timewait(&tmq->rspSem, 1000); From 3a9a9179d3cad7cacca655710538de8cc63b6111 Mon Sep 17 00:00:00 2001 From: chenhaoran Date: Wed, 22 Feb 2023 10:32:37 +0800 Subject: [PATCH 19/31] ci: install defined version python connector --- tests/parallel_test/container_build.sh | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/tests/parallel_test/container_build.sh b/tests/parallel_test/container_build.sh index ff854449bb..7664704ac1 100755 --- a/tests/parallel_test/container_build.sh +++ b/tests/parallel_test/container_build.sh @@ -55,7 +55,7 @@ fi date docker run \ -v $REP_MOUNT_PARAM \ - --rm --ulimit core=-1 taos_test:v1.0 sh -c "cd $REP_DIR;rm -rf debug;mkdir -p debug;cd debug;cmake .. -DBUILD_HTTP=false -DBUILD_TOOLS=true -DBUILD_TEST=true -DWEBSOCKET=true -DBUILD_TAOSX=true;make -j || exit 1" + --rm --ulimit core=-1 taos_test:v1.0 sh -c ";pip3 install taos-py==2.7.2;cd $REP_DIR;rm -rf debug;mkdir -p debug;cd debug;cmake .. -DBUILD_HTTP=false -DBUILD_TOOLS=true -DBUILD_TEST=true -DWEBSOCKET=true -DBUILD_TAOSX=true;make -j || exit 1" if [[ -d ${WORKDIR}/debugNoSan ]] ;then echo "delete ${WORKDIR}/debugNoSan" @@ -70,7 +70,7 @@ mv ${REP_REAL_PATH}/debug ${WORKDIR}/debugNoSan date docker run \ -v $REP_MOUNT_PARAM \ - --rm --ulimit core=-1 taos_test:v1.0 sh -c "cd $REP_DIR;rm -rf debug;mkdir -p debug;cd debug;cmake .. -DBUILD_HTTP=false -DBUILD_TOOLS=true -DBUILD_TEST=true -DWEBSOCKET=true -DBUILD_SANITIZER=1 -DTOOLS_SANITIZE=true -DTOOLS_BUILD_TYPE=Debug -DBUILD_TAOSX=true;make -j || exit 1 " + --rm --ulimit core=-1 taos_test:v1.0 sh -c ";pip3 install taos-py==2.7.2;cd $REP_DIR;rm -rf debug;mkdir -p debug;cd debug;cmake .. -DBUILD_HTTP=false -DBUILD_TOOLS=true -DBUILD_TEST=true -DWEBSOCKET=true -DBUILD_SANITIZER=1 -DTOOLS_SANITIZE=true -DTOOLS_BUILD_TYPE=Debug -DBUILD_TAOSX=true;make -j || exit 1" mv ${REP_REAL_PATH}/debug ${WORKDIR}/debugSan From d81b2365c4024054e500c76a1791a7880fe56d7c Mon Sep 17 00:00:00 2001 From: chenhaoran Date: Wed, 22 Feb 2023 10:43:54 +0800 Subject: [PATCH 20/31] ci: install defined version python connector --- tests/parallel_test/container_build.sh | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/tests/parallel_test/container_build.sh b/tests/parallel_test/container_build.sh index 7664704ac1..60e83246e2 100755 --- a/tests/parallel_test/container_build.sh +++ b/tests/parallel_test/container_build.sh @@ -55,7 +55,7 @@ fi date docker run \ -v $REP_MOUNT_PARAM \ - --rm --ulimit core=-1 taos_test:v1.0 sh -c ";pip3 install taos-py==2.7.2;cd $REP_DIR;rm -rf debug;mkdir -p debug;cd debug;cmake .. -DBUILD_HTTP=false -DBUILD_TOOLS=true -DBUILD_TEST=true -DWEBSOCKET=true -DBUILD_TAOSX=true;make -j || exit 1" + --rm --ulimit core=-1 taos_test:v1.0 sh -c "pip3 install taos-py==2.7.2;cd $REP_DIR;rm -rf debug;mkdir -p debug;cd debug;cmake .. -DBUILD_HTTP=false -DBUILD_TOOLS=true -DBUILD_TEST=true -DWEBSOCKET=true -DBUILD_TAOSX=true;make -j || exit 1" if [[ -d ${WORKDIR}/debugNoSan ]] ;then echo "delete ${WORKDIR}/debugNoSan" @@ -70,7 +70,7 @@ mv ${REP_REAL_PATH}/debug ${WORKDIR}/debugNoSan date docker run \ -v $REP_MOUNT_PARAM \ - --rm --ulimit core=-1 taos_test:v1.0 sh -c ";pip3 install taos-py==2.7.2;cd $REP_DIR;rm -rf debug;mkdir -p debug;cd debug;cmake .. -DBUILD_HTTP=false -DBUILD_TOOLS=true -DBUILD_TEST=true -DWEBSOCKET=true -DBUILD_SANITIZER=1 -DTOOLS_SANITIZE=true -DTOOLS_BUILD_TYPE=Debug -DBUILD_TAOSX=true;make -j || exit 1" + --rm --ulimit core=-1 taos_test:v1.0 sh -c "pip3 install taos-py==2.7.2;cd $REP_DIR;rm -rf debug;mkdir -p debug;cd debug;cmake .. -DBUILD_HTTP=false -DBUILD_TOOLS=true -DBUILD_TEST=true -DWEBSOCKET=true -DBUILD_SANITIZER=1 -DTOOLS_SANITIZE=true -DTOOLS_BUILD_TYPE=Debug -DBUILD_TAOSX=true;make -j || exit 1" mv ${REP_REAL_PATH}/debug ${WORKDIR}/debugSan From 4c3704fde01fb13ae596867c2e3d1ba71414cec4 Mon Sep 17 00:00:00 2001 From: chenhaoran Date: Wed, 22 Feb 2023 10:50:18 +0800 Subject: [PATCH 21/31] ci: install defined version python connector --- tests/parallel_test/container_build.sh | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/tests/parallel_test/container_build.sh b/tests/parallel_test/container_build.sh index 60e83246e2..4cdf9526fd 100755 --- a/tests/parallel_test/container_build.sh +++ b/tests/parallel_test/container_build.sh @@ -55,7 +55,7 @@ fi date docker run \ -v $REP_MOUNT_PARAM \ - --rm --ulimit core=-1 taos_test:v1.0 sh -c "pip3 install taos-py==2.7.2;cd $REP_DIR;rm -rf debug;mkdir -p debug;cd debug;cmake .. -DBUILD_HTTP=false -DBUILD_TOOLS=true -DBUILD_TEST=true -DWEBSOCKET=true -DBUILD_TAOSX=true;make -j || exit 1" + --rm --ulimit core=-1 taos_test:v1.0 sh -c "pip3 install taospy==2.7.2;cd $REP_DIR;rm -rf debug;mkdir -p debug;cd debug;cmake .. -DBUILD_HTTP=false -DBUILD_TOOLS=true -DBUILD_TEST=true -DWEBSOCKET=true -DBUILD_TAOSX=true;make -j || exit 1" if [[ -d ${WORKDIR}/debugNoSan ]] ;then echo "delete ${WORKDIR}/debugNoSan" @@ -70,7 +70,7 @@ mv ${REP_REAL_PATH}/debug ${WORKDIR}/debugNoSan date docker run \ -v $REP_MOUNT_PARAM \ - --rm --ulimit core=-1 taos_test:v1.0 sh -c "pip3 install taos-py==2.7.2;cd $REP_DIR;rm -rf debug;mkdir -p debug;cd debug;cmake .. -DBUILD_HTTP=false -DBUILD_TOOLS=true -DBUILD_TEST=true -DWEBSOCKET=true -DBUILD_SANITIZER=1 -DTOOLS_SANITIZE=true -DTOOLS_BUILD_TYPE=Debug -DBUILD_TAOSX=true;make -j || exit 1" + --rm --ulimit core=-1 taos_test:v1.0 sh -c "pip3 install taospy==2.7.2;cd $REP_DIR;rm -rf debug;mkdir -p debug;cd debug;cmake .. -DBUILD_HTTP=false -DBUILD_TOOLS=true -DBUILD_TEST=true -DWEBSOCKET=true -DBUILD_SANITIZER=1 -DTOOLS_SANITIZE=true -DTOOLS_BUILD_TYPE=Debug -DBUILD_TAOSX=true;make -j || exit 1" mv ${REP_REAL_PATH}/debug ${WORKDIR}/debugSan From 61a7001274703c11c36654a4d63731a91dafad04 Mon Sep 17 00:00:00 2001 From: Haojun Liao Date: Wed, 22 Feb 2023 10:50:42 +0800 Subject: [PATCH 22/31] fix(query): fix coverity issues. --- examples/c/tmq.c | 37 +++++++++++++++++++------ source/dnode/vnode/src/tsdb/tsdbFile.c | 4 +-- source/libs/executor/src/executil.c | 2 -- source/libs/executor/src/executor.c | 1 - source/libs/executor/src/scanoperator.c | 4 +-- 5 files changed, 32 insertions(+), 16 deletions(-) diff --git a/examples/c/tmq.c b/examples/c/tmq.c index eb41ad039a..8a2112fbcc 100644 --- a/examples/c/tmq.c +++ b/examples/c/tmq.c @@ -189,27 +189,46 @@ void tmq_commit_cb_print(tmq_t* tmq, int32_t code, void* param) { tmq_t* build_consumer() { tmq_conf_res_t code; + tmq_t* tmq = NULL; + tmq_conf_t* conf = tmq_conf_new(); code = tmq_conf_set(conf, "enable.auto.commit", "true"); - if (TMQ_CONF_OK != code) return NULL; + if (TMQ_CONF_OK != code) { + goto _end; + } code = tmq_conf_set(conf, "auto.commit.interval.ms", "1000"); - if (TMQ_CONF_OK != code) return NULL; + if (TMQ_CONF_OK != code) { + goto _end; + } code = tmq_conf_set(conf, "group.id", "cgrpName"); - if (TMQ_CONF_OK != code) return NULL; + if (TMQ_CONF_OK != code) { + goto _end; + } code = tmq_conf_set(conf, "client.id", "user defined name"); - if (TMQ_CONF_OK != code) return NULL; + if (TMQ_CONF_OK != code) { + goto _end; + } code = tmq_conf_set(conf, "td.connect.user", "root"); - if (TMQ_CONF_OK != code) return NULL; + if (TMQ_CONF_OK != code) { + goto _end; + } code = tmq_conf_set(conf, "td.connect.pass", "taosdata"); - if (TMQ_CONF_OK != code) return NULL; + if (TMQ_CONF_OK != code) { + goto _end; + } code = tmq_conf_set(conf, "auto.offset.reset", "earliest"); - if (TMQ_CONF_OK != code) return NULL; + if (TMQ_CONF_OK != code) { + goto _end; + } code = tmq_conf_set(conf, "experimental.snapshot.enable", "false"); - if (TMQ_CONF_OK != code) return NULL; + if (TMQ_CONF_OK != code) { + goto _end; + } tmq_conf_set_auto_commit_cb(conf, tmq_commit_cb_print, NULL); + tmq = tmq_consumer_new(conf, NULL, 0); - tmq_t* tmq = tmq_consumer_new(conf, NULL, 0); + _end: tmq_conf_destroy(conf); return tmq; } diff --git a/source/dnode/vnode/src/tsdb/tsdbFile.c b/source/dnode/vnode/src/tsdb/tsdbFile.c index 42728be657..9b3dbcd8ea 100644 --- a/source/dnode/vnode/src/tsdb/tsdbFile.c +++ b/source/dnode/vnode/src/tsdb/tsdbFile.c @@ -148,7 +148,7 @@ bool tsdbDelFileIsSame(SDelFile *pDelFile1, SDelFile *pDelFile2) { return pDelFi int32_t tsdbDFileRollback(STsdb *pTsdb, SDFileSet *pSet, EDataFileT ftype) { int32_t code = 0; - int64_t size; + int64_t size = 0; int64_t n; TdFilePtr pFD; char fname[TSDB_FILENAME_LEN]; @@ -167,7 +167,7 @@ int32_t tsdbDFileRollback(STsdb *pTsdb, SDFileSet *pSet, EDataFileT ftype) { tPutSmaFile(hdr, pSet->pSmaF); break; default: - ASSERT(0); + goto _err; // make the coverity scan happy } taosCalcChecksumAppend(0, hdr, TSDB_FHDR_SIZE); diff --git a/source/libs/executor/src/executil.c b/source/libs/executor/src/executil.c index 040e67713d..4f4a9fde0f 100644 --- a/source/libs/executor/src/executil.c +++ b/source/libs/executor/src/executil.c @@ -1148,7 +1148,6 @@ int32_t getGroupIdFromTagsVal(void* pMeta, uint64_t uid, SNodeList* pGroupNode, if (TSDB_CODE_SUCCESS == code) { REPLACE_NODE(pNew); } else { - taosMemoryFree(keyBuf); nodesDestroyList(groupNew); metaReaderClear(&mr); return code; @@ -1166,7 +1165,6 @@ int32_t getGroupIdFromTagsVal(void* pMeta, uint64_t uid, SNodeList* pGroupNode, if (pValue->node.resType.type == TSDB_DATA_TYPE_JSON) { if (tTagIsJson(data)) { terrno = TSDB_CODE_QRY_JSON_IN_GROUP_ERROR; - taosMemoryFree(keyBuf); nodesDestroyList(groupNew); metaReaderClear(&mr); return terrno; diff --git a/source/libs/executor/src/executor.c b/source/libs/executor/src/executor.c index 6c354c3d61..ccd03e1efd 100644 --- a/source/libs/executor/src/executor.c +++ b/source/libs/executor/src/executor.c @@ -353,7 +353,6 @@ int32_t qUpdateQualifiedTableId(qTaskInfo_t tinfo, const SArray* tableIdList, bo return code; } - // todo refactor STableList bool assignUid = false; size_t bufLen = (pScanInfo->pGroupTags != NULL) ? getTableTagsBufLen(pScanInfo->pGroupTags) : 0; char* keyBuf = NULL; diff --git a/source/libs/executor/src/scanoperator.c b/source/libs/executor/src/scanoperator.c index c06fc40b9b..1c3dafd0e1 100644 --- a/source/libs/executor/src/scanoperator.c +++ b/source/libs/executor/src/scanoperator.c @@ -3019,8 +3019,8 @@ int32_t tblCountScanGetInputs(SNodeList* groupTags, SName* tableName, STableCoun } } } else { - strncpy(supp->dbNameFilter, tNameGetDbNameP(tableName), TSDB_DB_NAME_LEN); - strncpy(supp->stbNameFilter, tNameGetTableName(tableName), TSDB_TABLE_NAME_LEN); + tstrncpy(supp->dbNameFilter, tNameGetDbNameP(tableName), TSDB_DB_NAME_LEN); + tstrncpy(supp->stbNameFilter, tNameGetTableName(tableName), TSDB_TABLE_NAME_LEN); } return TSDB_CODE_SUCCESS; } From 88d6b8dac7e07d28846fab30627463ef89bd7df6 Mon Sep 17 00:00:00 2001 From: chenhaoran Date: Wed, 22 Feb 2023 11:42:22 +0800 Subject: [PATCH 23/31] ci: install defined version python connector --- tests/parallel_test/container_build.sh | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/tests/parallel_test/container_build.sh b/tests/parallel_test/container_build.sh index 4cdf9526fd..8b88a8007d 100755 --- a/tests/parallel_test/container_build.sh +++ b/tests/parallel_test/container_build.sh @@ -55,7 +55,7 @@ fi date docker run \ -v $REP_MOUNT_PARAM \ - --rm --ulimit core=-1 taos_test:v1.0 sh -c "pip3 install taospy==2.7.2;cd $REP_DIR;rm -rf debug;mkdir -p debug;cd debug;cmake .. -DBUILD_HTTP=false -DBUILD_TOOLS=true -DBUILD_TEST=true -DWEBSOCKET=true -DBUILD_TAOSX=true;make -j || exit 1" + --rm --ulimit core=-1 taos_test:v1.0 sh -c "pip uninstall taospy -y;pip3 install taospy==2.7.2;cd $REP_DIR;rm -rf debug;mkdir -p debug;cd debug;cmake .. -DBUILD_HTTP=false -DBUILD_TOOLS=true -DBUILD_TEST=true -DWEBSOCKET=true -DBUILD_TAOSX=true;make -j || exit 1" if [[ -d ${WORKDIR}/debugNoSan ]] ;then echo "delete ${WORKDIR}/debugNoSan" @@ -70,7 +70,7 @@ mv ${REP_REAL_PATH}/debug ${WORKDIR}/debugNoSan date docker run \ -v $REP_MOUNT_PARAM \ - --rm --ulimit core=-1 taos_test:v1.0 sh -c "pip3 install taospy==2.7.2;cd $REP_DIR;rm -rf debug;mkdir -p debug;cd debug;cmake .. -DBUILD_HTTP=false -DBUILD_TOOLS=true -DBUILD_TEST=true -DWEBSOCKET=true -DBUILD_SANITIZER=1 -DTOOLS_SANITIZE=true -DTOOLS_BUILD_TYPE=Debug -DBUILD_TAOSX=true;make -j || exit 1" + --rm --ulimit core=-1 taos_test:v1.0 sh -c "pip uninstall taospy -y;pip3 install taospy==2.7.2;cd $REP_DIR;rm -rf debug;mkdir -p debug;cd debug;cmake .. -DBUILD_HTTP=false -DBUILD_TOOLS=true -DBUILD_TEST=true -DWEBSOCKET=true -DBUILD_SANITIZER=1 -DTOOLS_SANITIZE=true -DTOOLS_BUILD_TYPE=Debug -DBUILD_TAOSX=true;make -j || exit 1" mv ${REP_REAL_PATH}/debug ${WORKDIR}/debugSan From 82067f3bd29e2155bc84cfd3d6e43d06c9871e8d Mon Sep 17 00:00:00 2001 From: Haojun Liao Date: Wed, 22 Feb 2023 13:38:13 +0800 Subject: [PATCH 24/31] fix(query): fix coverity issues. --- source/dnode/vnode/src/tsdb/tsdbRead.c | 62 ++++++++++++------------- source/libs/executor/src/executorimpl.c | 7 ++- source/util/src/tutil.c | 2 +- 3 files changed, 38 insertions(+), 33 deletions(-) diff --git a/source/dnode/vnode/src/tsdb/tsdbRead.c b/source/dnode/vnode/src/tsdb/tsdbRead.c index 4e1b24750a..f3194a61b7 100644 --- a/source/dnode/vnode/src/tsdb/tsdbRead.c +++ b/source/dnode/vnode/src/tsdb/tsdbRead.c @@ -220,6 +220,8 @@ static bool hasDataInFileBlock(const SBlockData* pBlockData, const SFil static void initBlockDumpInfo(STsdbReader* pReader, SDataBlockIter* pBlockIter); static int32_t getInitialDelIndex(const SArray* pDelSkyline, int32_t order); +static STableBlockScanInfo* getTableBlockScanInfo(SHashObj* pTableMap, uint64_t uid, const char* id); + static FORCE_INLINE STSchema* getLatestTableSchema(STsdbReader* pReader, uint64_t uid); static bool outOfTimeWindow(int64_t ts, STimeWindow* pWindow) { return (ts > pWindow->ekey) || (ts < pWindow->skey); } @@ -699,13 +701,11 @@ static int32_t doLoadBlockIndex(STsdbReader* pReader, SDataFReader* pFileReader, if (pBlockIdx->uid == pList->tableUidList[j]) { // this block belongs to a table that is not queried. - void* p = taosHashGet(pReader->status.pTableMap, &pBlockIdx->uid, sizeof(uint64_t)); - if (p == NULL) { - tsdbError("failed to locate the tableBlockScan Info in hashmap, uid:%"PRIu64", %s", pBlockIdx->uid, pReader->idStr); - return TSDB_CODE_APP_ERROR; + STableBlockScanInfo* pScanInfo = getTableBlockScanInfo(pReader->status.pTableMap, pBlockIdx->uid, pReader->idStr); + if (pScanInfo == NULL) { + return terrno; } - STableBlockScanInfo* pScanInfo = *(STableBlockScanInfo**)p; if (pScanInfo->pBlockList == NULL) { pScanInfo->pBlockList = taosArrayInit(4, sizeof(SBlockIndex)); } @@ -753,9 +753,10 @@ static int32_t doLoadFileBlock(STsdbReader* pReader, SArray* pIndexList, SBlockN for (int32_t i = 0; i < numOfTables; ++i) { SBlockIdx* pBlockIdx = taosArrayGet(pIndexList, i); - - STableBlockScanInfo* pScanInfo = - *(STableBlockScanInfo**)taosHashGet(pReader->status.pTableMap, &pBlockIdx->uid, sizeof(int64_t)); + STableBlockScanInfo* pScanInfo = getTableBlockScanInfo(pReader->status.pTableMap, pBlockIdx->uid, pReader->idStr); + if (pScanInfo == NULL) { + return terrno; + } tMapDataReset(&pScanInfo->mapData); tsdbReadDataBlk(pReader->pFileReader, pBlockIdx, &pScanInfo->mapData); @@ -1257,14 +1258,13 @@ static int32_t fileDataBlockOrderCompar(const void* pLeft, const void* pRight, v static int32_t doSetCurrentBlock(SDataBlockIter* pBlockIter, const char* idStr) { SFileDataBlockInfo* pBlockInfo = getCurrentBlockInfo(pBlockIter); if (pBlockInfo != NULL) { - STableBlockScanInfo** pScanInfo = taosHashGet(pBlockIter->pTableMap, &pBlockInfo->uid, sizeof(pBlockInfo->uid)); + STableBlockScanInfo* pScanInfo = getTableBlockScanInfo(pBlockIter->pTableMap, pBlockInfo->uid, idStr); if (pScanInfo == NULL) { - tsdbError("failed to locate the uid:%" PRIu64 " in query table uid list, %s", pBlockInfo->uid, idStr); - return TSDB_CODE_INVALID_PARA; + return terrno; } - SBlockIndex* pIndex = taosArrayGet((*pScanInfo)->pBlockList, pBlockInfo->tbBlockIdx); - tMapDataGetItemByIdx(&(*pScanInfo)->mapData, pIndex->ordinalIndex, &pBlockIter->block, tGetDataBlk); + SBlockIndex* pIndex = taosArrayGet(pScanInfo->pBlockList, pBlockInfo->tbBlockIdx); + tMapDataGetItemByIdx(&pScanInfo->mapData, pIndex->ordinalIndex, &pBlockIter->block, tGetDataBlk); } #if 0 @@ -2507,16 +2507,11 @@ static int32_t buildComposedDataBlock(STsdbReader* pReader) { STableBlockScanInfo* pBlockScanInfo = NULL; if (pBlockInfo != NULL) { - void* p = taosHashGet(pReader->status.pTableMap, &pBlockInfo->uid, sizeof(pBlockInfo->uid)); - if (p == NULL) { - code = TSDB_CODE_INVALID_PARA; - tsdbError("failed to locate the uid:%" PRIu64 " in query table uid list, total tables:%d, %s", pBlockInfo->uid, - taosHashGetSize(pReader->status.pTableMap), pReader->idStr); + pBlockScanInfo = getTableBlockScanInfo(pReader->status.pTableMap, pBlockInfo->uid, pReader->idStr); + if (pBlockScanInfo == NULL) { goto _end; } - pBlockScanInfo = *(STableBlockScanInfo**)p; - SDataBlk* pBlock = getCurrentBlock(&pReader->status.blockIter); TSDBKEY keyInBuf = getCurrentKeyInBuf(pBlockScanInfo, pReader); @@ -2856,12 +2851,9 @@ static int32_t doBuildDataBlock(STsdbReader* pReader) { SLastBlockReader* pLastBlockReader = pReader->status.fileIter.pLastBlockReader; ASSERT(pBlockInfo != NULL); - - pScanInfo = *(STableBlockScanInfo**)taosHashGet(pReader->status.pTableMap, &pBlockInfo->uid, sizeof(pBlockInfo->uid)); + pScanInfo = getTableBlockScanInfo(pReader->status.pTableMap, pBlockInfo->uid, pReader->idStr); if (pScanInfo == NULL) { - tsdbError("failed to get table scan-info, %s", pReader->idStr); - code = TSDB_CODE_INVALID_PARA; - return code; + return terrno; } pBlock = getCurrentBlock(pBlockIter); @@ -4202,7 +4194,7 @@ int32_t tsdbRetrieveDatablockSMA(STsdbReader* pReader, SSDataBlock* pDataBlock, SSDataBlock* pResBlock = pReader->pResBlock; if (pResBlock->pBlockAgg == NULL) { size_t num = taosArrayGetSize(pResBlock->pDataBlock); - pResBlock->pBlockAgg = taosMemoryCalloc(num, sizeof(SColumnDataAgg)); + pResBlock->pBlockAgg = taosMemoryCalloc(num, POINTER_BYTES); } // do fill all null column value SMA info @@ -4232,6 +4224,18 @@ int32_t tsdbRetrieveDatablockSMA(STsdbReader* pReader, SSDataBlock* pDataBlock, return code; } +STableBlockScanInfo* getTableBlockScanInfo(SHashObj* pTableMap, uint64_t uid, const char* id) { + STableBlockScanInfo** p = taosHashGet(pTableMap, &uid, sizeof(uid)); + if (p == NULL || *p == NULL) { + terrno = TSDB_CODE_INVALID_PARA; + int32_t size = taosHashGetSize(pTableMap); + tsdbError("failed to locate the uid:%" PRIu64 " in query table uid list, total tables:%d, %s", uid, size, id); + return NULL; + } + + return *p; +} + static SSDataBlock* doRetrieveDataBlock(STsdbReader* pReader) { SReaderStatus* pStatus = &pReader->status; @@ -4240,12 +4244,8 @@ static SSDataBlock* doRetrieveDataBlock(STsdbReader* pReader) { } SFileDataBlockInfo* pBlockInfo = getCurrentBlockInfo(&pStatus->blockIter); - STableBlockScanInfo* pBlockScanInfo = - *(STableBlockScanInfo**)taosHashGet(pStatus->pTableMap, &pBlockInfo->uid, sizeof(pBlockInfo->uid)); + STableBlockScanInfo* pBlockScanInfo = getTableBlockScanInfo(pStatus->pTableMap, pBlockInfo->uid, pReader->idStr); if (pBlockScanInfo == NULL) { - terrno = TSDB_CODE_INVALID_PARA; - tsdbError("failed to locate the uid:%" PRIu64 " in query table uid list, total tables:%d, %s", pBlockInfo->uid, - taosHashGetSize(pReader->status.pTableMap), pReader->idStr); return NULL; } diff --git a/source/libs/executor/src/executorimpl.c b/source/libs/executor/src/executorimpl.c index d28c3cfe58..991756cbe9 100644 --- a/source/libs/executor/src/executorimpl.c +++ b/source/libs/executor/src/executorimpl.c @@ -2022,7 +2022,12 @@ int32_t extractTableSchemaInfo(SReadHandle* pHandle, SScanPhysiNode* pScanNode, tDecoderClear(&mr.coder); tb_uid_t suid = mr.me.ctbEntry.suid; - metaGetTableEntryByUidCache(&mr, suid); + code = metaGetTableEntryByUidCache(&mr, suid); + if (code != TSDB_CODE_SUCCESS) { + metaReaderClear(&mr); + return terrno; + } + pSchemaInfo->sw = tCloneSSchemaWrapper(&mr.me.stbEntry.schemaRow); pSchemaInfo->tversion = mr.me.stbEntry.schemaTag.version; } else { diff --git a/source/util/src/tutil.c b/source/util/src/tutil.c index 55d7d4f6e7..1f9ca7407e 100644 --- a/source/util/src/tutil.c +++ b/source/util/src/tutil.c @@ -118,7 +118,7 @@ char **strsplit(char *z, const char *delim, int32_t *num) { if ((*num) >= size) { size = (size << 1); split = taosMemoryRealloc(split, POINTER_BYTES * size); - ASSERTS(NULL != split, "realloc memory failed. size=%d", POINTER_BYTES * size); + ASSERTS(NULL != split, "realloc memory failed. size=%d", (int32_t) POINTER_BYTES * size); } } From 8532de31cf5a81a3aef0a39d44246428aca190db Mon Sep 17 00:00:00 2001 From: Haojun Liao Date: Wed, 22 Feb 2023 14:32:39 +0800 Subject: [PATCH 25/31] refactor: remove assert. --- source/client/src/clientTmq.c | 7 ++--- source/dnode/vnode/src/tsdb/tsdbRead.c | 5 +--- source/libs/executor/src/dataDeleter.c | 3 --- source/libs/executor/src/executil.c | 9 ------- source/libs/executor/src/executor.c | 27 ++++++++----------- source/libs/executor/src/executorimpl.c | 33 ++++++------------------ source/libs/executor/src/filloperator.c | 2 -- source/libs/executor/src/groupoperator.c | 3 --- source/libs/executor/src/sortoperator.c | 2 -- 9 files changed, 22 insertions(+), 69 deletions(-) diff --git a/source/client/src/clientTmq.c b/source/client/src/clientTmq.c index d358f4eecf..69e10bd649 100644 --- a/source/client/src/clientTmq.c +++ b/source/client/src/clientTmq.c @@ -422,7 +422,6 @@ int32_t tmqCommitDone(SMqCommitCbParamSet* pParamSet) { static void tmqCommitRspCountDown(SMqCommitCbParamSet* pParamSet) { int32_t waitingRspNum = atomic_sub_fetch_32(&pParamSet->waitingRspNum, 1); - ASSERT(waitingRspNum >= 0); if (waitingRspNum == 0) { tmqCommitDone(pParamSet); } @@ -551,7 +550,7 @@ static int32_t tmqSendCommitReq(tmq_t* tmq, SMqClientVg* pVg, SMqClientTopic* pT int32_t tmqCommitMsgImpl(tmq_t* tmq, const TAOS_RES* msg, int8_t async, tmq_commit_cb* userCb, void* userParam) { char* topic; int32_t vgId; - ASSERT(msg != NULL); + if (TD_RES_TMQ(msg)) { SMqRspObj* pRspObj = (SMqRspObj*)msg; topic = pRspObj->topic; @@ -994,14 +993,12 @@ tmq_t* tmq_consumer_new(tmq_conf_t* conf, char* errstr, int32_t errstrLen) { const char* user = conf->user == NULL ? TSDB_DEFAULT_USER : conf->user; const char* pass = conf->pass == NULL ? TSDB_DEFAULT_PASS : conf->pass; - ASSERT(conf->groupId[0]); - pTmq->clientTopics = taosArrayInit(0, sizeof(SMqClientTopic)); pTmq->mqueue = taosOpenQueue(); pTmq->qall = taosAllocateQall(); pTmq->delayedTask = taosOpenQueue(); - if (pTmq->clientTopics == NULL || pTmq->mqueue == NULL || pTmq->qall == NULL || pTmq->delayedTask == NULL) { + if (pTmq->clientTopics == NULL || pTmq->mqueue == NULL || pTmq->qall == NULL || pTmq->delayedTask == NULL || conf->groupId[0] == 0) { terrno = TSDB_CODE_OUT_OF_MEMORY; tscError("consumer:0x%" PRIx64 " setup failed since %s, consumer group %s", pTmq->consumerId, terrstr(), pTmq->groupId); diff --git a/source/dnode/vnode/src/tsdb/tsdbRead.c b/source/dnode/vnode/src/tsdb/tsdbRead.c index f3194a61b7..78793e6a95 100644 --- a/source/dnode/vnode/src/tsdb/tsdbRead.c +++ b/source/dnode/vnode/src/tsdb/tsdbRead.c @@ -855,9 +855,7 @@ static int doBinarySearchKey(TSKEY* keyList, int num, int pos, TSKEY key, int or s = pos; // check - assert(pos >= 0 && pos < num); - assert(num > 0); - + ASSERT(pos >= 0 && pos < num && num > 0); if (order == TSDB_ORDER_ASC) { // find the first position which is smaller than the key e = num - 1; @@ -2850,7 +2848,6 @@ static int32_t doBuildDataBlock(STsdbReader* pReader) { SFileDataBlockInfo* pBlockInfo = getCurrentBlockInfo(pBlockIter); SLastBlockReader* pLastBlockReader = pReader->status.fileIter.pLastBlockReader; - ASSERT(pBlockInfo != NULL); pScanInfo = getTableBlockScanInfo(pReader->status.pTableMap, pBlockInfo->uid, pReader->idStr); if (pScanInfo == NULL) { return terrno; diff --git a/source/libs/executor/src/dataDeleter.c b/source/libs/executor/src/dataDeleter.c index a8051ea7c3..2366d0c2f1 100644 --- a/source/libs/executor/src/dataDeleter.c +++ b/source/libs/executor/src/dataDeleter.c @@ -62,9 +62,6 @@ static void toDataCacheEntry(SDataDeleterHandle* pHandle, const SInputData* pInp pEntry->numOfCols = taosArrayGetSize(pInput->pData->pDataBlock); pEntry->dataLen = sizeof(SDeleterRes); -// ASSERT(1 == pEntry->numOfRows); -// ASSERT(3 == pEntry->numOfCols); - pBuf->useSize = sizeof(SDataCacheEntry); SColumnInfoData* pColRes = (SColumnInfoData*)taosArrayGet(pInput->pData->pDataBlock, 0); diff --git a/source/libs/executor/src/executil.c b/source/libs/executor/src/executil.c index 4f4a9fde0f..c185ab7675 100644 --- a/source/libs/executor/src/executil.c +++ b/source/libs/executor/src/executil.c @@ -170,7 +170,6 @@ void initGroupedResultInfo(SGroupResInfo* pGroupResInfo, SSHashObj* pHashmap, in } pGroupResInfo->index = 0; - assert(pGroupResInfo->index <= getNumOfTotalRes(pGroupResInfo)); } void initMultiResInfoFromArrayList(SGroupResInfo* pGroupResInfo, SArray* pArrayList) { @@ -334,10 +333,7 @@ int32_t isQualifiedTable(STableKeyInfo* info, SNode* pTagCond, void* metaHandle, return code; } - ASSERT(nodeType(pNew) == QUERY_NODE_VALUE); SValueNode* pValue = (SValueNode*)pNew; - - ASSERT(pValue->node.resType.type == TSDB_DATA_TYPE_BOOL); *pQualified = pValue->datum.b; nodesDestroyNode(pNew); @@ -1056,7 +1052,6 @@ int32_t getTableList(void* metaHandle, void* pVnode, SScanPhysiNode* pScanNode, } if (!pTagCond) { // no tag filter condition exists, let's fetch all tables of this super table - ASSERT(pTagIndexCond == NULL); vnodeGetCtbIdList(pVnode, pScanNode->suid, pUidList); } else { // failed to find the result in the cache, let try to calculate the results @@ -1366,7 +1361,6 @@ void createExprFromOneNode(SExprInfo* pExp, SNode* pNode, int16_t slotId) { if (!pFuncNode->pParameterList && (memcmp(pExprNode->_function.functionName, name, len) == 0) && pExprNode->_function.functionName[len] == 0) { pFuncNode->pParameterList = nodesMakeList(); - ASSERT(LIST_LENGTH(pFuncNode->pParameterList) == 0); SValueNode* res = (SValueNode*)nodesMakeNode(QUERY_NODE_VALUE); if (NULL == res) { // todo handle error } else { @@ -1806,7 +1800,6 @@ uint64_t getTableGroupId(const STableListInfo* pTableList, uint64_t tableUid) { // TODO handle the group offset info, fix it, the rule of group output will be broken by this function int32_t tableListAddTableInfo(STableListInfo* pTableList, uint64_t uid, uint64_t gid) { if (pTableList->map == NULL) { - ASSERT(taosArrayGetSize(pTableList->pTableList) == 0); pTableList->map = taosHashInit(32, taosGetDefaultHashFunction(TSDB_DATA_TYPE_BINARY), false, HASH_ENTRY_LOCK); } @@ -1956,7 +1949,6 @@ static int32_t sortTableGroup(STableListInfo* pTableListInfo) { int32_t buildGroupIdMapForAllTables(STableListInfo* pTableListInfo, SReadHandle* pHandle, SNodeList* group, bool groupSort) { int32_t code = TSDB_CODE_SUCCESS; - ASSERT(pTableListInfo->map != NULL); bool groupByTbname = groupbyTbname(group); size_t numOfTables = taosArrayGetSize(pTableListInfo->pTableList); @@ -2013,7 +2005,6 @@ int32_t createScanTableListInfo(SScanPhysiNode* pScanNode, SNodeList* pGroupTags } int32_t numOfTables = taosArrayGetSize(pTableListInfo->pTableList); - ASSERT(pTableListInfo->numOfOuputGroups == 1); int64_t st1 = taosGetTimestampUs(); pTaskInfo->cost.extractListTime = (st1 - st) / 1000.0; diff --git a/source/libs/executor/src/executor.c b/source/libs/executor/src/executor.c index ccd03e1efd..8f8b32ff07 100644 --- a/source/libs/executor/src/executor.c +++ b/source/libs/executor/src/executor.c @@ -35,7 +35,6 @@ static void initRefPool() { } static int32_t doSetSMABlock(SOperatorInfo* pOperator, void* input, size_t numOfBlocks, int32_t type, char* id) { - ASSERT(pOperator != NULL); if (pOperator->operatorType != QUERY_NODE_PHYSICAL_PLAN_STREAM_SCAN) { if (pOperator->numOfDownstream == 0) { qError("failed to find stream scan operator to set the input data block, %s" PRIx64, id); @@ -75,27 +74,23 @@ static int32_t doSetSMABlock(SOperatorInfo* pOperator, void* input, size_t numOf } static int32_t doSetStreamOpOpen(SOperatorInfo* pOperator, char* id) { - { - ASSERT(pOperator != NULL); - if (pOperator->operatorType != QUERY_NODE_PHYSICAL_PLAN_STREAM_SCAN) { - if (pOperator->numOfDownstream == 0) { - qError("failed to find stream scan operator to set the input data block, %s" PRIx64, id); - return TSDB_CODE_APP_ERROR; - } - - if (pOperator->numOfDownstream > 1) { // not handle this in join query - qError("join not supported for stream block scan, %s" PRIx64, id); - return TSDB_CODE_APP_ERROR; - } - pOperator->status = OP_NOT_OPENED; - return doSetStreamOpOpen(pOperator->pDownstream[0], id); + if (pOperator->operatorType != QUERY_NODE_PHYSICAL_PLAN_STREAM_SCAN) { + if (pOperator->numOfDownstream == 0) { + qError("failed to find stream scan operator to set the input data block, %s" PRIx64, id); + return TSDB_CODE_APP_ERROR; } + + if (pOperator->numOfDownstream > 1) { // not handle this in join query + qError("join not supported for stream block scan, %s" PRIx64, id); + return TSDB_CODE_APP_ERROR; + } + pOperator->status = OP_NOT_OPENED; + return doSetStreamOpOpen(pOperator->pDownstream[0], id); } return 0; } static int32_t doSetStreamBlock(SOperatorInfo* pOperator, void* input, size_t numOfBlocks, int32_t type, char* id) { - ASSERT(pOperator != NULL); if (pOperator->operatorType != QUERY_NODE_PHYSICAL_PLAN_STREAM_SCAN) { if (pOperator->numOfDownstream == 0) { qError("failed to find stream scan operator to set the input data block, %s" PRIx64, id); diff --git a/source/libs/executor/src/executorimpl.c b/source/libs/executor/src/executorimpl.c index 991756cbe9..35c884c1a1 100644 --- a/source/libs/executor/src/executorimpl.c +++ b/source/libs/executor/src/executorimpl.c @@ -240,7 +240,6 @@ SResultRow* doSetResultOutBufByKey(SDiskbasedBuf* pResultBuf, SResultRowInfo* pR // allocate a new buffer page if (pResult == NULL) { - ASSERT(pSup->resultRowSize > 0); pResult = getNewResultRow(pResultBuf, &pSup->currentPageId, pSup->resultRowSize); if (pResult == NULL) { T_LONG_JMP(pTaskInfo->env, terrno); @@ -310,7 +309,6 @@ static int32_t addNewWindowResultBuf(SResultRow* pWindowRes, SDiskbasedBuf* pRes pWindowRes->offset = (int32_t)pData->num; pData->num += size; - assert(pWindowRes->pageId >= 0); } return 0; @@ -488,7 +486,6 @@ static int32_t doSetInputDataBlock(SExprSupp* pExprSup, SSDataBlock* pBlock, int // todo: refactor this if (fmIsImplicitTsFunc(pCtx[i].functionId) && (j == pOneExpr->base.numOfParams - 1)) { pInput->pPTS = pInput->pData[j]; // in case of merge function, this is not always the ts column data. - // ASSERT(pInput->pPTS->info.type == TSDB_DATA_TYPE_TIMESTAMP); } ASSERT(pInput->pData[j] != NULL); } else if (pFuncParam->type == FUNC_PARAM_TYPE_VALUE) { @@ -1024,8 +1021,6 @@ void doSetTableGroupOutputBuf(SOperatorInfo* pOperator, int32_t numOfOutput, uin SResultRow* pResultRow = doSetResultOutBufByKey(pAggInfo->aggSup.pResultBuf, pResultRowInfo, (char*)&groupId, sizeof(groupId), true, groupId, pTaskInfo, false, &pAggInfo->aggSup); - assert(pResultRow != NULL); - /* * not assign result buffer yet, add new result buffer * all group belong to one result set, and each group result has different group id so set the id to be one @@ -1279,7 +1274,6 @@ void doBuildResultDatablock(SOperatorInfo* pOperator, SOptrBasicInfo* pbInfo, SG // STaskAttr *pQueryAttr = pRuntimeEnv->pQueryAttr; // SResultRowInfo *pWindowResInfo = &pRuntimeEnv->resultRowInfo; // -// assert(pQueryAttr->limit.offset == 0); // STimeWindow tw = *win; // getNextTimeWindow(pQueryAttr, &tw); // @@ -1294,7 +1288,6 @@ void doBuildResultDatablock(SOperatorInfo* pOperator, SOptrBasicInfo* pbInfo, SG // tw = *win; // int32_t startPos = // getNextQualifiedWindow(pQueryAttr, &tw, pBlockInfo, pColInfoData->pData, binarySearchForKey, -1); -// assert(startPos >= 0); // // // set the abort info // pQueryAttr->pos = startPos; @@ -1329,11 +1322,6 @@ void doBuildResultDatablock(SOperatorInfo* pOperator, SOptrBasicInfo* pbInfo, SG // static bool skipTimeInterval(STaskRuntimeEnv *pRuntimeEnv, TSKEY* start) { // STaskAttr *pQueryAttr = pRuntimeEnv->pQueryAttr; -// if (QUERY_IS_ASC_QUERY(pQueryAttr)) { -// assert(*start <= pRuntimeEnv->current->lastKey); -// } else { -// assert(*start >= pRuntimeEnv->current->lastKey); -// } // // // if queried with value filter, do NOT forward query start position // if (pQueryAttr->limit.offset <= 0 || pQueryAttr->numOfFilterCols > 0 || pRuntimeEnv->pTsBuf != NULL || @@ -1347,8 +1335,6 @@ void doBuildResultDatablock(SOperatorInfo* pOperator, SOptrBasicInfo* pbInfo, SG // value is // * not valid. otherwise, we only forward pQueryAttr->limit.offset number of points // */ -// assert(pRuntimeEnv->resultRowInfo.prevSKey == TSKEY_INITIAL_VAL); -// // STimeWindow w = TSWINDOW_INITIALIZER; // bool ascQuery = QUERY_IS_ASC_QUERY(pQueryAttr); // @@ -1418,8 +1404,6 @@ void doBuildResultDatablock(SOperatorInfo* pOperator, SOptrBasicInfo* pbInfo, SG // tw = win; // int32_t startPos = // getNextQualifiedWindow(pQueryAttr, &tw, &blockInfo, pColInfoData->pData, binarySearchForKey, -1); -// assert(startPos >= 0); -// // // set the abort info // pQueryAttr->pos = startPos; // pTableQueryInfo->lastKey = ((TSKEY *)pColInfoData->pData)[startPos]; @@ -1441,10 +1425,6 @@ void doBuildResultDatablock(SOperatorInfo* pOperator, SOptrBasicInfo* pbInfo, SG // } int32_t appendDownstream(SOperatorInfo* p, SOperatorInfo** pDownstream, int32_t num) { - if (p->pDownstream == NULL) { - assert(p->numOfDownstream == 0); - } - p->pDownstream = taosMemoryCalloc(1, num * POINTER_BYTES); if (p->pDownstream == NULL) { return TSDB_CODE_OUT_OF_MEMORY; @@ -1800,7 +1780,10 @@ int32_t initAggSup(SExprSupp* pSup, SAggSupporter* pAggSup, SExprInfo* pExprInfo } void initResultSizeInfo(SResultInfo* pResultInfo, int32_t numOfRows) { - ASSERT(numOfRows != 0); + if (numOfRows == 0) { + numOfRows = 4096; + } + pResultInfo->capacity = numOfRows; pResultInfo->threshold = numOfRows * 0.75; @@ -1941,7 +1924,6 @@ _error: } void cleanupBasicInfo(SOptrBasicInfo* pInfo) { - assert(pInfo != NULL); pInfo->pRes = blockDataDestroy(pInfo->pRes); } @@ -2253,7 +2235,8 @@ SOperatorInfo* createOperatorTree(SPhysiNode* pPhyNode, SExecTaskInfo* pTaskInfo } else if (QUERY_NODE_PHYSICAL_PLAN_PROJECT == type) { pOperator = createProjectOperatorInfo(NULL, (SProjectPhysiNode*)pPhyNode, pTaskInfo); } else { - ASSERT(0); + terrno = TSDB_CODE_INVALID_PARA; + return NULL; } if (pOperator != NULL) { @@ -2345,7 +2328,8 @@ SOperatorInfo* createOperatorTree(SPhysiNode* pPhyNode, SExecTaskInfo* pTaskInfo } else if (QUERY_NODE_PHYSICAL_PLAN_INTERP_FUNC == type) { pOptr = createTimeSliceOperatorInfo(ops[0], pPhyNode, pTaskInfo); } else { - ASSERT(0); + terrno = TSDB_CODE_INVALID_PARA; + return NULL; } taosMemoryFree(ops); @@ -2583,7 +2567,6 @@ int32_t setOutputBuf(SStreamState* pState, STimeWindow* win, SResultRow** pResul return TSDB_CODE_OUT_OF_MEMORY; } *pResult = (SResultRow*)value; - ASSERT(*pResult); // set time window for current result (*pResult)->win = (*win); setResultRowInitCtx(*pResult, pCtx, numOfOutput, rowEntryInfoOffset); diff --git a/source/libs/executor/src/filloperator.c b/source/libs/executor/src/filloperator.c index 16983cb507..f7de84d085 100644 --- a/source/libs/executor/src/filloperator.c +++ b/source/libs/executor/src/filloperator.c @@ -193,8 +193,6 @@ static SSDataBlock* doFillImpl(SOperatorInfo* pOperator) { return pResBlock; } } else if (pInfo->existNewGroupBlock) { // try next group - assert(pBlock != NULL); - blockDataCleanup(pResBlock); doHandleRemainBlockForNewGroupImpl(pOperator, pInfo, pResultInfo, pTaskInfo); diff --git a/source/libs/executor/src/groupoperator.c b/source/libs/executor/src/groupoperator.c index 9d88126220..4e17e1a3f2 100644 --- a/source/libs/executor/src/groupoperator.c +++ b/source/libs/executor/src/groupoperator.c @@ -204,7 +204,6 @@ static void recordNewGroupKeys(SArray* pGroupCols, SArray* pGroupColVals, SSData } static int32_t buildGroupKeys(void* pKey, const SArray* pGroupColVals) { - ASSERT(pKey != NULL); size_t numOfGroupCols = taosArrayGetSize(pGroupColVals); char* isNull = (char*)pKey; @@ -570,7 +569,6 @@ static void doHashPartition(SOperatorInfo* pOperator, SSDataBlock* pBlock) { } (*columnLen) += contentLen; - ASSERT(*columnLen >= 0); } (*rows) += 1; @@ -681,7 +679,6 @@ static int compareDataGroupInfo(const void* group1, const void* group2) { const SDataGroupInfo* pGroupInfo2 = group2; if (pGroupInfo1->groupId == pGroupInfo2->groupId) { - ASSERT(0); return 0; } diff --git a/source/libs/executor/src/sortoperator.c b/source/libs/executor/src/sortoperator.c index 98ef6b8a36..3db29f40fd 100644 --- a/source/libs/executor/src/sortoperator.c +++ b/source/libs/executor/src/sortoperator.c @@ -768,8 +768,6 @@ SOperatorInfo* createMultiwayMergeOperatorInfo(SOperatorInfo** downStreams, size pInfo->binfo.pRes = createDataBlockFromDescNode(pDescNode); int32_t rowSize = pInfo->binfo.pRes->info.rowSize; - ASSERT(rowSize < 100 * 1024 * 1024); - int32_t numOfOutputCols = 0; code = extractColMatchInfo(pMergePhyNode->pTargets, pDescNode, &numOfOutputCols, COL_MATCH_FROM_SLOT_ID, &pInfo->matchInfo); From 96389577d73549438bdcc941167c3c3581d05e40 Mon Sep 17 00:00:00 2001 From: chenhaoran Date: Wed, 22 Feb 2023 14:42:17 +0800 Subject: [PATCH 26/31] fix: ingore filter.c error found by asan --- tests/script/sh/checkAsan.sh | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/tests/script/sh/checkAsan.sh b/tests/script/sh/checkAsan.sh index 7225722791..5c7976a9fc 100755 --- a/tests/script/sh/checkAsan.sh +++ b/tests/script/sh/checkAsan.sh @@ -38,8 +38,9 @@ python_error=`cat ${LOG_DIR}/*.info | grep -w "stack" | wc -l` # /root/TDengine/source/libs/function/src/builtinsimpl.c:856:29: runtime error: signed integer overflow: 9223372036854775806 + 9223372036854775805 cannot be represented in type 'long int' # /root/TDengine/source/libs/scalar/src/sclvector.c:1075:66: runtime error: signed integer overflow: 9223372034707292160 + 1668838476672 cannot be represented in type 'long int' # /root/TDengine/source/common/src/tdataformat.c:1876:7: runtime error: signed integer overflow: 8252423483843671206 + 2406154664059062870 cannot be represented in type 'long int' +# /home/chr/TDengine/source/libs/scalar/src/filter.c:3149:14: runtime error: applying non-zero offset 18446744073709551615 to null pointer -runtime_error=`cat ${LOG_DIR}/*.asan | grep "runtime error" | grep -v "trees.c:873" | grep -v "sclfunc.c.*outside the range of representable values of type"| grep -v "signed integer overflow" |grep -v "strerror.c"| grep -v "asan_malloc_linux.cc" |wc -l` +runtime_error=`cat ${LOG_DIR}/*.asan | grep "runtime error" | grep -v "trees.c:873" | grep -v "sclfunc.c.*outside the range of representable values of type"| grep -v "signed integer overflow" |grep -v "strerror.c"| grep -v "asan_malloc_linux.cc" |grep -v "filter.c:3149:14"|wc -l` echo -e "\033[44;32;1m"asan error_num: $error_num"\033[0m" echo -e "\033[44;32;1m"asan memory_leak: $memory_leak"\033[0m" From cd1089acb3c9399512a853e7af3e5b9492cd4bae Mon Sep 17 00:00:00 2001 From: Haojun Liao Date: Wed, 22 Feb 2023 16:43:37 +0800 Subject: [PATCH 27/31] refactor: fix coverity issues. --- source/libs/executor/src/tsort.c | 1 + 1 file changed, 1 insertion(+) diff --git a/source/libs/executor/src/tsort.c b/source/libs/executor/src/tsort.c index 90f20f40b8..a656b8e3fd 100644 --- a/source/libs/executor/src/tsort.c +++ b/source/libs/executor/src/tsort.c @@ -213,6 +213,7 @@ static int32_t doAddToBuf(SSDataBlock* pDataBlock, SSortHandle* pHandle) { void* pPage = getNewBufPage(pHandle->pBuf, &pageId); if (pPage == NULL) { blockDataDestroy(p); + taosArrayDestroy(pPageIdList); return terrno; } From edc64122cb8d5a72a3dc092bd97d30c98b0d2e6d Mon Sep 17 00:00:00 2001 From: Haojun Liao Date: Wed, 22 Feb 2023 16:50:47 +0800 Subject: [PATCH 28/31] refactor: fix coverity issues. --- utils/test/c/tmqSim.c | 13 +++++++------ 1 file changed, 7 insertions(+), 6 deletions(-) diff --git a/utils/test/c/tmqSim.c b/utils/test/c/tmqSim.c index 9c1dc2e063..1acf50a7d8 100644 --- a/utils/test/c/tmqSim.c +++ b/utils/test/c/tmqSim.c @@ -400,7 +400,7 @@ TAOS* createNewTaosConnect() { int32_t retryCnt = 10; while (retryCnt--) { - TAOS* taos = taos_connect(NULL, "root", "taosdata", NULL, 0); + taos = taos_connect(NULL, "root", "taosdata", NULL, 0); if (NULL != taos) { return taos; } @@ -780,7 +780,8 @@ void loop_consume(SThreadInfo* pInfo) { if (pInfo->ifCheckData) { char filename[256] = {0}; - char tmpString[128]; + memset(tmpString, 0, tListLen(tmpString)); + // sprintf(filename, "%s/../log/consumerid_%d_%s.txt", configDir, pInfo->consumerId, // getCurrentTimeString(tmpString)); sprintf(filename, "%s/../log/consumerid_%d.txt", configDir, pInfo->consumerId); @@ -834,12 +835,12 @@ void loop_consume(SThreadInfo* pInfo) { } if ((totalRows >= pInfo->expectMsgCnt) || (totalMsgs >= pInfo->expectMsgCnt)) { - char tmpString[128]; + memset(tmpString, 0, tListLen(tmpString)); taosFprintfFile(g_fp, "%s over than expect rows, so break consume\n", getCurrentTimeString(tmpString)); break; } } else { - char tmpString[128]; + memset(tmpString, 0, tListLen(tmpString)); taosFprintfFile(g_fp, "%s no poll more msg when time over, break consume\n", getCurrentTimeString(tmpString)); break; } @@ -1113,7 +1114,7 @@ void omb_loop_consume(SThreadInfo* pInfo) { lastTotalLenOfMsg = totalLenOfMsg; } } else { - char tmpString[128]; + memset(tmpString, 0, tListLen(tmpString)); taosFprintfFile(g_fp, "%s no poll more msg when time over, break consume\n", getCurrentTimeString(tmpString)); printf("%s no poll more msg when time over, break consume\n", getCurrentTimeString(tmpString)); int64_t currentPrintTime = taosGetTimestampMs(); @@ -1381,7 +1382,7 @@ void startOmbConsume() { printf("SQL: %s\n", sql); queryDbExec(taos, sql, NO_INSERT_TYPE); - int32_t producerRate = ceil(g_stConfInfo.producerRate / g_stConfInfo.producers); + int32_t producerRate = ceil(((double)g_stConfInfo.producerRate) / g_stConfInfo.producers); printf("==== create %d produce thread ====\n", g_stConfInfo.producers); for (int32_t i = 0; i < g_stConfInfo.producers; ++i) { From d4fdf17cae01e5188d50a61b917d0a112860f3fe Mon Sep 17 00:00:00 2001 From: sunpeng Date: Wed, 22 Feb 2023 17:32:14 +0800 Subject: [PATCH 29/31] build: delete TaosConsumer and TaosTmq from taospy (#20076) (#20091) --- tests/pytest/crash_gen/crash_gen_main.py | 1153 +++++++++++---------- tests/system-test/0-others/user_manage.py | 112 +- 2 files changed, 645 insertions(+), 620 deletions(-) diff --git a/tests/pytest/crash_gen/crash_gen_main.py b/tests/pytest/crash_gen/crash_gen_main.py index 3c39b05e69..ec588659e9 100755 --- a/tests/pytest/crash_gen/crash_gen_main.py +++ b/tests/pytest/crash_gen/crash_gen_main.py @@ -18,7 +18,8 @@ from __future__ import annotations from typing import Any, Set, Tuple from typing import Dict from typing import List -from typing import Optional # Type hinting, ref: https://stackoverflow.com/questions/19202633/python-3-type-hinting-for-none +from typing import \ + Optional # Type hinting, ref: https://stackoverflow.com/questions/19202633/python-3-type-hinting-for-none import textwrap import time @@ -39,7 +40,6 @@ import gc import taos from taos.tmq import * - from .shared.types import TdColumns, TdTags # from crash_gen import ServiceManager, TdeInstance, TdeSubProcess @@ -65,10 +65,11 @@ if sys.version_info[0] < 3: # Command-line/Environment Configurations, will set a bit later # ConfigNameSpace = argparse.Namespace # gConfig: argparse.Namespace -gSvcMgr: Optional[ServiceManager] # TODO: refactor this hack, use dep injection +gSvcMgr: Optional[ServiceManager] # TODO: refactor this hack, use dep injection # logger: logging.Logger gContainer: Container + # def runThread(wt: WorkerThread): # wt.run() @@ -77,7 +78,7 @@ class WorkerThread: def __init__(self, pool: ThreadPool, tid, tc: ThreadCoordinator): """ Note: this runs in the main thread context - """ + """ # self._curStep = -1 self._pool = pool self._tid = tid @@ -91,15 +92,15 @@ class WorkerThread: if (Config.getConfig().per_thread_db_connection): # type: ignore # print("connector_type = {}".format(gConfig.connector_type)) tInst = gContainer.defTdeInstance - if Config.getConfig().connector_type == 'native': - self._dbConn = DbConn.createNative(tInst.getDbTarget()) + if Config.getConfig().connector_type == 'native': + self._dbConn = DbConn.createNative(tInst.getDbTarget()) elif Config.getConfig().connector_type == 'rest': - self._dbConn = DbConn.createRest(tInst.getDbTarget()) + self._dbConn = DbConn.createRest(tInst.getDbTarget()) elif Config.getConfig().connector_type == 'mixed': - if Dice.throw(2) == 0: # 1/2 chance - self._dbConn = DbConn.createNative(tInst.getDbTarget()) + if Dice.throw(2) == 0: # 1/2 chance + self._dbConn = DbConn.createNative(tInst.getDbTarget()) else: - self._dbConn = DbConn.createRest(tInst.getDbTarget()) + self._dbConn = DbConn.createRest(tInst.getDbTarget()) else: raise RuntimeError("Unexpected connector type: {}".format(Config.getConfig().connector_type)) @@ -138,7 +139,7 @@ class WorkerThread: # clean up if (Config.getConfig().per_thread_db_connection): # type: ignore - if self._dbConn.isOpen: #sometimes it is not open + if self._dbConn.isOpen: # sometimes it is not open self._dbConn.close() else: Logging.warning("Cleaning up worker thread, dbConn already closed") @@ -150,20 +151,19 @@ class WorkerThread: tc = self._tc # Thread Coordinator, the overall master try: tc.crossStepBarrier() # shared barrier first, INCLUDING the last one - except threading.BrokenBarrierError as err: # main thread timed out + except threading.BrokenBarrierError as err: # main thread timed out print("_bto", end="") Logging.debug("[TRD] Worker thread exiting due to main thread barrier time-out") break Logging.debug("[TRD] Worker thread [{}] exited barrier...".format(self._tid)) - self.crossStepGate() # then per-thread gate, after being tapped + self.crossStepGate() # then per-thread gate, after being tapped Logging.debug("[TRD] Worker thread [{}] exited step gate...".format(self._tid)) if not self._tc.isRunning(): print("_wts", end="") Logging.debug("[TRD] Thread Coordinator not running any more, worker thread now stopping...") break - # Before we fetch the task and run it, let's ensure we properly "use" the database (not needed any more) try: if (Config.getConfig().per_thread_db_connection): # most likely TRUE @@ -172,7 +172,8 @@ class WorkerThread: # self.useDb() # might encounter exceptions. TODO: catch except taos.error.ProgrammingError as err: errno = Helper.convertErrno(err.errno) - if errno in [0x383, 0x386, 0x00B, 0x014] : # invalid database, dropping, Unable to establish connection, Database not ready + if errno in [0x383, 0x386, 0x00B, + 0x014]: # invalid database, dropping, Unable to establish connection, Database not ready # ignore dummy = 0 else: @@ -180,12 +181,12 @@ class WorkerThread: raise # Fetch a task from the Thread Coordinator - Logging.debug( "[TRD] Worker thread [{}] about to fetch task".format(self._tid)) + Logging.debug("[TRD] Worker thread [{}] about to fetch task".format(self._tid)) task = tc.fetchTask() # Execute such a task Logging.debug("[TRD] Worker thread [{}] about to execute task: {}".format( - self._tid, task.__class__.__name__)) + self._tid, task.__class__.__name__)) task.execute(self) tc.saveExecutedTask(task) Logging.debug("[TRD] Worker thread [{}] finished executing task".format(self._tid)) @@ -228,7 +229,7 @@ class WorkerThread: self._stepGate.set() # wake up! time.sleep(0) # let the released thread run a bit else: - print("_tad", end="") # Thread already dead + print("_tad", end="") # Thread already dead def execSql(self, sql): # TODO: expose DbConn directly return self.getDbConn().execute(sql) @@ -239,7 +240,7 @@ class WorkerThread: def getQueryResult(self): return self.getDbConn().getQueryResult() - def getDbConn(self) -> DbConn : + def getDbConn(self) -> DbConn: if (Config.getConfig().per_thread_db_connection): return self._dbConn else: @@ -251,6 +252,7 @@ class WorkerThread: # else: # return self._tc.getDbState().getDbConn().query(sql) + # The coordinator of all worker threads, mostly running in main thread @@ -262,7 +264,7 @@ class ThreadCoordinator: self._pool = pool # self._wd = wd self._te = None # prepare for every new step - self._dbManager = dbManager # type: Optional[DbManager] # may be freed + self._dbManager = dbManager # type: Optional[DbManager] # may be freed self._executedTasks: List[Task] = [] # in a given step self._lock = threading.RLock() # sync access for a few things @@ -284,7 +286,7 @@ class ThreadCoordinator: return self._dbManager def crossStepBarrier(self, timeout=None): - self._stepBarrier.wait(timeout) + self._stepBarrier.wait(timeout) def requestToStop(self): self._runStatus = Status.STATUS_STOPPING @@ -292,7 +294,7 @@ class ThreadCoordinator: def _runShouldEnd(self, transitionFailed, hasAbortedTask, workerTimeout): maxSteps = Config.getConfig().max_steps # type: ignore - if self._curStep >= (maxSteps - 1): # maxStep==10, last curStep should be 9 + if self._curStep >= (maxSteps - 1): # maxStep==10, last curStep should be 9 return True if self._runStatus != Status.STATUS_RUNNING: return True @@ -304,7 +306,7 @@ class ThreadCoordinator: return True return False - def _hasAbortedTask(self): # from execution of previous step + def _hasAbortedTask(self): # from execution of previous step for task in self._executedTasks: if task.isAborted(): # print("Task aborted: {}".format(task)) @@ -319,17 +321,17 @@ class ThreadCoordinator: "--\r\n\n--> Step {} starts with main thread waking up".format(self._curStep)) # A new TE for the new step - self._te = None # set to empty first, to signal worker thread to stop + self._te = None # set to empty first, to signal worker thread to stop if not transitionFailed: # only if not failed self._te = TaskExecutor(self._curStep) Logging.debug("[TRD] Main thread waking up at step {}, tapping worker threads".format( - self._curStep)) # Now not all threads had time to go to sleep + self._curStep)) # Now not all threads had time to go to sleep # Worker threads will wake up at this point, and each execute it's own task - self.tapAllThreads() # release all worker thread from their "gates" + self.tapAllThreads() # release all worker thread from their "gates" def _syncAtBarrier(self): - # Now main thread (that's us) is ready to enter a step + # Now main thread (that's us) is ready to enter a step # let other threads go past the pool barrier, but wait at the # thread gate Logging.debug("[TRD] Main thread about to cross the barrier") @@ -341,7 +343,7 @@ class ThreadCoordinator: transitionFailed = False try: for x in self._dbs: - db = x # type: Database + db = x # type: Database sm = db.getStateMachine() Logging.debug("[STT] starting transitions for DB: {}".format(db.getName())) # at end of step, transiton the DB state @@ -357,8 +359,8 @@ class ThreadCoordinator: # for t in self._pool.threadList: # Logging.debug("[DB] use db for all worker threads") # t.useDb() - # t.execSql("use db") # main thread executing "use - # db" on behalf of every worker thread + # t.execSql("use db") # main thread executing "use + # db" on behalf of every worker thread except taos.error.ProgrammingError as err: if (err.msg == 'network unavailable'): # broken DB connection @@ -369,12 +371,13 @@ class ThreadCoordinator: self._execStats.registerFailure("Broken DB Connection") # continue # don't do that, need to tap all threads at # end, and maybe signal them to stop - if isinstance(err, CrashGenError): # our own transition failure + if isinstance(err, CrashGenError): # our own transition failure Logging.info("State transition error") # TODO: saw an error here once, let's print out stack info for err? - traceback.print_stack() # Stack frame to here. + traceback.print_stack() # Stack frame to here. Logging.info("Caused by:") - traceback.print_exception(*sys.exc_info()) # Ref: https://www.geeksforgeeks.org/how-to-print-exception-stack-trace-in-python/ + traceback.print_exception( + *sys.exc_info()) # Ref: https://www.geeksforgeeks.org/how-to-print-exception-stack-trace-in-python/ transitionFailed = True self._te = None # Not running any more self._execStats.registerFailure("State transition error: {}".format(err)) @@ -392,14 +395,14 @@ class ThreadCoordinator: # Coordinate all threads step by step self._curStep = -1 # not started yet - + self._execStats.startExec() # start the stop watch transitionFailed = False hasAbortedTask = False workerTimeout = False while not self._runShouldEnd(transitionFailed, hasAbortedTask, workerTimeout): - if not Config.getConfig().debug: # print this only if we are not in debug mode - Progress.emit(Progress.STEP_BOUNDARY) + if not Config.getConfig().debug: # print this only if we are not in debug mode + Progress.emit(Progress.STEP_BOUNDARY) # print(".", end="", flush=True) # if (self._curStep % 2) == 0: # print memory usage once every 10 steps # memUsage = resource.getrusage(resource.RUSAGE_SELF).ru_maxrss @@ -408,15 +411,14 @@ class ThreadCoordinator: # h = hpy() # print("\n") # print(h.heap()) - - + try: - self._syncAtBarrier() # For now just cross the barrier + self._syncAtBarrier() # For now just cross the barrier Progress.emit(Progress.END_THREAD_STEP) - if self._stepStartTime : + if self._stepStartTime: stepExecTime = time.time() - self._stepStartTime Progress.emitStr('{:.3f}s/{}'.format(stepExecTime, DbConnNative.totalRequests)) - DbConnNative.resetTotalRequests() # reset to zero + DbConnNative.resetTotalRequests() # reset to zero except threading.BrokenBarrierError as err: self._execStats.registerFailure("Aborted due to worker thread timeout") Logging.error("\n") @@ -439,15 +441,15 @@ class ThreadCoordinator: # At this point, all threads should be pass the overall "barrier" and before the per-thread "gate" # We use this period to do house keeping work, when all worker # threads are QUIET. - hasAbortedTask = self._hasAbortedTask() # from previous step - if hasAbortedTask: + hasAbortedTask = self._hasAbortedTask() # from previous step + if hasAbortedTask: Logging.info("Aborted task encountered, exiting test program") self._execStats.registerFailure("Aborted Task Encountered") - break # do transition only if tasks are error free + break # do transition only if tasks are error free # Ending previous step try: - transitionFailed = self._doTransition() # To start, we end step -1 first + transitionFailed = self._doTransition() # To start, we end step -1 first except taos.error.ProgrammingError as err: transitionFailed = True errno2 = Helper.convertErrno(err.errno) # correct error scheme @@ -459,32 +461,32 @@ class ThreadCoordinator: # Then we move on to the next step Progress.emit(Progress.BEGIN_THREAD_STEP) self._stepStartTime = time.time() - self._releaseAllWorkerThreads(transitionFailed) + self._releaseAllWorkerThreads(transitionFailed) - if hasAbortedTask or transitionFailed : # abnormal ending, workers waiting at "gate" + if hasAbortedTask or transitionFailed: # abnormal ending, workers waiting at "gate" Logging.debug("Abnormal ending of main thraed") elif workerTimeout: Logging.debug("Abnormal ending of main thread, due to worker timeout") - else: # regular ending, workers waiting at "barrier" + else: # regular ending, workers waiting at "barrier" Logging.debug("Regular ending, main thread waiting for all worker threads to stop...") self._syncAtBarrier() self._te = None # No more executor, time to end Logging.debug("Main thread tapping all threads one last time...") self.tapAllThreads() # Let the threads run one last time - #TODO: looks like we are not capturing the failures for the last step yet (i.e. calling registerFailure if neccessary) + # TODO: looks like we are not capturing the failures for the last step yet (i.e. calling registerFailure if neccessary) Logging.debug("\r\n\n--> Main thread ready to finish up...") Logging.debug("Main thread joining all threads") self._pool.joinAll() # Get all threads to finish - Logging.info(". . . All worker threads finished") # No CR/LF before + Logging.info(". . . All worker threads finished") # No CR/LF before self._execStats.endExec() - def cleanup(self): # free resources + def cleanup(self): # free resources self._pool.cleanup() self._pool = None - self._te = None + self._te = None self._dbManager = None self._executedTasks = [] self._lock = None @@ -492,7 +494,6 @@ class ThreadCoordinator: self._execStats = None self._runStatus = None - def printStats(self): self._execStats.printStats() @@ -523,21 +524,21 @@ class ThreadCoordinator: def _initDbs(self): ''' Initialize multiple databases, invoked at __ini__() time ''' - self._dbs = [] # type: List[Database] + self._dbs = [] # type: List[Database] dbc = self.getDbManager().getDbConn() if Config.getConfig().max_dbs == 0: self._dbs.append(Database(0, dbc)) - else: - baseDbNumber = int(datetime.datetime.now().timestamp( # Don't use Dice/random, as they are deterministic - )*333) % 888 if Config.getConfig().dynamic_db_table_names else 0 + else: + baseDbNumber = int(datetime.datetime.now().timestamp( # Don't use Dice/random, as they are deterministic + ) * 333) % 888 if Config.getConfig().dynamic_db_table_names else 0 for i in range(Config.getConfig().max_dbs): self._dbs.append(Database(baseDbNumber + i, dbc)) def pickDatabase(self): idxDb = 0 - if Config.getConfig().max_dbs != 0 : - idxDb = Dice.throw(Config.getConfig().max_dbs) # 0 to N-1 - db = self._dbs[idxDb] # type: Database + if Config.getConfig().max_dbs != 0: + idxDb = Dice.throw(Config.getConfig().max_dbs) # 0 to N-1 + db = self._dbs[idxDb] # type: Database return db def fetchTask(self) -> Task: @@ -549,12 +550,12 @@ class ThreadCoordinator: # pick a task type for current state db = self.pickDatabase() - if Dice.throw(2)==1: - taskType = db.getStateMachine().pickTaskType() # dynamic name of class + if Dice.throw(2) == 1: + taskType = db.getStateMachine().pickTaskType() # dynamic name of class else: - taskType = db.getStateMachine().balance_pickTaskType() # and an method can get balance task types + taskType = db.getStateMachine().balance_pickTaskType() # and an method can get balance task types pass - + return taskType(self._execStats, db) # create a task from it def resetExecutedTasks(self): @@ -564,6 +565,7 @@ class ThreadCoordinator: with self._lock: self._executedTasks.append(task) + class ThreadPool: def __init__(self, numThreads, maxSteps): self.numThreads = numThreads @@ -585,7 +587,8 @@ class ThreadPool: workerThread._thread.join() def cleanup(self): - self.threadList = [] # maybe clean up each? + self.threadList = [] # maybe clean up each? + # A queue of continguous POSITIVE integers, used by DbManager to generate continuous numbers # for new table names @@ -680,11 +683,11 @@ class AnyState: CAN_CREATE_DB = 1 # For below, if we can "drop the DB", but strictly speaking # only "under normal circumstances", as we may override it with the -b option - CAN_DROP_DB = 2 + CAN_DROP_DB = 2 CAN_CREATE_FIXED_SUPER_TABLE = 3 CAN_CREATE_STREAM = 3 # super table must exists CAN_CREATE_TOPIC = 3 # super table must exists - CAN_CREATE_CONSUMERS = 3 + CAN_CREATE_CONSUMERS = 3 CAN_DROP_FIXED_SUPER_TABLE = 4 CAN_DROP_TOPIC = 4 CAN_DROP_STREAM = 4 @@ -729,7 +732,7 @@ class AnyState: def canDropDb(self): # If user requests to run up to a number of DBs, # we'd then not do drop_db operations any more - if Config.getConfig().max_dbs > 0 or Config.getConfig().use_shadow_db : + if Config.getConfig().max_dbs > 0 or Config.getConfig().use_shadow_db: return False return self._info[self.CAN_DROP_DB] @@ -737,19 +740,19 @@ class AnyState: return self._info[self.CAN_CREATE_FIXED_SUPER_TABLE] def canDropFixedSuperTable(self): - if Config.getConfig().use_shadow_db: # duplicate writes to shaddow DB, in which case let's disable dropping s-table + if Config.getConfig().use_shadow_db: # duplicate writes to shaddow DB, in which case let's disable dropping s-table return False return self._info[self.CAN_DROP_FIXED_SUPER_TABLE] def canCreateTopic(self): return self._info[self.CAN_CREATE_TOPIC] - + def canDropTopic(self): return self._info[self.CAN_DROP_TOPIC] def canCreateConsumers(self): return self._info[self.CAN_CREATE_CONSUMERS] - + def canCreateStreams(self): return self._info[self.CAN_CREATE_STREAM] @@ -777,7 +780,7 @@ class AnyState: raise CrashGenError( "Unexpected more than 1 success at state: {}, with task: {}, in task set: {}".format( self.__class__.__name__, - cls.__name__, # verified just now that isinstance(task, cls) + cls.__name__, # verified just now that isinstance(task, cls) [c.__class__.__name__ for c in tasks] )) @@ -792,16 +795,17 @@ class AnyState: sCnt += 1 if (exists and sCnt <= 0): raise CrashGenError("Unexpected zero success at state: {}, with task: {}, in task set: {}".format( - self.__class__.__name__, - cls.__name__, # verified just now that isinstance(task, cls) - [c.__class__.__name__ for c in tasks] - )) + self.__class__.__name__, + cls.__name__, # verified just now that isinstance(task, cls) + [c.__class__.__name__ for c in tasks] + )) def assertNoTask(self, tasks, cls): for task in tasks: if isinstance(task, cls): raise CrashGenError( - "This task: {}, is not expected to be present, given the success/failure of others".format(cls.__name__)) + "This task: {}, is not expected to be present, given the success/failure of others".format( + cls.__name__)) def assertNoSuccess(self, tasks, cls): for task in tasks: @@ -848,7 +852,7 @@ class StateEmpty(AnyState): def verifyTasksToState(self, tasks, newState): if (self.hasSuccess(tasks, TaskCreateDb) - ): # at EMPTY, if there's succes in creating DB + ): # at EMPTY, if there's succes in creating DB if (not self.hasTask(tasks, TaskDropDb)): # and no drop_db tasks # we must have at most one. TODO: compare numbers self.assertAtMostOneSuccess(tasks, TaskCreateDb) @@ -885,19 +889,19 @@ class StateSuperTableOnly(AnyState): def verifyTasksToState(self, tasks, newState): if (self.hasSuccess(tasks, TaskDropSuperTable) - ): # we are able to drop the table - #self.assertAtMostOneSuccess(tasks, TaskDropSuperTable) + ): # we are able to drop the table + # self.assertAtMostOneSuccess(tasks, TaskDropSuperTable) # we must have had recreted it self.hasSuccess(tasks, TaskCreateSuperTable) # self._state = self.STATE_DB_ONLY # elif ( self.hasSuccess(tasks, AddFixedDataTask) ): # no success dropping the table, but added data # self.assertNoTask(tasks, DropFixedTableTask) # not true in massively parrallel cases - # self._state = self.STATE_HAS_DATA + # self._state = self.STATE_HAS_DATA # elif ( self.hasSuccess(tasks, ReadFixedDataTask) ): # no success in prev cases, but was able to read data - # self.assertNoTask(tasks, DropFixedTableTask) - # self.assertNoTask(tasks, AddFixedDataTask) - # self._state = self.STATE_TABLE_ONLY # no change + # self.assertNoTask(tasks, DropFixedTableTask) + # self.assertNoTask(tasks, AddFixedDataTask) + # self._state = self.STATE_TABLE_ONLY # no change # else: # did not drop table, did not insert data, did not read successfully, that is impossible # raise RuntimeError("Unexpected no-success scenarios") # TODO: need to revamp!! @@ -919,41 +923,41 @@ class StateHasData(AnyState): self.assertAtMostOneSuccess(tasks, TaskDropDb) # TODO: dicy elif (newState.equals(AnyState.STATE_DB_ONLY)): # in DB only if (not self.hasTask(tasks, TaskCreateDb) - ): # without a create_db task + ): # without a create_db task # we must have drop_db task self.assertNoTask(tasks, TaskDropDb) self.hasSuccess(tasks, TaskDropSuperTable) # self.assertAtMostOneSuccess(tasks, DropFixedSuperTableTask) # TODO: dicy # elif ( newState.equals(AnyState.STATE_TABLE_ONLY) ): # data deleted - # self.assertNoTask(tasks, TaskDropDb) - # self.assertNoTask(tasks, TaskDropSuperTable) - # self.assertNoTask(tasks, TaskAddData) - # self.hasSuccess(tasks, DeleteDataTasks) + # self.assertNoTask(tasks, TaskDropDb) + # self.assertNoTask(tasks, TaskDropSuperTable) + # self.assertNoTask(tasks, TaskAddData) + # self.hasSuccess(tasks, DeleteDataTasks) else: # should be STATE_HAS_DATA if (not self.hasTask(tasks, TaskCreateDb) - ): # only if we didn't create one + ): # only if we didn't create one # we shouldn't have dropped it self.assertNoTask(tasks, TaskDropDb) - if not( self.hasTask(tasks, TaskCreateSuperTable) - ): # if we didn't create the table + if not (self.hasTask(tasks, TaskCreateSuperTable) + ): # if we didn't create the table # we should not have a task that drops it self.assertNoTask(tasks, TaskDropSuperTable) # self.assertIfExistThenSuccess(tasks, ReadFixedDataTask) class StateMechine: - def __init__(self, db: Database): + def __init__(self, db: Database): self._db = db # transitition target probabilities, indexed with value of STATE_EMPTY, STATE_DB_ONLY, etc. self._stateWeights = [1, 2, 10, 40] - def init(self, dbc: DbConn): # late initailization, don't save the dbConn + def init(self, dbc: DbConn): # late initailization, don't save the dbConn try: self._curState = self._findCurrentState(dbc) # starting state - except taos.error.ProgrammingError as err: + except taos.error.ProgrammingError as err: Logging.error("Failed to initialized state machine, cannot find current state: {}".format(err)) traceback.print_stack() - raise # re-throw + raise # re-throw # TODO: seems no lnoger used, remove? def getCurrentState(self): @@ -999,28 +1003,27 @@ class StateMechine: def _findCurrentState(self, dbc: DbConn): ts = time.time() # we use this to debug how fast/slow it is to do the various queries to find the current DB state - dbName =self._db.getName() - if not dbc.existsDatabase(dbName): # dbc.hasDatabases(): # no database?! - Logging.debug( "[STT] empty database found, between {} and {}".format(ts, time.time())) + dbName = self._db.getName() + if not dbc.existsDatabase(dbName): # dbc.hasDatabases(): # no database?! + Logging.debug("[STT] empty database found, between {} and {}".format(ts, time.time())) return StateEmpty() # did not do this when openning connection, and this is NOT the worker # thread, which does this on their own dbc.use(dbName) - + if not dbc.hasTables(): # no tables - + Logging.debug("[STT] DB_ONLY found, between {} and {}".format(ts, time.time())) return StateDbOnly() # For sure we have tables, which means we must have the super table. # TODO: are we sure? - + sTable = self._db.getFixedSuperTable() - if sTable.hasRegTables(dbc): # no regular tables # print("debug=====*\n"*100) Logging.debug("[STT] SUPER_TABLE_ONLY found, between {} and {}".format(ts, time.time())) - + return StateSuperTableOnly() else: # has actual tables Logging.debug("[STT] HAS_DATA found, between {} and {}".format(ts, time.time())) @@ -1029,7 +1032,7 @@ class StateMechine: # We transition the system to a new state by examining the current state itself def transition(self, tasks, dbc: DbConn): global gSvcMgr - + if (len(tasks) == 0): # before 1st step, or otherwise empty Logging.debug("[STT] Starting State: {}".format(self._curState)) return # do nothing @@ -1038,39 +1041,39 @@ class StateMechine: dbc.execute("select * from information_schema.ins_dnodes") # Generic Checks, first based on the start state - if not Config.getConfig().ignore_errors: # verify state, only if we are asked not to ignore certain errors. + if not Config.getConfig().ignore_errors: # verify state, only if we are asked not to ignore certain errors. if self._curState.canCreateDb(): self._curState.assertIfExistThenSuccess(tasks, TaskCreateDb) # self.assertAtMostOneSuccess(tasks, CreateDbTask) # not really, in # case of multiple creation and drops if self._curState.canDropDb(): - if gSvcMgr == None: # only if we are running as client-only + if gSvcMgr == None: # only if we are running as client-only self._curState.assertIfExistThenSuccess(tasks, TaskDropDb) # self.assertAtMostOneSuccess(tasks, DropDbTask) # not really in # case of drop-create-drop # if self._state.canCreateFixedTable(): - # self.assertIfExistThenSuccess(tasks, CreateFixedTableTask) # Not true, DB may be dropped - # self.assertAtMostOneSuccess(tasks, CreateFixedTableTask) # not - # really, in case of create-drop-create + # self.assertIfExistThenSuccess(tasks, CreateFixedTableTask) # Not true, DB may be dropped + # self.assertAtMostOneSuccess(tasks, CreateFixedTableTask) # not + # really, in case of create-drop-create # if self._state.canDropFixedTable(): - # self.assertIfExistThenSuccess(tasks, DropFixedTableTask) # Not True, the whole DB may be dropped - # self.assertAtMostOneSuccess(tasks, DropFixedTableTask) # not - # really in case of drop-create-drop + # self.assertIfExistThenSuccess(tasks, DropFixedTableTask) # Not True, the whole DB may be dropped + # self.assertAtMostOneSuccess(tasks, DropFixedTableTask) # not + # really in case of drop-create-drop # if self._state.canAddData(): # self.assertIfExistThenSuccess(tasks, AddFixedDataTask) # not true # actually # if self._state.canReadData(): - # Nothing for sure + # Nothing for sure newState = self._findCurrentState(dbc) Logging.debug("[STT] New DB state determined: {}".format(newState)) # can old state move to new state through the tasks? - if not Config.getConfig().ignore_errors: # verify state, only if we are asked not to ignore certain errors. + if not Config.getConfig().ignore_errors: # verify state, only if we are asked not to ignore certain errors. self._curState.verifyTasksToState(tasks, newState) self._curState = newState @@ -1096,22 +1099,24 @@ class StateMechine: weightsTypes = BasicTypes.copy() # this matrixs can balance the Frequency of TaskTypes - balance_TaskType_matrixs = {'TaskDropDb': 5 , 'TaskDropTopics': 20 , 'TaskDropStreams':10 , 'TaskDropStreamTables':10 , - 'TaskReadData':50 , 'TaskDropSuperTable':5 , 'TaskAlterTags':3 , 'TaskAddData':10, - 'TaskDeleteData':10 , 'TaskCreateDb':10 , 'TaskCreateStream': 3, 'TaskCreateTopic' :3, - 'TaskCreateConsumers':10, 'TaskCreateSuperTable': 10 } # TaskType : balance_matrixs of task - - for task , weights in balance_TaskType_matrixs.items(): - + balance_TaskType_matrixs = {'TaskDropDb': 5, 'TaskDropTopics': 20, 'TaskDropStreams': 10, + 'TaskDropStreamTables': 10, + 'TaskReadData': 50, 'TaskDropSuperTable': 5, 'TaskAlterTags': 3, 'TaskAddData': 10, + 'TaskDeleteData': 10, 'TaskCreateDb': 10, 'TaskCreateStream': 3, + 'TaskCreateTopic': 3, + 'TaskCreateConsumers': 10, + 'TaskCreateSuperTable': 10} # TaskType : balance_matrixs of task + + for task, weights in balance_TaskType_matrixs.items(): + for basicType in BasicTypes: if basicType.__name__ == task: for _ in range(weights): weightsTypes.append(basicType) - task = random.sample(weightsTypes,1) + task = random.sample(weightsTypes, 1) return task[0] - # ref: # https://eli.thegreenplace.net/2010/01/22/weighted-random-generation-in-python/ def _weighted_choice_sub(self, weights) -> int: @@ -1123,6 +1128,7 @@ class StateMechine: return i raise CrashGenError("Unexpected no choice") + class Database: ''' We use this to represent an actual TDengine database inside a service instance, possibly in a cluster environment. @@ -1131,16 +1137,16 @@ class Database: TODO: consider moving, but keep in mind it contains "StateMachine" ''' - _clsLock = threading.Lock() # class wide lock + _clsLock = threading.Lock() # class wide lock _lastInt = 101 # next one is initial integer - _lastTick = None # Optional[datetime] - _lastLaggingTick = None # Optional[datetime] # lagging tick, for out-of-sequence (oos) data insertions + _lastTick = None # Optional[datetime] + _lastLaggingTick = None # Optional[datetime] # lagging tick, for out-of-sequence (oos) data insertions - def __init__(self, dbNum: int, dbc: DbConn): # TODO: remove dbc - self._dbNum = dbNum # we assign a number to databases, for our testing purpose + def __init__(self, dbNum: int, dbc: DbConn): # TODO: remove dbc + self._dbNum = dbNum # we assign a number to databases, for our testing purpose self._stateMachine = StateMechine(self) self._stateMachine.init(dbc) - + self._lock = threading.RLock() def getStateMachine(self) -> StateMechine: @@ -1152,7 +1158,7 @@ class Database: def getName(self): return "db_{}".format(self._dbNum) - def filterTasks(self, inTasks: List[Task]): # Pick out those belonging to us + def filterTasks(self, inTasks: List[Task]): # Pick out those belonging to us outTasks = [] for task in inTasks: if task.getDb().isSame(self): @@ -1184,38 +1190,42 @@ class Database: # start time will be auto generated , start at 10 years ago local time local_time = datetime.datetime.now().strftime('%Y-%m-%d %H:%M:%S.%f')[:-16] local_epoch_time = [int(i) for i in local_time.split("-")] - #local_epoch_time will be such as : [2022, 7, 18] + # local_epoch_time will be such as : [2022, 7, 18] - t1 = datetime.datetime(local_epoch_time[0]-5, local_epoch_time[1], local_epoch_time[2]) + t1 = datetime.datetime(local_epoch_time[0] - 5, local_epoch_time[1], local_epoch_time[2]) t2 = datetime.datetime.now() # maybe a very large number, takes 69 years to exceed Python int range elSec = int(t2.timestamp() - t1.timestamp()) elSec2 = (elSec % (8 * 12 * 30 * 24 * 60 * 60 / 500)) * \ - 500 # a number representing seconds within 10 years + 500 # a number representing seconds within 10 years # print("elSec = {}".format(elSec)) - t3 = datetime.datetime(local_epoch_time[0]-10, local_epoch_time[1], local_epoch_time[2]) # default "keep" is 10 years + t3 = datetime.datetime(local_epoch_time[0] - 10, local_epoch_time[1], + local_epoch_time[2]) # default "keep" is 10 years t4 = datetime.datetime.fromtimestamp( t3.timestamp() + elSec2) # see explanation above Logging.debug("Setting up TICKS to start from: {}".format(t4)) return t4 @classmethod - def getNextTick(cls): + def getNextTick(cls): ''' Fetch a timestamp tick, with some random factor, may not be unique. - ''' + ''' with cls._clsLock: # prevent duplicate tick - if cls._lastLaggingTick is None or cls._lastTick is None : # not initialized + if cls._lastLaggingTick is None or cls._lastTick is None: # not initialized # 10k at 1/20 chance, should be enough to avoid overlaps tick = cls.setupLastTick() cls._lastTick = tick - cls._lastLaggingTick = tick + datetime.timedelta(0, -60*2) # lagging behind 2 minutes, should catch up fast + cls._lastLaggingTick = tick + datetime.timedelta(0, + -60 * 2) # lagging behind 2 minutes, should catch up fast # if : # should be quite a bit into the future - if Config.isSet('mix_oos_data') and Dice.throw(20) == 0: # if asked to do so, and 1 in 20 chance, return lagging tick - cls._lastLaggingTick += datetime.timedelta(0, 1) # pick the next sequence from the lagging tick sequence - return cls._lastLaggingTick + if Config.isSet('mix_oos_data') and Dice.throw( + 20) == 0: # if asked to do so, and 1 in 20 chance, return lagging tick + cls._lastLaggingTick += datetime.timedelta(0, + 1) # pick the next sequence from the lagging tick sequence + return cls._lastLaggingTick else: # regular # add one second to it cls._lastTick += datetime.timedelta(0, 1) @@ -1332,9 +1342,7 @@ class Task(): # Logging.debug("Creating new task {}...".format(self._taskNum)) self._execStats = execStats - self._db = db # A task is always associated/for a specific DB - - + self._db = db # A task is always associated/for a specific DB def isSuccess(self): return self._err is None @@ -1367,82 +1375,78 @@ class Task(): def _isServiceStable(self): if not gSvcMgr: return True # we don't run service, so let's assume it's stable - return gSvcMgr.isStable() # otherwise let's examine the service + return gSvcMgr.isStable() # otherwise let's examine the service def _isErrAcceptable(self, errno, msg): if errno in [ - # TDengine 2.x Error Codes: - 0x05, # TSDB_CODE_RPC_NOT_READY - 0x0B, # Unable to establish connection, more details in TD-1648 - # 0x200, # invalid SQL， TODO: re-examine with TD-934 - 0x20F, # query terminated, possibly due to vnoding being dropped, see TD-1776 - 0x213, # "Disconnected from service", result of "kill connection ???" - 0x217, # "db not selected", client side defined error code - # 0x218, # "Table does not exist" client side defined error code - 0x360, # Table already exists - 0x362, - # 0x369, # tag already exists - 0x36A, 0x36B, 0x36D, - 0x381, - 0x380, # "db not selected" - 0x383, - 0x386, # DB is being dropped?! - 0x503, - 0x510, # vnode not in ready state - 0x14, # db not ready, errno changed - 0x600, # Invalid table ID, why? - 0x218, # Table does not exist + # TDengine 2.x Error Codes: + 0x05, # TSDB_CODE_RPC_NOT_READY + 0x0B, # Unable to establish connection, more details in TD-1648 + # 0x200, # invalid SQL， TODO: re-examine with TD-934 + 0x20F, # query terminated, possibly due to vnoding being dropped, see TD-1776 + 0x213, # "Disconnected from service", result of "kill connection ???" + 0x217, # "db not selected", client side defined error code + # 0x218, # "Table does not exist" client side defined error code + 0x360, # Table already exists + 0x362, + # 0x369, # tag already exists + 0x36A, 0x36B, 0x36D, + 0x381, + 0x380, # "db not selected" + 0x383, + 0x386, # DB is being dropped?! + 0x503, + 0x510, # vnode not in ready state + 0x14, # db not ready, errno changed + 0x600, # Invalid table ID, why? + 0x218, # Table does not exist - # TDengine 3.0 Error Codes: - 0x0333, # Object is creating # TODO: this really is NOT an acceptable error - 0x0369, # Tag already exists - 0x0388, # Database not exist - 0x03A0, # STable already exists - 0x03A1, # STable [does] not exist - 0x03AA, # Tag already exists - 0x0603, # Table already exists - 0x2603, # Table does not exist, replaced by 2662 below - 0x260d, # Tags number not matched - 0x2662, # Table does not exist #TODO: what about 2603 above? - 0x2600, # database not specified, SQL: show stables , database droped , and show tables - 0x032C, # Object is creating - 0x032D, # Object is dropping - 0x03D3, # Conflict transaction not completed - 0x0707, # Query not ready , it always occur at replica 3 - 0x707, # Query not ready - 0x396, # Database in creating status - 0x386, # Database in droping status - 0x03E1, # failed on tmq_subscribe ,topic not exist - 0x03ed , # Topic must be dropped first, SQL: drop database db_0 - 0x0203 , # Invalid value - 0x03f0 , # Stream already exist , topic already exists + # TDengine 3.0 Error Codes: + 0x0333, # Object is creating # TODO: this really is NOT an acceptable error + 0x0369, # Tag already exists + 0x0388, # Database not exist + 0x03A0, # STable already exists + 0x03A1, # STable [does] not exist + 0x03AA, # Tag already exists + 0x0603, # Table already exists + 0x2603, # Table does not exist, replaced by 2662 below + 0x260d, # Tags number not matched + 0x2662, # Table does not exist #TODO: what about 2603 above? + 0x2600, # database not specified, SQL: show stables , database droped , and show tables + 0x032C, # Object is creating + 0x032D, # Object is dropping + 0x03D3, # Conflict transaction not completed + 0x0707, # Query not ready , it always occur at replica 3 + 0x707, # Query not ready + 0x396, # Database in creating status + 0x386, # Database in droping status + 0x03E1, # failed on tmq_subscribe ,topic not exist + 0x03ed, # Topic must be dropped first, SQL: drop database db_0 + 0x0203, # Invalid value + 0x03f0, # Stream already exist , topic already exists - - - - 1000 # REST catch-all error - ]: - return True # These are the ALWAYS-ACCEPTABLE ones + 1000 # REST catch-all error + ]: + return True # These are the ALWAYS-ACCEPTABLE ones # This case handled below already. # elif (errno in [ 0x0B ]) and Settings.getConfig().auto_start_service: # return True # We may get "network unavilable" when restarting service - elif Config.getConfig().ignore_errors: # something is specified on command line + elif Config.getConfig().ignore_errors: # something is specified on command line moreErrnos = [int(v, 0) for v in Config.getConfig().ignore_errors.split(',')] if errno in moreErrnos: return True - elif errno == 0x200 : # invalid SQL, we need to div in a bit more + elif errno == 0x200: # invalid SQL, we need to div in a bit more if msg.find("invalid column name") != -1: - return True - elif msg.find("tags number not matched") != -1: # mismatched tags after modification return True - elif msg.find("duplicated column names") != -1: # also alter table tag issues + elif msg.find("tags number not matched") != -1: # mismatched tags after modification return True - elif not self._isServiceStable(): # We are managing service, and ... + elif msg.find("duplicated column names") != -1: # also alter table tag issues + return True + elif not self._isServiceStable(): # We are managing service, and ... Logging.info("Ignoring error when service starting/stopping: errno = {}, msg = {}".format(errno, msg)) return True - - return False # Not an acceptable error + return False # Not an acceptable error def execute(self, wt: WorkerThread): wt.verifyThreadSelf() @@ -1453,7 +1457,7 @@ class Task(): self.logDebug( "[-] executing task {}...".format(self.__class__.__name__)) - self._err = None # TODO: type hint mess up? + self._err = None # TODO: type hint mess up? self._execStats.beginTaskType(self.__class__.__name__) # mark beginning errno2 = None @@ -1465,19 +1469,19 @@ class Task(): errno2 = Helper.convertErrno(err.errno) if (Config.getConfig().continue_on_exception): # user choose to continue self.logDebug("[=] Continue after TAOS exception: errno=0x{:X}, msg: {}, SQL: {}".format( - errno2, err, wt.getDbConn().getLastSql())) + errno2, err, wt.getDbConn().getLastSql())) self._err = err elif self._isErrAcceptable(errno2, err.__str__()): self.logDebug("[=] Acceptable Taos library exception: errno=0x{:X}, msg: {}, SQL: {}".format( - errno2, err, wt.getDbConn().getLastSql())) + errno2, err, wt.getDbConn().getLastSql())) # print("_", end="", flush=True) Progress.emit(Progress.ACCEPTABLE_ERROR) self._err = err - else: # not an acceptable error + else: # not an acceptable error shortTid = threading.get_ident() % 10000 errMsg = "[=] Unexpected Taos library exception ({}): errno=0x{:X}, thread={}, msg: {}, SQL: {}".format( self.__class__.__name__, - errno2, + errno2, shortTid, err, wt.getDbConn().getLastSql()) self.logDebug(errMsg) @@ -1485,7 +1489,8 @@ class Task(): # raise # so that we see full stack traceback.print_exc() print( - "\n\n----------------------------\nProgram ABORTED Due to Unexpected TAOS Error: \n\n{}\n".format(errMsg) + + "\n\n----------------------------\nProgram ABORTED Due to Unexpected TAOS Error: \n\n{}\n".format( + errMsg) + "----------------------------\n") # sys.exit(-1) self._err = err @@ -1502,10 +1507,10 @@ class Task(): traceback.print_exc() # except BaseException: # TODO: what is this again??!! # raise RuntimeError("Punt") - # self.logDebug( - # "[=] Unexpected exception, SQL: {}".format( - # wt.getDbConn().getLastSql())) - # raise + # self.logDebug( + # "[=] Unexpected exception, SQL: {}".format( + # wt.getDbConn().getLastSql())) + # raise self._execStats.endTaskType(self.__class__.__name__, self.isSuccess()) self.logDebug("[X] task execution completed, {}, status: {}".format( @@ -1524,12 +1529,12 @@ class Task(): def getQueryResult(self, wt: WorkerThread): # execute an SQL on the worker thread return wt.getQueryResult() - def lockTable(self, ftName): # full table name + def lockTable(self, ftName): # full table name # print(" <<" + ftName + '_', end="", flush=True) - with Task._lock: # SHORT lock! so we only protect lock creation - if not ftName in Task._tableLocks: # Create new lock and add to list, if needed + with Task._lock: # SHORT lock! so we only protect lock creation + if not ftName in Task._tableLocks: # Create new lock and add to list, if needed Task._tableLocks[ftName] = threading.Lock() - + # No lock protection, anybody can do this any time lock = Task._tableLocks[ftName] # Logging.info("Acquiring lock: {}, {}".format(ftName, lock)) @@ -1538,7 +1543,7 @@ class Task(): def unlockTable(self, ftName): # print('_' + ftName + ">> ", end="", flush=True) - with Task._lock: + with Task._lock: if not ftName in self._tableLocks: raise RuntimeError("Corrupt state, no such lock") lock = Task._tableLocks[ftName] @@ -1588,11 +1593,11 @@ class ExecutionStats: t[0] += 1 # index 0 has the "total" execution times if isSuccess: t[1] += 1 # index 1 has the "success" execution times - if eno != None: + if eno != None: if klassName not in self._errors: self._errors[klassName] = {} errors = self._errors[klassName] - errors[eno] = errors[eno]+1 if eno in errors else 1 + errors[eno] = errors[eno] + 1 if eno in errors else 1 def beginTaskType(self, klassName): with self._lock: @@ -1615,7 +1620,7 @@ class ExecutionStats: Logging.info( "----------------------------------------------------------------------") Logging.info( - "| Crash_Gen test {}, with the following stats:". format( + "| Crash_Gen test {}, with the following stats:".format( "FAILED (reason: {})".format( self._failureReason) if self._failed else "SUCCEEDED")) Logging.info("| Task Execution Times (success/total):") @@ -1628,7 +1633,7 @@ class ExecutionStats: # print("errors = {}".format(errors)) errStrs = ["0x{:X}:{}".format(eno, n) for (eno, n) in errors.items()] # print("error strings = {}".format(errStrs)) - errStr = ", ".join(errStrs) + errStr = ", ".join(errStrs) Logging.info("| {0:<24}: {1}/{2} (Errors: {3})".format(k, n[1], n[0], errStr)) Logging.info( @@ -1647,8 +1652,8 @@ class ExecutionStats: Logging.info("| Top numbers written: {}".format(TaskExecutor.getBoundedList())) Logging.info("| Active DB Native Connections (now): {}".format(DbConnNative.totalConnections)) Logging.info("| Longest native query time: {:.3f} seconds, started: {}". - format(MyTDSql.longestQueryTime, - time.strftime("%x %X", time.localtime(MyTDSql.lqStartTime))) ) + format(MyTDSql.longestQueryTime, + time.strftime("%x %X", time.localtime(MyTDSql.lqStartTime)))) Logging.info("| Longest native query: {}".format(MyTDSql.longestQuery)) Logging.info( "----------------------------------------------------------------------") @@ -1662,12 +1667,12 @@ class StateTransitionTask(Task): _baseTableNumber = None - _endState = None # TODO: no longter used? + _endState = None # TODO: no longter used? @classmethod def getInfo(cls): # each sub class should supply their own information raise RuntimeError("Overriding method expected") - + @classmethod def getEndState(cls): # TODO: optimize by calling it fewer times raise RuntimeError("Overriding method expected") @@ -1687,7 +1692,7 @@ class StateTransitionTask(Task): @classmethod def getRegTableName(cls, i): - if ( StateTransitionTask._baseTableNumber is None): # Set it one time + if (StateTransitionTask._baseTableNumber is None): # Set it one time StateTransitionTask._baseTableNumber = Dice.throw( 999) if Config.getConfig().dynamic_db_table_names else 0 return "reg_table_{}".format(StateTransitionTask._baseTableNumber + i) @@ -1711,16 +1716,21 @@ class TaskCreateDb(StateTransitionTask): repStr = "" if Config.getConfig().num_replicas != 1: # numReplica = Dice.throw(Settings.getConfig().max_replicas) + 1 # 1,2 ... N - numReplica = Config.getConfig().num_replicas # fixed, always + numReplica = Config.getConfig().num_replicas # fixed, always repStr = "replica {}".format(numReplica) - updatePostfix = "" if Config.getConfig().verify_data else "" # allow update only when "verify data" is active , 3.0 version default is update 1 - vg_nums = random.randint(1,8) - cache_model = Dice.choice(['none' , 'last_row' , 'last_value' , 'both']) - buffer = random.randint(3,128) + updatePostfix = "" if Config.getConfig().verify_data else "" # allow update only when "verify data" is active , 3.0 version default is update 1 + vg_nums = random.randint(1, 8) + cache_model = Dice.choice(['none', 'last_row', 'last_value', 'both']) + buffer = random.randint(3, 128) dbName = self._db.getName() - self.execWtSql(wt, "create database {} {} {} vgroups {} cachemodel '{}' buffer {} ".format(dbName, repStr, updatePostfix, vg_nums, cache_model,buffer ) ) + self.execWtSql(wt, "create database {} {} {} vgroups {} cachemodel '{}' buffer {} ".format(dbName, repStr, + updatePostfix, + vg_nums, + cache_model, + buffer)) if dbName == "db_0" and Config.getConfig().use_shadow_db: - self.execWtSql(wt, "create database {} {} {} ".format("db_s", repStr, updatePostfix ) ) + self.execWtSql(wt, "create database {} {} {} ".format("db_s", repStr, updatePostfix)) + class TaskDropDb(StateTransitionTask): @classmethod @@ -1732,19 +1742,20 @@ class TaskDropDb(StateTransitionTask): return state.canDropDb() def _executeInternal(self, te: TaskExecutor, wt: WorkerThread): - + try: - self.queryWtSql(wt, "drop database {}".format(self._db.getName())) # drop database maybe failed ,because topic exists + self.queryWtSql(wt, "drop database {}".format( + self._db.getName())) # drop database maybe failed ,because topic exists except taos.error.ProgrammingError as err: errno = Helper.convertErrno(err.errno) - if errno in [0x0203]: # drop maybe failed + if errno in [0x0203]: # drop maybe failed pass Logging.debug("[OPS] database dropped at {}".format(time.time())) class TaskCreateStream(StateTransitionTask): - + @classmethod def getEndState(cls): return StateHasData() @@ -1755,39 +1766,40 @@ class TaskCreateStream(StateTransitionTask): def _executeInternal(self, te: TaskExecutor, wt: WorkerThread): dbname = self._db.getName() - - sub_stream_name = dbname+ '_sub_stream' + + sub_stream_name = dbname + '_sub_stream' sub_stream_tb_name = 'stream_tb_sub' - super_stream_name = dbname+ '_super_stream' + super_stream_name = dbname + '_super_stream' super_stream_tb_name = 'stream_tb_super' if not self._db.exists(wt.getDbConn()): Logging.debug("Skipping task, no DB yet") return - sTable = self._db.getFixedSuperTable() # type: TdSuperTable + sTable = self._db.getFixedSuperTable() # type: TdSuperTable # wt.execSql("use db") # should always be in place - stbname =sTable.getName() + stbname = sTable.getName() sub_tables = sTable.getRegTables(wt.getDbConn()) - aggExpr = Dice.choice([ - 'count(*)', 'avg(speed)', 'sum(speed)', 'stddev(speed)','min(speed)', 'max(speed)', 'first(speed)', 'last(speed)', - 'apercentile(speed, 10)', 'last_row(*)', 'twa(speed)']) - - stream_sql = '' # set default value + aggExpr = Dice.choice([ + 'count(*)', 'avg(speed)', 'sum(speed)', 'stddev(speed)', 'min(speed)', 'max(speed)', 'first(speed)', + 'last(speed)', + 'apercentile(speed, 10)', 'last_row(*)', 'twa(speed)']) + + stream_sql = '' # set default value if sub_tables: sub_tbname = sub_tables[0] # create stream with query above sub_table - stream_sql = 'create stream {} into {}.{} as select {}, avg(speed) FROM {}.{} PARTITION BY tbname INTERVAL(5s) SLIDING(3s) '.\ - format(sub_stream_name,dbname,sub_stream_tb_name ,aggExpr,dbname,sub_tbname) + stream_sql = 'create stream {} into {}.{} as select {}, avg(speed) FROM {}.{} PARTITION BY tbname INTERVAL(5s) SLIDING(3s) '. \ + format(sub_stream_name, dbname, sub_stream_tb_name, aggExpr, dbname, sub_tbname) else: - stream_sql = 'create stream {} into {}.{} as select {}, avg(speed) FROM {}.{} PARTITION BY tbname INTERVAL(5s) SLIDING(3s) '.\ - format(super_stream_name,dbname,super_stream_tb_name,aggExpr, dbname,stbname) + stream_sql = 'create stream {} into {}.{} as select {}, avg(speed) FROM {}.{} PARTITION BY tbname INTERVAL(5s) SLIDING(3s) '. \ + format(super_stream_name, dbname, super_stream_tb_name, aggExpr, dbname, stbname) self.execWtSql(wt, stream_sql) Logging.debug("[OPS] stream is creating at {}".format(time.time())) class TaskCreateTopic(StateTransitionTask): - + @classmethod def getEndState(cls): return StateHasData() @@ -1798,40 +1810,46 @@ class TaskCreateTopic(StateTransitionTask): def _executeInternal(self, te: TaskExecutor, wt: WorkerThread): dbname = self._db.getName() - - sub_topic_name = dbname+ '_sub_topic' - super_topic_name = dbname+ '_super_topic' - stable_topic = dbname+ '_stable_topic' - db_topic = 'database_' + dbname+ '_topics' + + sub_topic_name = dbname + '_sub_topic' + super_topic_name = dbname + '_super_topic' + stable_topic = dbname + '_stable_topic' + db_topic = 'database_' + dbname + '_topics' if not self._db.exists(wt.getDbConn()): Logging.debug("Skipping task, no DB yet") return - sTable = self._db.getFixedSuperTable() # type: TdSuperTable + sTable = self._db.getFixedSuperTable() # type: TdSuperTable # wt.execSql("use db") # should always be in place # create topic if not exists topic_ctb_column as select ts, c1, c2, c3 from stb1; - stbname =sTable.getName() + stbname = sTable.getName() sub_tables = sTable.getRegTables(wt.getDbConn()) - scalarExpr = Dice.choice([ '*','speed','color','abs(speed)','acos(speed)','asin(speed)','atan(speed)','ceil(speed)','cos(speed)','cos(speed)', - 'floor(speed)','log(speed,2)','pow(speed,2)','round(speed)','sin(speed)','sqrt(speed)','char_length(color)','concat(color,color)', - 'concat_ws(" ", color,color," ")','length(color)', 'lower(color)', 'ltrim(color)','substr(color , 2)','upper(color)','cast(speed as double)', - 'cast(ts as bigint)']) + scalarExpr = Dice.choice( + ['*', 'speed', 'color', 'abs(speed)', 'acos(speed)', 'asin(speed)', 'atan(speed)', 'ceil(speed)', + 'cos(speed)', 'cos(speed)', + 'floor(speed)', 'log(speed,2)', 'pow(speed,2)', 'round(speed)', 'sin(speed)', 'sqrt(speed)', + 'char_length(color)', 'concat(color,color)', + 'concat_ws(" ", color,color," ")', 'length(color)', 'lower(color)', 'ltrim(color)', 'substr(color , 2)', + 'upper(color)', 'cast(speed as double)', + 'cast(ts as bigint)']) topic_sql = '' # set default value - if Dice.throw(3)==0: # create topic : source data from sub query - if sub_tables: # if not empty + if Dice.throw(3) == 0: # create topic : source data from sub query + if sub_tables: # if not empty sub_tbname = sub_tables[0] # create topic : source data from sub query of sub stable - topic_sql = 'create topic {} as select {} FROM {}.{} ; '.format(sub_topic_name,scalarExpr,dbname,sub_tbname) - + topic_sql = 'create topic {} as select {} FROM {}.{} ; '.format(sub_topic_name, scalarExpr, dbname, + sub_tbname) + else: # create topic : source data from sub query of stable - topic_sql = 'create topic {} as select {} FROM {}.{} '.format(super_topic_name,scalarExpr, dbname,stbname) - elif Dice.throw(3)==1: # create topic : source data from super table - topic_sql = 'create topic {} AS STABLE {}.{} '.format(stable_topic,dbname,stbname) - - elif Dice.throw(3)==2: # create topic : source data from whole database - topic_sql = 'create topic {} AS DATABASE {} '.format(db_topic,dbname) + topic_sql = 'create topic {} as select {} FROM {}.{} '.format(super_topic_name, scalarExpr, dbname, + stbname) + elif Dice.throw(3) == 1: # create topic : source data from super table + topic_sql = 'create topic {} AS STABLE {}.{} '.format(stable_topic, dbname, stbname) + + elif Dice.throw(3) == 2: # create topic : source data from whole database + topic_sql = 'create topic {} AS DATABASE {} '.format(db_topic, dbname) else: pass @@ -1840,8 +1858,9 @@ class TaskCreateTopic(StateTransitionTask): self.execWtSql(wt, topic_sql) Logging.debug("[OPS] db topic is creating at {}".format(time.time())) + class TaskDropTopics(StateTransitionTask): - + @classmethod def getEndState(cls): return StateHasData() @@ -1852,21 +1871,21 @@ class TaskDropTopics(StateTransitionTask): def _executeInternal(self, te: TaskExecutor, wt: WorkerThread): dbname = self._db.getName() - if not self._db.exists(wt.getDbConn()): Logging.debug("Skipping task, no DB yet") return - sTable = self._db.getFixedSuperTable() # type: TdSuperTable + sTable = self._db.getFixedSuperTable() # type: TdSuperTable # wt.execSql("use db") # should always be in place tblName = sTable.getName() if sTable.hasTopics(wt.getDbConn()): - sTable.dropTopics(wt.getDbConn(),dbname,None) # drop topics of database - sTable.dropTopics(wt.getDbConn(),dbname,tblName) # drop topics of stable + sTable.dropTopics(wt.getDbConn(), dbname, None) # drop topics of database + sTable.dropTopics(wt.getDbConn(), dbname, tblName) # drop topics of stable + class TaskDropStreams(StateTransitionTask): - + @classmethod def getEndState(cls): return StateHasData() @@ -1877,20 +1896,20 @@ class TaskDropStreams(StateTransitionTask): def _executeInternal(self, te: TaskExecutor, wt: WorkerThread): # dbname = self._db.getName() - if not self._db.exists(wt.getDbConn()): Logging.debug("Skipping task, no DB yet") return - sTable = self._db.getFixedSuperTable() # type: TdSuperTable + sTable = self._db.getFixedSuperTable() # type: TdSuperTable # wt.execSql("use db") # should always be in place # tblName = sTable.getName() if sTable.hasStreams(wt.getDbConn()): sTable.dropStreams(wt.getDbConn()) # drop stream of database + class TaskDropStreamTables(StateTransitionTask): - + @classmethod def getEndState(cls): return StateHasData() @@ -1901,42 +1920,42 @@ class TaskDropStreamTables(StateTransitionTask): def _executeInternal(self, te: TaskExecutor, wt: WorkerThread): # dbname = self._db.getName() - if not self._db.exists(wt.getDbConn()): Logging.debug("Skipping task, no DB yet") return - sTable = self._db.getFixedSuperTable() # type: TdSuperTable - wt.execSql("use db") # should always be in place + sTable = self._db.getFixedSuperTable() # type: TdSuperTable + wt.execSql("use db") # should always be in place # tblName = sTable.getName() if sTable.hasStreamTables(wt.getDbConn()): - sTable.dropStreamTables(wt.getDbConn()) # drop stream tables + sTable.dropStreamTables(wt.getDbConn()) # drop stream tables + class TaskCreateConsumers(StateTransitionTask): - + @classmethod def getEndState(cls): return StateHasData() @classmethod def canBeginFrom(cls, state: AnyState): - return state.canCreateConsumers() + return state.canCreateConsumers() def _executeInternal(self, te: TaskExecutor, wt: WorkerThread): - if Config.getConfig().connector_type == 'native': - - sTable = self._db.getFixedSuperTable() # type: TdSuperTable + if Config.getConfig().connector_type == 'native': + + sTable = self._db.getFixedSuperTable() # type: TdSuperTable # wt.execSql("use db") # should always be in place if sTable.hasTopics(wt.getDbConn()): - sTable.createConsumer(wt.getDbConn(),random.randint(1,10)) + sTable.createConsumer(wt.getDbConn(), random.randint(1, 10)) pass else: print(" restful not support tmq consumers") - return + return + - class TaskCreateSuperTable(StateTransitionTask): @classmethod def getEndState(cls): @@ -1951,9 +1970,9 @@ class TaskCreateSuperTable(StateTransitionTask): Logging.debug("Skipping task, no DB yet") return - sTable = self._db.getFixedSuperTable() # type: TdSuperTable + sTable = self._db.getFixedSuperTable() # type: TdSuperTable # wt.execSql("use db") # should always be in place - + sTable.create(wt.getDbConn(), {'ts': TdDataType.TIMESTAMP, 'speed': TdDataType.INT, 'color': TdDataType.BINARY16}, { 'b': TdDataType.BINARY200, 'f': TdDataType.FLOAT}, @@ -1974,11 +1993,10 @@ class TdSuperTable: def getName(self): return self._stName - - def drop(self, dbc, skipCheck = False): + def drop(self, dbc, skipCheck=False): dbName = self._dbName - if self.exists(dbc) : # if myself exists - fullTableName = dbName + '.' + self._stName + if self.exists(dbc): # if myself exists + fullTableName = dbName + '.' + self._stName dbc.execute("DROP TABLE {}".format(fullTableName)) else: if not skipCheck: @@ -1989,64 +2007,55 @@ class TdSuperTable: return dbc.existsSuperTable(self._stName) # TODO: odd semantic, create() method is usually static? - def create(self, dbc, cols: TdColumns, tags: TdTags, dropIfExists = False): + def create(self, dbc, cols: TdColumns, tags: TdTags, dropIfExists=False): '''Creating a super table''' dbName = self._dbName dbc.execute("USE " + dbName) - fullTableName = dbName + '.' + self._stName + fullTableName = dbName + '.' + self._stName if dbc.existsSuperTable(self._stName): - if dropIfExists: - dbc.execute("DROP TABLE {}".format(fullTableName)) - - else: # error + if dropIfExists: + dbc.execute("DROP TABLE {}".format(fullTableName)) + + else: # error raise CrashGenError("Cannot create super table, already exists: {}".format(self._stName)) # Now let's create sql = "CREATE TABLE {} ({})".format( fullTableName, - ",".join(['%s %s'%(k,v.value) for (k,v) in cols.items()])) - if tags : + ",".join(['%s %s' % (k, v.value) for (k, v) in cols.items()])) + if tags: sql += " TAGS ({})".format( - ",".join(['%s %s'%(k,v.value) for (k,v) in tags.items()]) - ) + ",".join(['%s %s' % (k, v.value) for (k, v) in tags.items()]) + ) else: sql += " TAGS (dummy int) " - dbc.execute(sql) + dbc.execute(sql) + + def createConsumer(self, dbc, Consumer_nums): - def createConsumer(self, dbc,Consumer_nums): - def generateConsumer(current_topic_list): - conf = TaosTmqConf() - conf.set("group.id", "tg2") - conf.set("td.connect.user", "root") - conf.set("td.connect.pass", "taosdata") -# conf.set("enable.auto.commit", "true") -# def tmq_commit_cb_print(tmq, resp, offset, param=None): -# print(f"commit: {resp}, tmq: {tmq}, offset: {offset}, param: {param}") -# conf.set_auto_commit_cb(tmq_commit_cb_print, None) - consumer = conf.new_consumer() - topic_list = TaosTmqList() + consumer = Consumer({"group.id": "tg2", "td.connect.user": "root", "td.connect.pass": "taosdata"}) + topic_list = [] for topic in current_topic_list: topic_list.append(topic) - try: - consumer.subscribe(topic_list) - except TmqError as e : - pass + + consumer.subscribe(topic_list) # consumer with random work life time_start = time.time() while 1: - res = consumer.poll(1000) - if time.time() - time_start >random.randint(5,50) : + res = consumer.poll(1) + consumer.commit(res) + if time.time() - time_start > random.randint(5, 50): break try: consumer.unsubscribe() - except TmqError as e : + except TmqError as e: pass return - + # mulit Consumer current_topic_list = self.getTopicLists(dbc) for i in range(Consumer_nums): @@ -2067,84 +2076,86 @@ class TdSuperTable: def getRegTables(self, dbc: DbConn): dbName = self._dbName try: - dbc.query("select distinct TBNAME from {}.{}".format(dbName, self._stName)) # TODO: analyze result set later - except taos.error.ProgrammingError as err: - errno2 = Helper.convertErrno(err.errno) + dbc.query("select distinct TBNAME from {}.{}".format(dbName, + self._stName)) # TODO: analyze result set later + except taos.error.ProgrammingError as err: + errno2 = Helper.convertErrno(err.errno) Logging.debug("[=] Failed to get tables from super table: errno=0x{:X}, msg: {}".format(errno2, err)) raise qr = dbc.getQueryResult() - return [v[0] for v in qr] # list transformation, ref: https://stackoverflow.com/questions/643823/python-list-transformation + return [v[0] for v in + qr] # list transformation, ref: https://stackoverflow.com/questions/643823/python-list-transformation def hasRegTables(self, dbc: DbConn): - + if dbc.existsSuperTable(self._stName): return dbc.query("SELECT * FROM {}.{}".format(self._dbName, self._stName)) > 0 else: return False - def hasStreamTables(self,dbc: DbConn): - + def hasStreamTables(self, dbc: DbConn): + return dbc.query("show {}.stables like 'stream_tb%'".format(self._dbName)) > 0 - def hasStreams(self,dbc: DbConn): + def hasStreams(self, dbc: DbConn): return dbc.query("show streams") > 0 - def hasTopics(self,dbc: DbConn): - + def hasTopics(self, dbc: DbConn): + return dbc.query("show topics") > 0 - def dropTopics(self,dbc: DbConn , dbname=None,stb_name=None): + def dropTopics(self, dbc: DbConn, dbname=None, stb_name=None): dbc.query("show topics ") topics = dbc.getQueryResult() - if dbname !=None and stb_name == None : - + if dbname != None and stb_name == None: + for topic in topics: if dbname in topic[0] and topic[0].startswith("database"): try: dbc.execute('drop topic {}'.format(topic[0])) - Logging.debug("[OPS] topic {} is droping at {}".format(topic,time.time())) - except taos.error.ProgrammingError as err: + Logging.debug("[OPS] topic {} is droping at {}".format(topic, time.time())) + except taos.error.ProgrammingError as err: errno = Helper.convertErrno(err.errno) - if errno in [0x03EB]: # Topic subscribed cannot be dropped - pass + if errno in [0x03EB]: # Topic subscribed cannot be dropped + pass # for subsript in subscriptions: - + else: pass pass return True - elif dbname !=None and stb_name!= None: + elif dbname != None and stb_name != None: for topic in topics: if topic[0].startswith(self._dbName) and topic[0].endswith('topic'): dbc.execute('drop topic {}'.format(topic[0])) - Logging.debug("[OPS] topic {} is droping at {}".format(topic,time.time())) + Logging.debug("[OPS] topic {} is droping at {}".format(topic, time.time())) return True else: return True pass - def dropStreams(self,dbc:DbConn): + def dropStreams(self, dbc: DbConn): dbc.query("show streams ") Streams = dbc.getQueryResult() for Stream in Streams: if Stream[0].startswith(self._dbName): dbc.execute('drop stream {}'.format(Stream[0])) - + return not dbc.query("show streams ") > 0 def dropStreamTables(self, dbc: DbConn): dbc.query("show {}.stables like 'stream_tb%'".format(self._dbName)) - + StreamTables = dbc.getQueryResult() - for StreamTable in StreamTables: + for StreamTable in StreamTables: if self.dropStreams(dbc): - dbc.execute('drop table {}.{}'.format(self._dbName,StreamTable[0])) - + dbc.execute('drop table {}.{}'.format(self._dbName, StreamTable[0])) + return not dbc.query("show {}.stables like 'stream_tb%'".format(self._dbName)) def ensureRegTable(self, task: Optional[Task], dbc: DbConn, regTableName: str): @@ -2155,16 +2166,16 @@ class TdSuperTable: ''' dbName = self._dbName sql = "select tbname from {}.{} where tbname in ('{}')".format(dbName, self._stName, regTableName) - if dbc.query(sql) >= 1 : # reg table exists already + if dbc.query(sql) >= 1: # reg table exists already return # acquire a lock first, so as to be able to *verify*. More details in TD-1471 - fullTableName = dbName + '.' + regTableName + fullTableName = dbName + '.' + regTableName if task is not None: # Somethime thie operation is requested on behalf of a "task" # Logging.info("Locking table for creation: {}".format(fullTableName)) - task.lockTable(fullTableName) # in which case we'll lock this table to ensure serialized access + task.lockTable(fullTableName) # in which case we'll lock this table to ensure serialized access # Logging.info("Table locked for creation".format(fullTableName)) - Progress.emit(Progress.CREATE_TABLE_ATTEMPT) # ATTEMPT to create a new table + Progress.emit(Progress.CREATE_TABLE_ATTEMPT) # ATTEMPT to create a new table # print("(" + fullTableName[-3:] + ")", end="", flush=True) try: sql = "CREATE TABLE {} USING {}.{} tags ({})".format( @@ -2176,17 +2187,17 @@ class TdSuperTable: finally: if task is not None: # Logging.info("Unlocking table after creation: {}".format(fullTableName)) - task.unlockTable(fullTableName) # no matter what + task.unlockTable(fullTableName) # no matter what # Logging.info("Table unlocked after creation: {}".format(fullTableName)) - def _getTagStrForSql(self, dbc) : + def _getTagStrForSql(self, dbc): tags = self._getTags(dbc) tagStrs = [] - for tagName in tags: + for tagName in tags: tagType = tags[tagName] if tagType == 'BINARY': tagStrs.append("'Beijing-Shanghai-LosAngeles'") - elif tagType== 'VARCHAR': + elif tagType == 'VARCHAR': tagStrs.append("'London-Paris-Berlin'") elif tagType == 'FLOAT': tagStrs.append('9.9') @@ -2200,12 +2211,12 @@ class TdSuperTable: dbc.query("DESCRIBE {}.{}".format(self._dbName, self._stName)) stCols = dbc.getQueryResult() # print(stCols) - ret = {row[0]:row[1] for row in stCols if row[3]=='TAG'} # name:type + ret = {row[0]: row[1] for row in stCols if row[3] == 'TAG'} # name:type # print("Tags retrieved: {}".format(ret)) return ret def addTag(self, dbc, tagName, tagType): - if tagName in self._getTags(dbc): # already + if tagName in self._getTags(dbc): # already return # sTable.addTag("extraTag", "int") sql = "alter table {}.{} add tag {} {}".format( @@ -2213,33 +2224,33 @@ class TdSuperTable: dbc.execute(sql) def dropTag(self, dbc, tagName): - if not tagName in self._getTags(dbc): # don't have this tag + if not tagName in self._getTags(dbc): # don't have this tag return sql = "alter table {}.{} drop tag {}".format(self._dbName, self._stName, tagName) dbc.execute(sql) def changeTag(self, dbc, oldTag, newTag): tags = self._getTags(dbc) - if not oldTag in tags: # don't have this tag + if not oldTag in tags: # don't have this tag return - if newTag in tags: # already have this tag + if newTag in tags: # already have this tag return sql = "alter table {}.{} change tag {} {}".format(self._dbName, self._stName, oldTag, newTag) dbc.execute(sql) def generateQueries(self, dbc: DbConn) -> List[SqlQuery]: ''' Generate queries to test/exercise this super table ''' - ret = [] # type: List[SqlQuery] + ret = [] # type: List[SqlQuery] for rTbName in self.getRegTables(dbc): # regular tables - - filterExpr = Dice.choice([ # TODO: add various kind of WHERE conditions + + filterExpr = Dice.choice([ # TODO: add various kind of WHERE conditions None ]) # Run the query against the regular table first - doAggr = (Dice.throw(2) == 0) # 1 in 2 chance - if not doAggr: # don't do aggregate query, just simple one + doAggr = (Dice.throw(2) == 0) # 1 in 2 chance + if not doAggr: # don't do aggregate query, just simple one commonExpr = Dice.choice([ '*', 'abs(speed)', @@ -2256,7 +2267,7 @@ class TdSuperTable: 'sin(speed)', 'sqrt(speed)', 'char_length(color)', - 'concat(color,color)', + 'concat(color,color)', 'concat_ws(" ", color,color," ")', 'length(color)', 'lower(color)', @@ -2276,26 +2287,26 @@ class TdSuperTable: 'distinct(color)' ] ) - ret.append(SqlQuery( # reg table + ret.append(SqlQuery( # reg table "select {} from {}.{}".format(commonExpr, self._dbName, rTbName))) - ret.append(SqlQuery( # super table + ret.append(SqlQuery( # super table "select {} from {}.{}".format(commonExpr, self._dbName, self.getName()))) - else: # Aggregate query - aggExpr = Dice.choice([ + else: # Aggregate query + aggExpr = Dice.choice([ 'count(*)', 'avg(speed)', # 'twa(speed)', # TODO: this one REQUIRES a where statement, not reasonable - 'sum(speed)', - 'stddev(speed)', + 'sum(speed)', + 'stddev(speed)', # SELECTOR functions - 'min(speed)', - 'max(speed)', - 'first(speed)', + 'min(speed)', + 'max(speed)', + 'first(speed)', 'last(speed)', - 'top(speed, 50)', # TODO: not supported? - 'bottom(speed, 50)', # TODO: not supported? - 'apercentile(speed, 10)', # TODO: TD-1316 - 'last_row(*)', # TODO: commented out per TD-3231, we should re-create + 'top(speed, 50)', # TODO: not supported? + 'bottom(speed, 50)', # TODO: not supported? + 'apercentile(speed, 10)', # TODO: TD-1316 + 'last_row(*)', # TODO: commented out per TD-3231, we should re-create # Transformation Functions # 'diff(speed)', # TODO: no supported?! 'spread(speed)', @@ -2313,21 +2324,21 @@ class TdSuperTable: 'sample(speed,5)', 'STATECOUNT(speed,"LT",1)', 'STATEDURATION(speed,"LT",1)', - 'twa(speed)' - - ]) # TODO: add more from 'top' + 'twa(speed)' + + ]) # TODO: add more from 'top' - # if aggExpr not in ['stddev(speed)']: # STDDEV not valid for super tables?! (Done in TD-1049) sql = "select {} from {}.{}".format(aggExpr, self._dbName, self.getName()) - if Dice.throw(3) == 0: # 1 in X chance - partion_expr = Dice.choice(['color','tbname']) + if Dice.throw(3) == 0: # 1 in X chance + partion_expr = Dice.choice(['color', 'tbname']) sql = sql + ' partition BY ' + partion_expr + ' order by ' + partion_expr Progress.emit(Progress.QUERY_GROUP_BY) # Logging.info("Executing GROUP-BY query: " + sql) ret.append(SqlQuery(sql)) - return ret + return ret + class TaskReadData(StateTransitionTask): @classmethod @@ -2345,60 +2356,61 @@ class TaskReadData(StateTransitionTask): def _reconnectIfNeeded(self, wt): # 1 in 20 chance, simulate a broken connection, only if service stable (not restarting) - if random.randrange(20)==0: # and self._canRestartService(): # TODO: break connection in all situations + if random.randrange(20) == 0: # and self._canRestartService(): # TODO: break connection in all situations # Logging.info("Attempting to reconnect to server") # TODO: change to DEBUG - Progress.emit(Progress.SERVICE_RECONNECT_START) + Progress.emit(Progress.SERVICE_RECONNECT_START) try: wt.getDbConn().close() wt.getDbConn().open() - except ConnectionError as err: # may fail + except ConnectionError as err: # may fail if not gSvcMgr: Logging.error("Failed to reconnect in client-only mode") - raise # Not OK if we are running in client-only mode - if gSvcMgr.isRunning(): # may have race conditon, but low prob, due to + raise # Not OK if we are running in client-only mode + if gSvcMgr.isRunning(): # may have race conditon, but low prob, due to Logging.error("Failed to reconnect when managed server is running") - raise # Not OK if we are running normally + raise # Not OK if we are running normally - Progress.emit(Progress.SERVICE_RECONNECT_FAILURE) + Progress.emit(Progress.SERVICE_RECONNECT_FAILURE) # Logging.info("Ignoring DB reconnect error") # print("_r", end="", flush=True) - Progress.emit(Progress.SERVICE_RECONNECT_SUCCESS) + Progress.emit(Progress.SERVICE_RECONNECT_SUCCESS) # The above might have taken a lot of time, service might be running # by now, causing error below to be incorrectly handled due to timing issue - return # TODO: fix server restart status race condtion - + return # TODO: fix server restart status race condtion def _executeInternal(self, te: TaskExecutor, wt: WorkerThread): self._reconnectIfNeeded(wt) dbc = wt.getDbConn() sTable = self._db.getFixedSuperTable() - + for q in sTable.generateQueries(dbc): # regular tables try: sql = q.getSql() # if 'GROUP BY' in sql: # Logging.info("Executing GROUP-BY query: " + sql) dbc.execute(sql) - except taos.error.ProgrammingError as err: + except taos.error.ProgrammingError as err: errno2 = Helper.convertErrno(err.errno) Logging.debug("[=] Read Failure: errno=0x{:X}, msg: {}, SQL: {}".format(errno2, err, dbc.getLastSql())) raise + class SqlQuery: @classmethod def buildRandom(cls, db: Database): '''Build a random query against a certain database''' - + dbName = db.getName() - def __init__(self, sql:str = None): + def __init__(self, sql: str = None): self._sql = sql def getSql(self): return self._sql - + + class TaskDropSuperTable(StateTransitionTask): @classmethod def getEndState(cls): @@ -2421,7 +2433,7 @@ class TaskDropSuperTable(StateTransitionTask): regTableName = self.getRegTableName(i) # "db.reg_table_{}".format(i) try: self.execWtSql(wt, "drop table {}.{}". - format(self._db.getName(), regTableName)) # nRows always 0, like MySQL + format(self._db.getName(), regTableName)) # nRows always 0, like MySQL except taos.error.ProgrammingError as err: # correcting for strange error number scheme errno2 = Helper.convertErrno(err.errno) @@ -2429,7 +2441,6 @@ class TaskDropSuperTable(StateTransitionTask): isSuccess = False Logging.debug("[DB] Acceptable error when dropping a table") continue # try to delete next regular table - if (not tickOutput): tickOutput = True # Print only one time @@ -2441,8 +2452,6 @@ class TaskDropSuperTable(StateTransitionTask): # Drop the super table itself tblName = self._db.getFixedSuperTableName() self.execWtSql(wt, "drop table {}.{}".format(self._db.getName(), tblName)) - - class TaskAlterTags(StateTransitionTask): @@ -2472,6 +2481,7 @@ class TaskAlterTags(StateTransitionTask): sTable.changeTag(dbc, "extraTag", "newTag") # sql = "alter table db.{} change tag extraTag newTag".format(tblName) + class TaskRestartService(StateTransitionTask): _isRunning = False _classLock = threading.Lock() @@ -2484,11 +2494,12 @@ class TaskRestartService(StateTransitionTask): def canBeginFrom(cls, state: AnyState): if Config.getConfig().auto_start_service: return state.canDropFixedSuperTable() # Basicallly when we have the super table - return False # don't run this otherwise + return False # don't run this otherwise CHANCE_TO_RESTART_SERVICE = 200 + def _executeInternal(self, te: TaskExecutor, wt: WorkerThread): - if not Config.getConfig().auto_start_service: # only execute when we are in -a mode + if not Config.getConfig().auto_start_service: # only execute when we are in -a mode print("_a", end="", flush=True) return @@ -2498,20 +2509,22 @@ class TaskRestartService(StateTransitionTask): return self._isRunning = True - if Dice.throw(self.CHANCE_TO_RESTART_SERVICE) == 0: # 1 in N chance + if Dice.throw(self.CHANCE_TO_RESTART_SERVICE) == 0: # 1 in N chance dbc = wt.getDbConn() - dbc.execute("select * from information_schema.ins_databases") # simple delay, align timing with other workers + dbc.execute( + "select * from information_schema.ins_databases") # simple delay, align timing with other workers gSvcMgr.restart() self._isRunning = False + class TaskAddData(StateTransitionTask): # Track which table is being actively worked on activeTable: Set[int] = set() # We use these two files to record operations to DB, useful for power-off tests - fAddLogReady = None # type: Optional[io.TextIOWrapper] - fAddLogDone = None # type: Optional[io.TextIOWrapper] + fAddLogReady = None # type: Optional[io.TextIOWrapper] + fAddLogDone = None # type: Optional[io.TextIOWrapper] @classmethod def prepToRecordOps(cls): @@ -2532,12 +2545,12 @@ class TaskAddData(StateTransitionTask): def canBeginFrom(cls, state: AnyState): return state.canAddData() - def _lockTableIfNeeded(self, fullTableName, extraMsg = ''): + def _lockTableIfNeeded(self, fullTableName, extraMsg=''): if Config.getConfig().verify_data: # Logging.info("Locking table: {}".format(fullTableName)) - self.lockTable(fullTableName) + self.lockTable(fullTableName) # Logging.info("Table locked {}: {}".format(extraMsg, fullTableName)) - # print("_w" + str(nextInt % 100), end="", flush=True) # Trace what was written + # print("_w" + str(nextInt % 100), end="", flush=True) # Trace what was written else: # Logging.info("Skipping locking table") pass @@ -2545,15 +2558,15 @@ class TaskAddData(StateTransitionTask): def _unlockTableIfNeeded(self, fullTableName): if Config.getConfig().verify_data: # Logging.info("Unlocking table: {}".format(fullTableName)) - self.unlockTable(fullTableName) + self.unlockTable(fullTableName) # Logging.info("Table unlocked: {}".format(fullTableName)) else: pass # Logging.info("Skipping unlocking table") - def _addDataInBatch(self, db, dbc, regTableName, te: TaskExecutor): - numRecords = self.LARGE_NUMBER_OF_RECORDS if Config.getConfig().larger_data else self.SMALL_NUMBER_OF_RECORDS - + def _addDataInBatch(self, db, dbc, regTableName, te: TaskExecutor): + numRecords = self.LARGE_NUMBER_OF_RECORDS if Config.getConfig().larger_data else self.SMALL_NUMBER_OF_RECORDS + fullTableName = db.getName() + '.' + regTableName self._lockTableIfNeeded(fullTableName, 'batch') @@ -2571,10 +2584,8 @@ class TaskAddData(StateTransitionTask): # Logging.info("Data added in batch: {}".format(sql)) self._unlockTableIfNeeded(fullTableName) - - - def _addData(self, db: Database, dbc, regTableName, te: TaskExecutor): # implied: NOT in batches - numRecords = self.LARGE_NUMBER_OF_RECORDS if Config.getConfig().larger_data else self.SMALL_NUMBER_OF_RECORDS + def _addData(self, db: Database, dbc, regTableName, te: TaskExecutor): # implied: NOT in batches + numRecords = self.LARGE_NUMBER_OF_RECORDS if Config.getConfig().larger_data else self.SMALL_NUMBER_OF_RECORDS for j in range(numRecords): # number of records per table intToWrite = db.getNextInt() @@ -2587,13 +2598,14 @@ class TaskAddData(StateTransitionTask): self.fAddLogReady.write("Ready to write {} to {}\n".format(intToWrite, regTableName)) self.fAddLogReady.flush() os.fsync(self.fAddLogReady.fileno()) - + # TODO: too ugly trying to lock the table reliably, refactor... fullTableName = db.getName() + '.' + regTableName - self._lockTableIfNeeded(fullTableName) # so that we are verify read-back. TODO: deal with exceptions before unlock - + self._lockTableIfNeeded( + fullTableName) # so that we are verify read-back. TODO: deal with exceptions before unlock + try: - sql = "INSERT INTO {} VALUES ('{}', {}, '{}');".format( # removed: tags ('{}', {}) + sql = "INSERT INTO {} VALUES ('{}', {}, '{}');".format( # removed: tags ('{}', {}) fullTableName, # ds.getFixedSuperTableName(), # ds.getNextBinary(), ds.getNextFloat(), @@ -2604,55 +2616,56 @@ class TaskAddData(StateTransitionTask): intWrote = intToWrite # Quick hack, attach an update statement here. TODO: create an "update" task - if (not Config.getConfig().use_shadow_db) and Dice.throw(5) == 0: # 1 in N chance, plus not using shaddow DB - intToUpdate = db.getNextInt() # Updated， but should not succeed + if (not Config.getConfig().use_shadow_db) and Dice.throw( + 5) == 0: # 1 in N chance, plus not using shaddow DB + intToUpdate = db.getNextInt() # Updated， but should not succeed nextColor = db.getNextColor() - sql = "INSERt INTO {} VALUES ('{}', {}, '{}');".format( # "INSERt" means "update" here - fullTableName, - nextTick, intToUpdate, nextColor) + sql = "INSERt INTO {} VALUES ('{}', {}, '{}');".format( # "INSERt" means "update" here + fullTableName, + nextTick, intToUpdate, nextColor) # sql = "UPDATE {} set speed={}, color='{}' WHERE ts='{}'".format( # fullTableName, db.getNextInt(), db.getNextColor(), nextTick) dbc.execute(sql) - intWrote = intToUpdate # We updated, seems TDengine non-cluster accepts this. + intWrote = intToUpdate # We updated, seems TDengine non-cluster accepts this. - except: # Any exception at all - self._unlockTableIfNeeded(fullTableName) + except: # Any exception at all + self._unlockTableIfNeeded(fullTableName) raise # Now read it back and verify, we might encounter an error if table is dropped - if Config.getConfig().verify_data: # only if command line asks for it + if Config.getConfig().verify_data: # only if command line asks for it try: readBack = dbc.queryScalar("SELECT speed from {}.{} WHERE ts='{}'". - format(db.getName(), regTableName, nextTick)) - if readBack != intWrote : + format(db.getName(), regTableName, nextTick)) + if readBack != intWrote: raise taos.error.ProgrammingError( "Failed to read back same data, wrote: {}, read: {}" .format(intWrote, readBack), 0x999) except taos.error.ProgrammingError as err: errno = Helper.convertErrno(err.errno) - if errno == CrashGenError.INVALID_EMPTY_RESULT: # empty result + if errno == CrashGenError.INVALID_EMPTY_RESULT: # empty result raise taos.error.ProgrammingError( "Failed to read back same data for tick: {}, wrote: {}, read: EMPTY" .format(nextTick, intWrote), errno) - elif errno == CrashGenError.INVALID_MULTIPLE_RESULT : # multiple results + elif errno == CrashGenError.INVALID_MULTIPLE_RESULT: # multiple results raise taos.error.ProgrammingError( "Failed to read back same data for tick: {}, wrote: {}, read: MULTIPLE RESULTS" .format(nextTick, intWrote), errno) - elif errno in [0x218, 0x362]: # table doesn't exist + elif errno in [0x218, 0x362]: # table doesn't exist # do nothing pass else: # Re-throw otherwise raise finally: - self._unlockTableIfNeeded(fullTableName) # Quite ugly, refactor lock/unlock + self._unlockTableIfNeeded(fullTableName) # Quite ugly, refactor lock/unlock # Done with read-back verification, unlock the table now else: - self._unlockTableIfNeeded(fullTableName) + self._unlockTableIfNeeded(fullTableName) - # Successfully wrote the data into the DB, let's record it somehow + # Successfully wrote the data into the DB, let's record it somehow te.recordDataMark(intWrote) if Config.getConfig().record_ops: @@ -2666,17 +2679,17 @@ class TaskAddData(StateTransitionTask): # ds = self._dbManager # Quite DANGEROUS here, may result in multi-thread client access db = self._db dbc = wt.getDbConn() - numTables = self.LARGE_NUMBER_OF_TABLES if Config.getConfig().larger_data else self.SMALL_NUMBER_OF_TABLES + numTables = self.LARGE_NUMBER_OF_TABLES if Config.getConfig().larger_data else self.SMALL_NUMBER_OF_TABLES numRecords = self.LARGE_NUMBER_OF_RECORDS if Config.getConfig().larger_data else self.SMALL_NUMBER_OF_RECORDS - tblSeq = list(range(numTables )) - random.shuffle(tblSeq) # now we have random sequence + tblSeq = list(range(numTables)) + random.shuffle(tblSeq) # now we have random sequence for i in tblSeq: if (i in self.activeTable): # wow already active # print("x", end="", flush=True) # concurrent insertion Progress.emit(Progress.CONCURRENT_INSERTION) else: self.activeTable.add(i) # marking it active - + dbName = db.getName() sTable = db.getFixedSuperTable() regTableName = self.getRegTableName(i) # "db.reg_table_{}".format(i) @@ -2684,21 +2697,22 @@ class TaskAddData(StateTransitionTask): # self._lockTable(fullTableName) # "create table" below. Stop it if the table is "locked" sTable.ensureRegTable(self, wt.getDbConn(), regTableName) # Ensure the table exists # self._unlockTable(fullTableName) - - if Dice.throw(1) == 0: # 1 in 2 chance + + if Dice.throw(1) == 0: # 1 in 2 chance self._addData(db, dbc, regTableName, te) else: self._addDataInBatch(db, dbc, regTableName, te) self.activeTable.discard(i) # not raising an error, unlike remove + class TaskDeleteData(StateTransitionTask): # Track which table is being actively worked on activeTable: Set[int] = set() # We use these two files to record operations to DB, useful for power-off tests - fAddLogReady = None # type: Optional[io.TextIOWrapper] - fAddLogDone = None # type: Optional[io.TextIOWrapper] + fAddLogReady = None # type: Optional[io.TextIOWrapper] + fAddLogDone = None # type: Optional[io.TextIOWrapper] @classmethod def prepToRecordOps(cls): @@ -2719,12 +2733,12 @@ class TaskDeleteData(StateTransitionTask): def canBeginFrom(cls, state: AnyState): return state.canDeleteData() - def _lockTableIfNeeded(self, fullTableName, extraMsg = ''): + def _lockTableIfNeeded(self, fullTableName, extraMsg=''): if Config.getConfig().verify_data: # Logging.info("Locking table: {}".format(fullTableName)) - self.lockTable(fullTableName) + self.lockTable(fullTableName) # Logging.info("Table locked {}: {}".format(extraMsg, fullTableName)) - # print("_w" + str(nextInt % 100), end="", flush=True) # Trace what was written + # print("_w" + str(nextInt % 100), end="", flush=True) # Trace what was written else: # Logging.info("Skipping locking table") pass @@ -2732,15 +2746,15 @@ class TaskDeleteData(StateTransitionTask): def _unlockTableIfNeeded(self, fullTableName): if Config.getConfig().verify_data: # Logging.info("Unlocking table: {}".format(fullTableName)) - self.unlockTable(fullTableName) + self.unlockTable(fullTableName) # Logging.info("Table unlocked: {}".format(fullTableName)) else: pass # Logging.info("Skipping unlocking table") - def _deleteData(self, db: Database, dbc, regTableName, te: TaskExecutor): # implied: NOT in batches - numRecords = self.LARGE_NUMBER_OF_RECORDS if Config.getConfig().larger_data else self.SMALL_NUMBER_OF_RECORDS - del_Records = int(numRecords/5) + def _deleteData(self, db: Database, dbc, regTableName, te: TaskExecutor): # implied: NOT in batches + numRecords = self.LARGE_NUMBER_OF_RECORDS if Config.getConfig().larger_data else self.SMALL_NUMBER_OF_RECORDS + del_Records = int(numRecords / 5) if Dice.throw(2) == 0: for j in range(del_Records): # number of records per table intToWrite = db.getNextInt() @@ -2753,13 +2767,14 @@ class TaskDeleteData(StateTransitionTask): self.fAddLogReady.write("Ready to delete {} to {}\n".format(intToWrite, regTableName)) self.fAddLogReady.flush() os.fsync(self.fAddLogReady.fileno()) - + # TODO: too ugly trying to lock the table reliably, refactor... fullTableName = db.getName() + '.' + regTableName - self._lockTableIfNeeded(fullTableName) # so that we are verify read-back. TODO: deal with exceptions before unlock - + self._lockTableIfNeeded( + fullTableName) # so that we are verify read-back. TODO: deal with exceptions before unlock + try: - sql = "delete from {} where ts = '{}' ;".format( # removed: tags ('{}', {}) + sql = "delete from {} where ts = '{}' ;".format( # removed: tags ('{}', {}) fullTableName, # ds.getFixedSuperTableName(), # ds.getNextBinary(), ds.getNextFloat(), @@ -2772,45 +2787,46 @@ class TaskDeleteData(StateTransitionTask): intWrote = intToWrite # Quick hack, attach an update statement here. TODO: create an "update" task - if (not Config.getConfig().use_shadow_db) and Dice.throw(5) == 0: # 1 in N chance, plus not using shaddow DB - intToUpdate = db.getNextInt() # Updated， but should not succeed + if (not Config.getConfig().use_shadow_db) and Dice.throw( + 5) == 0: # 1 in N chance, plus not using shaddow DB + intToUpdate = db.getNextInt() # Updated， but should not succeed # nextColor = db.getNextColor() - sql = "delete from {} where ts = '{}' ;".format( # "INSERt" means "update" here - fullTableName, - nextTick) + sql = "delete from {} where ts = '{}' ;".format( # "INSERt" means "update" here + fullTableName, + nextTick) # sql = "UPDATE {} set speed={}, color='{}' WHERE ts='{}'".format( # fullTableName, db.getNextInt(), db.getNextColor(), nextTick) dbc.execute(sql) - intWrote = intToUpdate # We updated, seems TDengine non-cluster accepts this. + intWrote = intToUpdate # We updated, seems TDengine non-cluster accepts this. - except: # Any exception at all - self._unlockTableIfNeeded(fullTableName) + except: # Any exception at all + self._unlockTableIfNeeded(fullTableName) raise # Now read it back and verify, we might encounter an error if table is dropped - if Config.getConfig().verify_data: # only if command line asks for it + if Config.getConfig().verify_data: # only if command line asks for it try: dbc.query("SELECT * from {}.{} WHERE ts='{}'". - format(db.getName(), regTableName, nextTick)) + format(db.getName(), regTableName, nextTick)) result = dbc.getQueryResult() - if len(result)==0: + if len(result) == 0: # means data has been delete - print("D1",end="") # DF means delete failed + print("D1", end="") # DF means delete failed else: - print("DF",end="") # DF means delete failed + print("DF", end="") # DF means delete failed except taos.error.ProgrammingError as err: errno = Helper.convertErrno(err.errno) # if errno == CrashGenError.INVALID_EMPTY_RESULT: # empty result # print("D1",end="") # D1 means delete data success and only 1 record - if errno in [0x218, 0x362,0x2662]: # table doesn't exist + if errno in [0x218, 0x362, 0x2662]: # table doesn't exist # do nothing pass else: # Re-throw otherwise raise finally: - self._unlockTableIfNeeded(fullTableName) # Quite ugly, refactor lock/unlock + self._unlockTableIfNeeded(fullTableName) # Quite ugly, refactor lock/unlock # Done with read-back verification, unlock the table now # Successfully wrote the data into the DB, let's record it somehow te.recordDataMark(intWrote) @@ -2824,52 +2840,54 @@ class TaskDeleteData(StateTransitionTask): self.fAddLogReady.write("Ready to delete {} to {}\n".format(intToWrite, regTableName)) self.fAddLogReady.flush() os.fsync(self.fAddLogReady.fileno()) - + # TODO: too ugly trying to lock the table reliably, refactor... fullTableName = db.getName() + '.' + regTableName - self._lockTableIfNeeded(fullTableName) # so that we are verify read-back. TODO: deal with exceptions before unlock - + self._lockTableIfNeeded( + fullTableName) # so that we are verify read-back. TODO: deal with exceptions before unlock + try: - sql = "delete from {} ;".format( # removed: tags ('{}', {}) + sql = "delete from {} ;".format( # removed: tags ('{}', {}) fullTableName) # Logging.info("Adding data: {}".format(sql)) dbc.execute(sql) # Logging.info("Data added: {}".format(sql)) - + # Quick hack, attach an update statement here. TODO: create an "update" task - if (not Config.getConfig().use_shadow_db) and Dice.throw(5) == 0: # 1 in N chance, plus not using shaddow DB - sql = "delete from {} ;".format( # "INSERt" means "update" here - fullTableName) + if (not Config.getConfig().use_shadow_db) and Dice.throw( + 5) == 0: # 1 in N chance, plus not using shaddow DB + sql = "delete from {} ;".format( # "INSERt" means "update" here + fullTableName) dbc.execute(sql) - except: # Any exception at all - self._unlockTableIfNeeded(fullTableName) + except: # Any exception at all + self._unlockTableIfNeeded(fullTableName) raise # Now read it back and verify, we might encounter an error if table is dropped - if Config.getConfig().verify_data: # only if command line asks for it + if Config.getConfig().verify_data: # only if command line asks for it try: dbc.query("SELECT * from {}.{} WHERE ts='{}'". - format(db.getName(), regTableName, nextTick)) + format(db.getName(), regTableName, nextTick)) result = dbc.getQueryResult() - if len(result)==0: + if len(result) == 0: # means data has been delete - print("DA",end="") + print("DA", end="") else: - print("DF",end="") # DF means delete failed + print("DF", end="") # DF means delete failed except taos.error.ProgrammingError as err: errno = Helper.convertErrno(err.errno) # if errno == CrashGenError.INVALID_EMPTY_RESULT: # empty result # print("Da",end="") # Da means delete data success and for all datas - if errno in [0x218, 0x362,0x2662]: # table doesn't exist + if errno in [0x218, 0x362, 0x2662]: # table doesn't exist # do nothing pass else: # Re-throw otherwise raise finally: - self._unlockTableIfNeeded(fullTableName) # Quite ugly, refactor lock/unlock + self._unlockTableIfNeeded(fullTableName) # Quite ugly, refactor lock/unlock # Done with read-back verification, unlock the table now if Config.getConfig().record_ops: @@ -2883,17 +2901,17 @@ class TaskDeleteData(StateTransitionTask): # ds = self._dbManager # Quite DANGEROUS here, may result in multi-thread client access db = self._db dbc = wt.getDbConn() - numTables = self.LARGE_NUMBER_OF_TABLES if Config.getConfig().larger_data else self.SMALL_NUMBER_OF_TABLES + numTables = self.LARGE_NUMBER_OF_TABLES if Config.getConfig().larger_data else self.SMALL_NUMBER_OF_TABLES numRecords = self.LARGE_NUMBER_OF_RECORDS if Config.getConfig().larger_data else self.SMALL_NUMBER_OF_RECORDS - tblSeq = list(range(numTables )) - random.shuffle(tblSeq) # now we have random sequence + tblSeq = list(range(numTables)) + random.shuffle(tblSeq) # now we have random sequence for i in tblSeq: if (i in self.activeTable): # wow already active # print("x", end="", flush=True) # concurrent insertion Progress.emit(Progress.CONCURRENT_INSERTION) else: self.activeTable.add(i) # marking it active - + dbName = db.getName() sTable = db.getFixedSuperTable() regTableName = self.getRegTableName(i) # "db.reg_table_{}".format(i) @@ -2901,54 +2919,57 @@ class TaskDeleteData(StateTransitionTask): # self._lockTable(fullTableName) # "create table" below. Stop it if the table is "locked" sTable.ensureRegTable(self, wt.getDbConn(), regTableName) # Ensure the table exists # self._unlockTable(fullTableName) - + self._deleteData(db, dbc, regTableName, te) - + self.activeTable.discard(i) # not raising an error, unlike remove -class ThreadStacks: # stack info for all threads +class ThreadStacks: # stack info for all threads def __init__(self): self._allStacks = {} - allFrames = sys._current_frames() # All current stack frames, keyed with "ident" + allFrames = sys._current_frames() # All current stack frames, keyed with "ident" for th in threading.enumerate(): # For each thread - stack = traceback.extract_stack(allFrames[th.ident]) #type: ignore # Get stack for a thread - shortTid = th.native_id % 10000 #type: ignore - self._allStacks[shortTid] = stack # Was using th.native_id + stack = traceback.extract_stack(allFrames[th.ident]) # type: ignore # Get stack for a thread + shortTid = th.native_id % 10000 # type: ignore + self._allStacks[shortTid] = stack # Was using th.native_id - def record_current_time(self,current_time): + def record_current_time(self, current_time): self.current_time = current_time - def print(self, filteredEndName = None, filterInternal = False): - for shortTid, stack in self._allStacks.items(): # for each thread, stack frames top to bottom + def print(self, filteredEndName=None, filterInternal=False): + for shortTid, stack in self._allStacks.items(): # for each thread, stack frames top to bottom lastFrame = stack[-1] - if filteredEndName: # we need to filter out stacks that match this name - if lastFrame.name == filteredEndName : # end did not match + if filteredEndName: # we need to filter out stacks that match this name + if lastFrame.name == filteredEndName: # end did not match continue if filterInternal: - if lastFrame.name in ['wait', 'invoke_excepthook', - '_wait', # The Barrier exception - 'svcOutputReader', # the svcMgr thread - '__init__']: # the thread that extracted the stack - continue # ignore + if lastFrame.name in ['wait', 'invoke_excepthook', + '_wait', # The Barrier exception + 'svcOutputReader', # the svcMgr thread + '__init__']: # the thread that extracted the stack + continue # ignore # Now print print("\n<----- Thread Info for LWP/ID: {} (most recent call last) <-----".format(shortTid)) - + lastSqlForThread = DbConn.fetchSqlForThread(shortTid) last_sql_commit_time = DbConn.get_save_sql_time(shortTid) # time_cost = DbConn.get_time_cost() - print("Last SQL statement attempted from thread {} ({:.4f} sec ago) is: {}".format(shortTid, self.current_time-last_sql_commit_time ,lastSqlForThread)) + print("Last SQL statement attempted from thread {} ({:.4f} sec ago) is: {}".format(shortTid, + self.current_time - last_sql_commit_time, + lastSqlForThread)) stackFrame = 0 - for frame in stack: # was using: reversed(stack) + for frame in stack: # was using: reversed(stack) # print(frame) print("[{sf}] File {filename}, line {lineno}, in {name}".format( sf=stackFrame, filename=frame.filename, lineno=frame.lineno, name=frame.name)) print(" {}".format(frame.line)) stackFrame += 1 print("-----> End of Thread Info ----->\n") - if self.current_time-last_sql_commit_time >100: # dead lock occured + if self.current_time - last_sql_commit_time > 100: # dead lock occured print("maybe dead locked of thread {} ".format(shortTid)) + class ClientManager: def __init__(self): Logging.info("Starting service manager") @@ -3041,36 +3062,35 @@ class ClientManager: # time.sleep(2.0) # dbManager = None # release? - def run(self, svcMgr): + def run(self, svcMgr): # self._printLastNumbers() # global gConfig # Prepare Tde Instance global gContainer - tInst = gContainer.defTdeInstance = TdeInstance() # "subdir to hold the instance" + tInst = gContainer.defTdeInstance = TdeInstance() # "subdir to hold the instance" cfg = Config.getConfig() dbManager = DbManager(cfg.connector_type, tInst.getDbTarget()) # Regular function thPool = ThreadPool(cfg.num_threads, cfg.max_steps) self.tc = ThreadCoordinator(thPool, dbManager) - + Logging.info("Starting client instance: {}".format(tInst)) self.tc.run() # print("exec stats: {}".format(self.tc.getExecStats())) # print("TC failed = {}".format(self.tc.isFailed())) - if svcMgr: # gConfig.auto_start_service: + if svcMgr: # gConfig.auto_start_service: svcMgr.stopTaosServices() svcMgr = None - # Release global variables # gConfig = None Config.clearConfig() gSvcMgr = None logger = None - + thPool = None - dbManager.cleanUp() # destructor wouldn't run in time + dbManager.cleanUp() # destructor wouldn't run in time dbManager = None # Print exec status, etc., AFTER showing messages from the server @@ -3082,7 +3102,7 @@ class ClientManager: # Release variables here self.tc = None - gc.collect() # force garbage collection + gc.collect() # force garbage collection # h = hpy() # print("\n----- Final Python Heap -----\n") # print(h.heap()) @@ -3093,37 +3113,38 @@ class ClientManager: # self.tc.getDbManager().cleanUp() # clean up first, so we can show ZERO db connections self.tc.printStats() + class MainExec: - def __init__(self): + def __init__(self): self._clientMgr = None - self._svcMgr = None # type: Optional[ServiceManager] + self._svcMgr = None # type: Optional[ServiceManager] signal.signal(signal.SIGTERM, self.sigIntHandler) - signal.signal(signal.SIGINT, self.sigIntHandler) + signal.signal(signal.SIGINT, self.sigIntHandler) signal.signal(signal.SIGUSR1, self.sigUsrHandler) # different handler! def sigUsrHandler(self, signalNumber, frame): if self._clientMgr: self._clientMgr.sigUsrHandler(signalNumber, frame) - elif self._svcMgr: # Only if no client mgr, we are running alone + elif self._svcMgr: # Only if no client mgr, we are running alone self._svcMgr.sigUsrHandler(signalNumber, frame) - + def sigIntHandler(self, signalNumber, frame): - if self._svcMgr: + if self._svcMgr: self._svcMgr.sigIntHandler(signalNumber, frame) - if self._clientMgr: + if self._clientMgr: self._clientMgr.sigIntHandler(signalNumber, frame) def runClient(self): global gSvcMgr if Config.getConfig().auto_start_service: - gSvcMgr = self._svcMgr = ServiceManager(1) # hack alert - gSvcMgr.startTaosServices() # we start, don't run - + gSvcMgr = self._svcMgr = ServiceManager(1) # hack alert + gSvcMgr.startTaosServices() # we start, don't run + self._clientMgr = ClientManager() ret = None - try: - ret = self._clientMgr.run(self._svcMgr) # stop TAOS service inside + try: + ret = self._clientMgr.run(self._svcMgr) # stop TAOS service inside except requests.exceptions.ConnectionError as err: Logging.warning("Failed to open REST connection to DB: {}".format(err)) # don't raise @@ -3131,10 +3152,11 @@ class MainExec: def runService(self): global gSvcMgr - gSvcMgr = self._svcMgr = ServiceManager(Config.getConfig().num_dnodes) # save it in a global variable TODO: hack alert + gSvcMgr = self._svcMgr = ServiceManager( + Config.getConfig().num_dnodes) # save it in a global variable TODO: hack alert - gSvcMgr.run() # run to some end state - gSvcMgr = self._svcMgr = None + gSvcMgr.run() # run to some end state + gSvcMgr = self._svcMgr = None def _buildCmdLineParser(self): parser = argparse.ArgumentParser( @@ -3145,7 +3167,7 @@ class MainExec: 1. You build TDengine in the top level ./build directory, as described in offical docs 2. You run the server there before this script: ./build/bin/taosd -c test/cfg - ''')) + ''')) parser.add_argument( '-a', @@ -3209,7 +3231,7 @@ class MainExec: '-n', '--dynamic-db-table-names', action='store_true', - help='Use non-fixed names for dbs/tables, for -b, useful for multi-instance executions (default: false)') + help='Use non-fixed names for dbs/tables, for -b, useful for multi-instance executions (default: false)') parser.add_argument( '-o', '--num-dnodes', @@ -3259,19 +3281,18 @@ class MainExec: return parser - - def init(self): # TODO: refactor + def init(self): # TODO: refactor global gContainer - gContainer = Container() # micky-mouse DI + gContainer = Container() # micky-mouse DI - global gSvcMgr # TODO: refactor away + global gSvcMgr # TODO: refactor away gSvcMgr = None parser = self._buildCmdLineParser() Config.init(parser) # Sanity check for arguments - if Config.getConfig().use_shadow_db and Config.getConfig().max_dbs>1 : + if Config.getConfig().use_shadow_db and Config.getConfig().max_dbs > 1: raise CrashGenError("Cannot combine use-shadow-db with max-dbs of more than 1") Logging.clsInit(Config.getConfig().debug) @@ -3282,10 +3303,10 @@ class MainExec: if Config.getConfig().run_tdengine: # run server try: self.runService() - return 0 # success + return 0 # success except ConnectionError as err: Logging.error("Failed to make DB connection, please check DB instance manually") - return -1 # failure + return -1 # failure else: return self.runClient() @@ -3294,7 +3315,7 @@ class Container(): _propertyList = {'defTdeInstance'} def __init__(self): - self._cargo = {} # No cargo at the beginning + self._cargo = {} # No cargo at the beginning def _verifyValidProperty(self, name): if not name in self._propertyList: @@ -3303,10 +3324,10 @@ class Container(): # Called for an attribute, when other mechanisms fail (compare to __getattribute__) def __getattr__(self, name): self._verifyValidProperty(name) - return self._cargo[name] # just a simple lookup + return self._cargo[name] # just a simple lookup def __setattr__(self, name, value): - if name == '_cargo' : # reserved vars + if name == '_cargo': # reserved vars super().__setattr__(name, value) return self._verifyValidProperty(name) diff --git a/tests/system-test/0-others/user_manage.py b/tests/system-test/0-others/user_manage.py index 5148e26b39..6f90a2873a 100644 --- a/tests/system-test/0-others/user_manage.py +++ b/tests/system-test/0-others/user_manage.py @@ -12,12 +12,13 @@ # -*- coding: utf-8 -*- import taos -from util.log import * -from util.cases import * -from util.sql import * -from util.common import * -from util.sqlset import * from taos.tmq import * +from util.cases import * +from util.common import * +from util.log import * +from util.sql import * +from util.sqlset import * + class TDTestCase: def init(self, conn, logSql, replicaVar=1): @@ -26,10 +27,10 @@ class TDTestCase: tdSql.init(conn.cursor()) self.setsql = TDSetSql() self.stbname = 'stb' - self.binary_length = 20 # the length of binary for column_dict + self.binary_length = 20 # the length of binary for column_dict self.nchar_length = 20 # the length of nchar for column_dict self.column_dict = { - 'ts' : 'timestamp', + 'ts': 'timestamp', 'col1': 'tinyint', 'col2': 'smallint', 'col3': 'int', @@ -45,7 +46,7 @@ class TDTestCase: 'col13': f'nchar({self.nchar_length})' } self.tag_dict = { - 'ts_tag' : 'timestamp', + 'ts_tag': 'timestamp', 't1': 'tinyint', 't2': 'smallint', 't3': 'int', @@ -67,25 +68,28 @@ class TDTestCase: f'now,1,2,3,4,5,6,7,8,9.9,10.1,true,"abcd","涛思数据"' ] self.tbnum = 1 + def prepare_data(self): - tdSql.execute(self.setsql.set_create_stable_sql(self.stbname,self.column_dict,self.tag_dict)) + tdSql.execute(self.setsql.set_create_stable_sql(self.stbname, self.column_dict, self.tag_dict)) for i in range(self.tbnum): tdSql.execute(f'create table {self.stbname}_{i} using {self.stbname} tags({self.tag_list[i]})') for j in self.values_list: tdSql.execute(f'insert into {self.stbname}_{i} values({j})') + def create_user(self): - for user_name in ['jiacy1_all','jiacy1_read','jiacy1_write','jiacy1_none','jiacy0_all','jiacy0_read','jiacy0_write','jiacy0_none']: + for user_name in ['jiacy1_all', 'jiacy1_read', 'jiacy1_write', 'jiacy1_none', 'jiacy0_all', 'jiacy0_read', + 'jiacy0_write', 'jiacy0_none']: if 'jiacy1' in user_name.lower(): tdSql.execute(f'create user {user_name} pass "123" sysinfo 1') elif 'jiacy0' in user_name.lower(): tdSql.execute(f'create user {user_name} pass "123" sysinfo 0') - for user_name in ['jiacy1_all','jiacy1_read','jiacy0_all','jiacy0_read']: + for user_name in ['jiacy1_all', 'jiacy1_read', 'jiacy0_all', 'jiacy0_read']: tdSql.execute(f'grant read on db to {user_name}') - for user_name in ['jiacy1_all','jiacy1_write','jiacy0_all','jiacy0_write']: + for user_name in ['jiacy1_all', 'jiacy1_write', 'jiacy0_all', 'jiacy0_write']: tdSql.execute(f'grant write on db to {user_name}') def user_privilege_check(self): - jiacy1_read_conn = taos.connect(user='jiacy1_read',password='123') + jiacy1_read_conn = taos.connect(user='jiacy1_read', password='123') sql = "create table ntb (ts timestamp,c0 int)" expectErrNotOccured = True try: @@ -94,32 +98,34 @@ class TDTestCase: expectErrNotOccured = False if expectErrNotOccured: caller = inspect.getframeinfo(inspect.stack()[1][0]) - tdLog.exit(f"{caller.filename}({caller.lineno}) failed: sql:{sql}, expect error not occured" ) + tdLog.exit(f"{caller.filename}({caller.lineno}) failed: sql:{sql}, expect error not occured") else: self.queryRows = 0 self.queryCols = 0 self.queryResult = None tdLog.info(f"sql:{sql}, expect error occured") pass + def drop_topic(self): - jiacy1_all_conn = taos.connect(user='jiacy1_all',password='123') - jiacy1_read_conn = taos.connect(user='jiacy1_read',password='123') - jiacy1_write_conn = taos.connect(user='jiacy1_write',password='123') - jiacy1_none_conn = taos.connect(user='jiacy1_none',password='123') - jiacy0_all_conn = taos.connect(user='jiacy0_all',password='123') - jiacy0_read_conn = taos.connect(user='jiacy0_read',password='123') - jiacy0_write_conn = taos.connect(user='jiacy0_write',password='123') - jiacy0_none_conn = taos.connect(user='jiacy0_none',password='123') + jiacy1_all_conn = taos.connect(user='jiacy1_all', password='123') + jiacy1_read_conn = taos.connect(user='jiacy1_read', password='123') + jiacy1_write_conn = taos.connect(user='jiacy1_write', password='123') + jiacy1_none_conn = taos.connect(user='jiacy1_none', password='123') + jiacy0_all_conn = taos.connect(user='jiacy0_all', password='123') + jiacy0_read_conn = taos.connect(user='jiacy0_read', password='123') + jiacy0_write_conn = taos.connect(user='jiacy0_write', password='123') + jiacy0_none_conn = taos.connect(user='jiacy0_none', password='123') tdSql.execute('create topic root_db as select * from db.stb') - for user in [jiacy1_all_conn,jiacy1_read_conn,jiacy0_all_conn,jiacy0_read_conn]: + for user in [jiacy1_all_conn, jiacy1_read_conn, jiacy0_all_conn, jiacy0_read_conn]: user.execute(f'create topic db_jiacy as select * from db.stb') user.execute('drop topic db_jiacy') - for user in [jiacy1_write_conn,jiacy1_none_conn,jiacy0_write_conn,jiacy0_none_conn,jiacy1_all_conn,jiacy1_read_conn,jiacy0_all_conn,jiacy0_read_conn]: + for user in [jiacy1_write_conn, jiacy1_none_conn, jiacy0_write_conn, jiacy0_none_conn, jiacy1_all_conn, + jiacy1_read_conn, jiacy0_all_conn, jiacy0_read_conn]: sql_list = [] - if user in [jiacy1_all_conn,jiacy1_read_conn,jiacy0_all_conn,jiacy0_read_conn]: + if user in [jiacy1_all_conn, jiacy1_read_conn, jiacy0_all_conn, jiacy0_read_conn]: sql_list = ['drop topic root_db'] - elif user in [jiacy1_write_conn,jiacy1_none_conn,jiacy0_write_conn,jiacy0_none_conn]: - sql_list = ['drop topic root_db','create topic db_jiacy as select * from db.stb'] + elif user in [jiacy1_write_conn, jiacy1_none_conn, jiacy0_write_conn, jiacy0_none_conn]: + sql_list = ['drop topic root_db', 'create topic db_jiacy as select * from db.stb'] for sql in sql_list: expectErrNotOccured = True try: @@ -128,33 +134,26 @@ class TDTestCase: expectErrNotOccured = False if expectErrNotOccured: caller = inspect.getframeinfo(inspect.stack()[1][0]) - tdLog.exit(f"{caller.filename}({caller.lineno}) failed: sql:{sql}, expect error not occured" ) + tdLog.exit(f"{caller.filename}({caller.lineno}) failed: sql:{sql}, expect error not occured") else: self.queryRows = 0 self.queryCols = 0 self.queryResult = None tdLog.info(f"sql:{sql}, expect error occured") + def tmq_commit_cb_print(tmq, resp, param=None): print(f"commit: {resp}, tmq: {tmq}, param: {param}") + def subscribe_topic(self): print("create topic") tdSql.execute('create topic db_topic as select * from db.stb') tdSql.execute('grant subscribe on db_topic to jiacy1_all') print("build consumer") - conf = TaosTmqConf() - conf.set("group.id", "tg2") - conf.set("td.connect.user", "jiacy1_all") - conf.set("td.connect.pass", "123") - conf.set("enable.auto.commit", "true") - conf.set_auto_commit_cb(self.tmq_commit_cb_print, None) - tmq = conf.new_consumer() + tmq = Consumer({"group.id": "tg2", "td.connect.user": "jiacy1_all", "td.connect.pass": "123", + "enable.auto.commit": "true"}) print("build topic list") - topic_list = TaosTmqList() - topic_list.append("db_topic") + tmq.subscribe(["db_topic"]) print("basic consume loop") - tmq.subscribe(topic_list) - sub_list = tmq.subscription() - print("subscribed topics: ", sub_list) c = 0 l = 0 for i in range(10): @@ -163,20 +162,23 @@ class TDTestCase: res = tmq.poll(10) print(f"loop {l}") l += 1 - if res: - c += 1 - topic = res.get_topic_name() - vg = res.get_vgroup_id() - db = res.get_db_name() - print(f"topic: {topic}\nvgroup id: {vg}\ndb: {db}") - for row in res: - print(row) - print("* committed") - tmq.commit(res) - else: + if not res: print(f"received empty message at loop {l} (committed {c})") - pass - + continue + if res.error(): + print(f"consumer error at loop {l} (committed {c}) {res.error()}") + continue + + c += 1 + topic = res.topic() + db = res.database() + print(f"topic: {topic}\ndb: {db}") + + for row in res: + print(row.fetchall()) + print("* committed") + tmq.commit(res) + def run(self): tdSql.prepare() self.create_user() @@ -184,9 +186,11 @@ class TDTestCase: self.drop_topic() self.user_privilege_check() self.subscribe_topic() + def stop(self): tdSql.close() tdLog.success("%s successfully executed" % __file__) + tdCases.addWindows(__file__, TDTestCase()) -tdCases.addLinux(__file__, TDTestCase()) \ No newline at end of file +tdCases.addLinux(__file__, TDTestCase()) From 532fb0830908648293b0c0ab2ad3bdf7db2d23fe Mon Sep 17 00:00:00 2001 From: dapan1121 Date: Wed, 22 Feb 2023 18:13:24 +0800 Subject: [PATCH 30/31] fix: rsma invalid read issue --- source/dnode/vnode/src/sma/smaEnv.c | 10 +++++----- 1 file changed, 5 insertions(+), 5 deletions(-) diff --git a/source/dnode/vnode/src/sma/smaEnv.c b/source/dnode/vnode/src/sma/smaEnv.c index a272f5fc97..886f8d5c07 100644 --- a/source/dnode/vnode/src/sma/smaEnv.c +++ b/source/dnode/vnode/src/sma/smaEnv.c @@ -279,10 +279,7 @@ static void tdDestroyRSmaStat(void *pRSmaStat) { // step 1: set rsma trigger stat cancelled atomic_store_8(RSMA_TRIGGER_STAT(pStat), TASK_TRIGGER_STAT_CANCELLED); - // step 2: destroy the rsma info and associated fetch tasks - taosHashCleanup(RSMA_INFO_HASH(pStat)); - - // step 3: wait for all triggered fetch tasks to finish + // step 2: wait for all triggered fetch tasks to finish int32_t nLoops = 0; while (1) { if (T_REF_VAL_GET((SSmaStat *)pStat) == 0) { @@ -298,9 +295,12 @@ static void tdDestroyRSmaStat(void *pRSmaStat) { } } - // step 4: + // step 3: tdRsmaStopExecutor(pSma); + // step 4: destroy the rsma info and associated fetch tasks + taosHashCleanup(RSMA_INFO_HASH(pStat)); + // step 5: tdRSmaFSClose(RSMA_FS(pStat)); From 20b0766bc6a07dfc5aa02fd202f461d5273687f6 Mon Sep 17 00:00:00 2001 From: Shuduo Sang Date: Wed, 22 Feb 2023 19:54:12 +0800 Subject: [PATCH 31/31] fix: taosbenchmark data gen refactor main (#20060) * fix: taosbenchmark data generating refactor for main * fix: taosbenchmark sml json refactor for main * fix: update taos-tools 2942ca0 * fix: update taos-tools 61cbfd2 --- cmake/taostools_CMakeLists.txt.in | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/cmake/taostools_CMakeLists.txt.in b/cmake/taostools_CMakeLists.txt.in index 94ed46e5e2..ae3b626f88 100644 --- a/cmake/taostools_CMakeLists.txt.in +++ b/cmake/taostools_CMakeLists.txt.in @@ -2,7 +2,7 @@ # taos-tools ExternalProject_Add(taos-tools GIT_REPOSITORY https://github.com/taosdata/taos-tools.git - GIT_TAG 634399d + GIT_TAG 61cbfd2 SOURCE_DIR "${TD_SOURCE_DIR}/tools/taos-tools" BINARY_DIR "" #BUILD_IN_SOURCE TRUE