Merge pull request #2561 from taosdata/feature/query

Feature/query

src/client/src/tscLocal.c

@@ -406,7 +406,7 @@ int tscProcessLocalCmd(SSqlObj *pSql) {
     pSql->res.qhandle = 0x1;
     pSql->res.numOfRows = 0;
   } else if (pCmd->command == TSDB_SQL_RESET_CACHE) {
-    taosCacheEmpty(tscCacheHandle);
+    taosCacheEmpty(tscCacheHandle,false);
   } else if (pCmd->command == TSDB_SQL_SERV_VERSION) {
     tscProcessServerVer(pSql);
   } else if (pCmd->command == TSDB_SQL_CLI_VERSION) {

src/client/src/tscServer.c

@@ -1950,7 +1950,7 @@ int tscProcessUseDbRsp(SSqlObj *pSql) {
 }
 
 int tscProcessDropDbRsp(SSqlObj *UNUSED_PARAM(pSql)) {
-  taosCacheEmpty(tscCacheHandle);
+  taosCacheEmpty(tscCacheHandle, false);
   return 0;
 }
 
@@ -1996,7 +1996,7 @@ int tscProcessAlterTableMsgRsp(SSqlObj *pSql) {
 
     if (isSuperTable) {  // if it is a super table, reset whole query cache
       tscDebug("%p reset query cache since table:%s is stable", pSql, pTableMetaInfo->name);
-      taosCacheEmpty(tscCacheHandle);
+      taosCacheEmpty(tscCacheHandle, false);
     }
   }
 

src/dnode/src/dnodeVRead.c

@@ -98,11 +98,7 @@ void dnodeDispatchToVnodeReadQueue(SRpcMsg *pMsg) {
     pHead->vgId    = htonl(pHead->vgId);
     pHead->contLen = htonl(pHead->contLen);
 
-    if (pMsg->msgType == TSDB_MSG_TYPE_FETCH) {
-      pVnode = vnodeGetVnode(pHead->vgId);
-    } else {
     pVnode = vnodeAccquireVnode(pHead->vgId);
-    }
 
     if (pVnode == NULL) {
       leftLen -= pHead->contLen;
@@ -179,24 +175,17 @@ void dnodeFreeVnodeRqueue(void *rqueue) {
   // dynamically adjust the number of threads
 }
 
-static void dnodeContinueExecuteQuery(void* pVnode, void* qhandle, SReadMsg *pMsg) {
+void dnodePutQhandleIntoReadQueue(void *pVnode, void *qhandle) {
   SReadMsg *pRead = (SReadMsg *)taosAllocateQitem(sizeof(SReadMsg));
-  pRead->rpcMsg      = pMsg->rpcMsg;
+  pRead->rpcMsg.msgType = TSDB_MSG_TYPE_QUERY;
   pRead->pCont = qhandle;
   pRead->contLen = 0;
-  pRead->rpcMsg.msgType = TSDB_MSG_TYPE_QUERY;
 
-  taos_queue queue = vnodeGetRqueue(pVnode);
+  taos_queue queue = vnodeAccquireRqueue(pVnode);
-  taosWriteQitem(queue, TAOS_QTYPE_RPC, pRead);
+  taosWriteQitem(queue, TAOS_QTYPE_QUERY, pRead);
 }
 
 void dnodeSendRpcReadRsp(void *pVnode, SReadMsg *pRead, int32_t code) {
-  if (code == TSDB_CODE_VND_ACTION_IN_PROGRESS) return;
-  if (code == TSDB_CODE_VND_ACTION_NEED_REPROCESSED) {
-    dnodeContinueExecuteQuery(pVnode, pRead->rspRet.qhandle, pRead);
-    code = TSDB_CODE_SUCCESS;
-  }
-
   SRpcMsg rpcRsp = {
     .handle  = pRead->rpcMsg.handle,
     .pCont   = pRead->rspRet.rsp,
@@ -206,6 +195,12 @@ void dnodeSendRpcReadRsp(void *pVnode, SReadMsg *pRead, int32_t code) {
 
   rpcSendResponse(&rpcRsp);
   rpcFreeCont(pRead->rpcMsg.pCont);
+  vnodeRelease(pVnode);
+}
+
+void dnodeDispatchNonRspMsg(void *pVnode, SReadMsg *pRead, int32_t code) {
+  vnodeRelease(pVnode);
+  return;
 }
 
 static void *dnodeProcessReadQueue(void *param) {
@@ -219,9 +214,16 @@ static void *dnodeProcessReadQueue(void *param) {
       break;
     }
 
-    dDebug("%p, msg:%s will be processed in vread queue", pReadMsg->rpcMsg.ahandle, taosMsg[pReadMsg->rpcMsg.msgType]);
+    dDebug("%p, msg:%s will be processed in vread queue, qtype:%d", pReadMsg->rpcMsg.ahandle,
+           taosMsg[pReadMsg->rpcMsg.msgType], type);
     int32_t code = vnodeProcessRead(pVnode, pReadMsg);
+
+    if (type == TAOS_QTYPE_RPC) {
       dnodeSendRpcReadRsp(pVnode, pReadMsg, code);
+    } else {
+      dnodeDispatchNonRspMsg(pVnode, pReadMsg, code);
+    }
+
     taosFreeQitem(pReadMsg);
   }
 

src/inc/dnode.h

@@ -53,6 +53,7 @@ void *dnodeAllocateVnodeWqueue(void *pVnode);
 void  dnodeFreeVnodeWqueue(void *queue);
 void *dnodeAllocateVnodeRqueue(void *pVnode);
 void  dnodeFreeVnodeRqueue(void *rqueue);
+void  dnodePutQhandleIntoReadQueue(void *pVnode, void *qhandle);
 void  dnodeSendRpcVnodeWriteRsp(void *pVnode, void *param, int32_t code);
 
 int32_t dnodeAllocateMnodePqueue();

src/inc/query.h

@@ -84,6 +84,13 @@ bool qHasMoreResultsToRetrieve(qinfo_t qinfo);
  */
 int32_t qKillQuery(qinfo_t qinfo);
 
+void* qOpenQueryMgmt(int32_t vgId);
+void  qSetQueryMgmtClosed(void* pExecutor);
+void  qCleanupQueryMgmt(void* pExecutor);
+void** qRegisterQInfo(void* pMgmt, void* qInfo);
+void** qAcquireQInfo(void* pMgmt, void** key);
+void** qReleaseQInfo(void* pMgmt, void* pQInfo, bool needFree);
+
 #ifdef __cplusplus
 }
 #endif

src/inc/taosdef.h

@@ -365,6 +365,7 @@ void tsDataSwap(void *pLeft, void *pRight, int32_t type, int32_t size);
 #define TAOS_QTYPE_FWD      1
 #define TAOS_QTYPE_WAL      2 
 #define TAOS_QTYPE_CQ       3
+#define TAOS_QTYPE_QUERY    4
 
 typedef enum {
   TSDB_SUPER_TABLE        = 0,  // super table

src/inc/vnode.h

@@ -52,6 +52,7 @@ void    vnodeRelease(void *pVnode);
 void*   vnodeAccquireVnode(int32_t vgId); // add refcount 
 void*   vnodeGetVnode(int32_t vgId);      // keep refcount unchanged
 
+void*   vnodeAccquireRqueue(void *);
 void*   vnodeGetRqueue(void *);
 void*   vnodeGetWqueue(int32_t vgId);
 void*   vnodeGetWal(void *pVnode);

src/mnode/src/mnodeProfile.c

@@ -68,7 +68,7 @@ int32_t mnodeInitProfile() {
   mnodeAddWriteMsgHandle(TSDB_MSG_TYPE_CM_KILL_STREAM, mnodeProcessKillStreamMsg);
   mnodeAddWriteMsgHandle(TSDB_MSG_TYPE_CM_KILL_CONN, mnodeProcessKillConnectionMsg);
 
-  tsMnodeConnCache = taosCacheInitWithCb(TSDB_DATA_TYPE_INT, CONN_CHECK_TIME, false, mnodeFreeConn, "conn");
+  tsMnodeConnCache = taosCacheInit(TSDB_DATA_TYPE_INT, CONN_CHECK_TIME, false, mnodeFreeConn, "conn");
   return 0;
 }
 

src/mnode/src/mnodeShow.c

@@ -65,7 +65,7 @@ int32_t mnodeInitShow() {
   mnodeAddReadMsgHandle(TSDB_MSG_TYPE_CM_CONNECT, mnodeProcessConnectMsg);
   mnodeAddReadMsgHandle(TSDB_MSG_TYPE_CM_USE_DB, mnodeProcessUseMsg);
   
-  tsMnodeShowCache = taosCacheInitWithCb(TSDB_DATA_TYPE_INT, 5, false, mnodeFreeShowObj, "show");
+  tsMnodeShowCache = taosCacheInit(TSDB_DATA_TYPE_INT, 5, false, mnodeFreeShowObj, "show");
   return 0;
 }
 

src/plugins/http/src/httpContext.c

@@ -58,7 +58,7 @@ static void httpDestroyContext(void *data) {
 }
 
 bool httpInitContexts() {
-  tsHttpServer.contextCache = taosCacheInitWithCb(TSDB_DATA_TYPE_BIGINT, 2, false, httpDestroyContext, "restc");
+  tsHttpServer.contextCache = taosCacheInit(TSDB_DATA_TYPE_BIGINT, 2, false, httpDestroyContext, "restc");
   if (tsHttpServer.contextCache == NULL) {
     httpError("failed to init context cache");
     return false;

src/plugins/http/src/httpSession.c

@@ -115,7 +115,7 @@ void httpCleanUpSessions() {
 }
 
 bool httpInitSessions() {
-  tsHttpServer.sessionCache = taosCacheInitWithCb(TSDB_DATA_TYPE_BINARY, 5, false, httpDestroySession, "rests");
+  tsHttpServer.sessionCache = taosCacheInit(TSDB_DATA_TYPE_BINARY, 5, false, httpDestroySession, "rests");
   if (tsHttpServer.sessionCache == NULL) {
     httpError("failed to init session cache");
     return false;

src/query/src/qExecutor.c

@@ -13,8 +13,10 @@
  * along with this program. If not, see <http://www.gnu.org/licenses/>.
  */
 #include "os.h"
-#include "taosmsg.h"
+#include "tcache.h"
+#include "tglobal.h"
 #include "qfill.h"
+#include "taosmsg.h"
 
 #include "hash.h"
 #include "qExecutor.h"
@@ -87,16 +89,18 @@ typedef struct {
   STSCursor   cur;
 } SQueryStatusInfo;
 
+#if 0
 static UNUSED_FUNC void *u_malloc (size_t __size) {
-//  uint32_t v = rand();
+  uint32_t v = rand();
-//  if (v % 5 <= 1) {
+  if (v % 5 <= 1) {
-//    return NULL;
+    return NULL;
-//  } else {
+  } else {
     return malloc(__size);
-//  }
+  }
 }
 
 #define malloc  u_malloc
+#endif
 
 #define CLEAR_QUERY_STATUS(q, st)   ((q)->status &= (~(st)))
 #define GET_NUM_OF_TABLEGROUP(q)    taosArrayGetSize((q)->tableqinfoGroupInfo.pGroupList)
@@ -1520,7 +1524,6 @@ static void teardownQueryRuntimeEnv(SQueryRuntimeEnv *pRuntimeEnv) {
 }
 
 static bool isQueryKilled(SQInfo *pQInfo) {
-  return false;
   return (pQInfo->code == TSDB_CODE_TSC_QUERY_CANCELLED);
 }
 
@@ -5910,9 +5913,16 @@ static int32_t doDumpQueryResult(SQInfo *pQInfo, char *data) {
   return TSDB_CODE_SUCCESS;
 }
 
+typedef struct SQueryMgmt {
+  SCacheObj      *qinfoPool;      // query handle pool
+  int32_t         vgId;
+  bool            closed;
+  pthread_mutex_t lock;
+} SQueryMgmt;
+
 int32_t qCreateQueryInfo(void* tsdb, int32_t vgId, SQueryTableMsg* pQueryMsg, void* param, _qinfo_free_fn_t fn,
     qinfo_t* pQInfo) {
-  assert(pQueryMsg != NULL);
+  assert(pQueryMsg != NULL && tsdb != NULL);
 
   int32_t code = TSDB_CODE_SUCCESS;
 
@@ -6356,3 +6366,112 @@ static void buildTagQueryResult(SQInfo* pQInfo) {
   setQueryStatus(pQuery, QUERY_COMPLETED);
 }
 
+void freeqinfoFn(void *qhandle) {
+  void** handle = qhandle;
+  if (handle == NULL || *handle == NULL) {
+    return;
+  }
+
+  qKillQuery(*handle);
+}
+
+void* qOpenQueryMgmt(int32_t vgId) {
+  const int32_t REFRESH_HANDLE_INTERVAL = 2; // every 2 seconds, refresh handle pool
+
+  char cacheName[128] = {0};
+  sprintf(cacheName, "qhandle_%d", vgId);
+
+  SQueryMgmt* pQueryHandle = calloc(1, sizeof(SQueryMgmt));
+
+  pQueryHandle->qinfoPool = taosCacheInit(TSDB_DATA_TYPE_BIGINT, REFRESH_HANDLE_INTERVAL, true, freeqinfoFn, cacheName);
+  pQueryHandle->closed    = false;
+  pthread_mutex_init(&pQueryHandle->lock, NULL);
+
+  qDebug("vgId:%d, open querymgmt success", vgId);
+  return pQueryHandle;
+}
+
+void qSetQueryMgmtClosed(void* pQMgmt) {
+  if (pQMgmt == NULL) {
+    return;
+  }
+
+  SQueryMgmt* pQueryMgmt = pQMgmt;
+  qDebug("vgId:%d, set querymgmt closed, wait for all queries cancelled", pQueryMgmt->vgId);
+
+  pthread_mutex_lock(&pQueryMgmt->lock);
+  pQueryMgmt->closed = true;
+  pthread_mutex_unlock(&pQueryMgmt->lock);
+
+  taosCacheEmpty(pQueryMgmt->qinfoPool, true);
+}
+
+void qCleanupQueryMgmt(void* pQMgmt) {
+  if (pQMgmt == NULL) {
+    return;
+  }
+
+  SQueryMgmt* pQueryMgmt = pQMgmt;
+  int32_t vgId = pQueryMgmt->vgId;
+
+  assert(pQueryMgmt->closed);
+
+  SCacheObj* pqinfoPool = pQueryMgmt->qinfoPool;
+  pQueryMgmt->qinfoPool = NULL;
+
+  taosCacheCleanup(pqinfoPool);
+  pthread_mutex_destroy(&pQueryMgmt->lock);
+  tfree(pQueryMgmt);
+
+  qDebug("vgId:%d querymgmt cleanup completed", vgId);
+}
+
+void** qRegisterQInfo(void* pMgmt, void* qInfo) {
+  if (pMgmt == NULL) {
+    return NULL;
+  }
+
+  SQueryMgmt *pQueryMgmt = pMgmt;
+  if (pQueryMgmt->qinfoPool == NULL) {
+    return NULL;
+  }
+
+  pthread_mutex_lock(&pQueryMgmt->lock);
+  if (pQueryMgmt->closed) {
+    pthread_mutex_unlock(&pQueryMgmt->lock);
+
+    return NULL;
+  } else {
+    void** handle = taosCachePut(pQueryMgmt->qinfoPool, qInfo, POINTER_BYTES, &qInfo, POINTER_BYTES, tsShellActivityTimer*2);
+    pthread_mutex_unlock(&pQueryMgmt->lock);
+
+    return handle;
+  }
+}
+
+void** qAcquireQInfo(void* pMgmt, void** key) {
+  SQueryMgmt *pQueryMgmt = pMgmt;
+
+  if (pQueryMgmt->qinfoPool == NULL || pQueryMgmt->closed) {
+    return NULL;
+  }
+
+  void** handle = taosCacheAcquireByKey(pQueryMgmt->qinfoPool, key, POINTER_BYTES);
+  if (handle == NULL || *handle == NULL) {
+    return NULL;
+  } else {
+    return handle;
+  }
+}
+
+void** qReleaseQInfo(void* pMgmt, void* pQInfo, bool needFree) {
+  SQueryMgmt *pQueryMgmt = pMgmt;
+
+  if (pQueryMgmt->qinfoPool == NULL) {
+    return NULL;
+  }
+
+  taosCacheRelease(pQueryMgmt->qinfoPool, pQInfo, needFree);
+  return 0;
+}
+

src/util/inc/tcache.h

@@ -65,7 +65,7 @@ typedef struct {
   int64_t         totalSize;          // total allocated buffer in this hash table, SCacheObj is not included.
   int64_t         refreshTime;
   STrashElem *    pTrash;
-  const char *    cacheName;
+  char*           name;
 //  void *          tmrCtrl;
 //  void *          pTimer;
   SCacheStatis    statistics;
@@ -163,8 +163,9 @@ void taosCacheRelease(SCacheObj *pCacheObj, void **data, bool _remove);
 /**
  *  move all data node into trash, clear node in trash can if it is not referenced by any clients
  * @param handle
+ * @param _remove  remove the data or not if refcount is greater than 0
  */
-void taosCacheEmpty(SCacheObj *pCacheObj);
+void taosCacheEmpty(SCacheObj *pCacheObj, bool _remove);
 
 /**
  * release all allocated memory and destroy the cache object.

src/util/src/tcache.c

@@ -119,9 +119,8 @@ static FORCE_INLINE void taosCacheReleaseNode(SCacheObj *pCacheObj, SCacheDataNo
   int32_t size = pNode->size;
   taosHashRemove(pCacheObj->pHashTable, pNode->key, pNode->keySize);
 
-  uDebug("key:%p, %p is destroyed from cache, totalNum:%d totalSize:%" PRId64 "bytes size:%dbytes, cacheName:%s",
+  uDebug("cache:%s, key:%p, %p is destroyed from cache, totalNum:%d totalSize:%" PRId64 "bytes size:%dbytes",
-         pNode->key, pNode->data, (int32_t)taosHashGetSize(pCacheObj->pHashTable), pCacheObj->totalSize, size,
+         pCacheObj->name, pNode->key, pNode->data, (int32_t)taosHashGetSize(pCacheObj->pHashTable), pCacheObj->totalSize, size);
-         pCacheObj->cacheName);
   if (pCacheObj->freeFp) pCacheObj->freeFp(pNode->data);
   free(pNode);
 }
@@ -226,7 +225,7 @@ static void doCleanupDataCache(SCacheObj *pCacheObj);
  */
 static void* taosCacheRefresh(void *handle);
 
-SCacheObj *taosCacheInitWithCb(int32_t keyType, int64_t refreshTimeInSeconds, bool extendLifespan, __cache_freeres_fn_t fn, const char* cacheName) {
+SCacheObj *taosCacheInit(int32_t keyType, int64_t refreshTimeInSeconds, bool extendLifespan, __cache_freeres_fn_t fn, const char* cacheName) {
   if (refreshTimeInSeconds <= 0) {
     return NULL;
   }
@@ -238,7 +237,7 @@ SCacheObj *taosCacheInitWithCb(int32_t keyType, int64_t refreshTimeInSeconds, bo
   }
   
   pCacheObj->pHashTable = taosHashInit(128, taosGetDefaultHashFunction(keyType), false);
-  pCacheObj->cacheName = cacheName;
+  pCacheObj->name = strdup(cacheName);
   if (pCacheObj->pHashTable == NULL) {
     free(pCacheObj);
     uError("failed to allocate memory, reason:%s", strerror(errno));
@@ -268,10 +267,6 @@ SCacheObj *taosCacheInitWithCb(int32_t keyType, int64_t refreshTimeInSeconds, bo
   return pCacheObj;
 }
 
-SCacheObj *taosCacheInit(int32_t keyType, int64_t refreshTimeInSeconds, bool extendLifespan, __cache_freeres_fn_t fn, const char* cacheName) {
-  return taosCacheInitWithCb(keyType, refreshTimeInSeconds, extendLifespan, fn, cacheName);
-}
-
 void *taosCachePut(SCacheObj *pCacheObj, const void *key, size_t keyLen, const void *pData, size_t dataSize, int duration) {
   SCacheDataNode *pNode;
   
@@ -288,16 +283,16 @@ void *taosCachePut(SCacheObj *pCacheObj, const void *key, size_t keyLen, const v
     if (NULL != pNode) {
       pCacheObj->totalSize += pNode->size;
 
-      uDebug("key:%p, %p added into cache, added:%" PRIu64 ", expire:%" PRIu64 ", totalNum:%d totalSize:%" PRId64
+      uDebug("cache:%s, key:%p, %p added into cache, added:%" PRIu64 ", expire:%" PRIu64 ", totalNum:%d totalSize:%" PRId64
-             "bytes size:%" PRId64 "bytes, cacheName:%s",
+             "bytes size:%" PRId64 "bytes",
-             key, pNode->data, pNode->addedTime, (pNode->lifespan * pNode->extendFactor + pNode->addedTime),
+             pCacheObj->name, key, pNode->data, pNode->addedTime, (pNode->lifespan * pNode->extendFactor + pNode->addedTime),
-             (int32_t)taosHashGetSize(pCacheObj->pHashTable), pCacheObj->totalSize, dataSize, pCacheObj->cacheName);
+             (int32_t)taosHashGetSize(pCacheObj->pHashTable), pCacheObj->totalSize, dataSize);
     } else {
-      uError("key:%p, failed to added into cache, out of memory, cacheName:%s", key, pCacheObj->cacheName);
+      uError("cache:%s, key:%p, failed to added into cache, out of memory", pCacheObj->name, key);
     }
   } else {  // old data exists, update the node
     pNode = taosUpdateCacheImpl(pCacheObj, pOld, key, keyLen, pData, dataSize, duration * 1000L);
-    uDebug("key:%p, %p exist in cache, updated, cacheName:%s", key, pNode->data, pCacheObj->cacheName);
+    uDebug("cache:%s, key:%p, %p exist in cache, updated", pCacheObj->name, key, pNode->data);
   }
 
   __cache_unlock(pCacheObj);
@@ -332,10 +327,10 @@ void *taosCacheAcquireByKey(SCacheObj *pCacheObj, const void *key, size_t keyLen
   
   if (ptNode != NULL) {
     atomic_add_fetch_32(&pCacheObj->statistics.hitCount, 1);
-    uDebug("key:%p, %p is retrieved from cache, refcnt:%d, cacheName:%s", key, (*ptNode)->data, ref, pCacheObj->cacheName);
+    uDebug("cache:%s, key:%p, %p is retrieved from cache, refcnt:%d", pCacheObj->name, key, (*ptNode)->data, ref);
   } else {
     atomic_add_fetch_32(&pCacheObj->statistics.missCount, 1);
-    uDebug("key:%p, not in cache, retrieved failed, cacheName:%s", key, pCacheObj->cacheName);
+    uDebug("cache:%s, key:%p, not in cache, retrieved failed", pCacheObj->name, key);
   }
   
   atomic_add_fetch_32(&pCacheObj->statistics.totalAccess, 1);
@@ -360,11 +355,11 @@ void* taosCacheUpdateExpireTimeByName(SCacheObj *pCacheObj, void *key, size_t ke
 
   if (ptNode != NULL) {
     atomic_add_fetch_32(&pCacheObj->statistics.hitCount, 1);
-    uDebug("key:%p, %p expireTime is updated in cache, refcnt:%d, cacheName:%s", key, (*ptNode)->data,
+    uDebug("cache:%s, key:%p, %p expireTime is updated in cache, refcnt:%d", pCacheObj->name, key,
-           T_REF_VAL_GET(*ptNode), pCacheObj->cacheName);
+        (*ptNode)->data, T_REF_VAL_GET(*ptNode));
   } else {
     atomic_add_fetch_32(&pCacheObj->statistics.missCount, 1);
-    uDebug("key:%p, not in cache, retrieved failed, cacheName:%s", key, pCacheObj->cacheName);
+    uDebug("cache:%s, key:%p, not in cache, retrieved failed", pCacheObj->name, key);
   }
 
   atomic_add_fetch_32(&pCacheObj->statistics.totalAccess, 1);
@@ -383,7 +378,7 @@ void *taosCacheAcquireByData(SCacheObj *pCacheObj, void *data) {
   }
 
   int32_t ref = T_REF_INC(ptNode);
-  uDebug("%p acquired by data in cache, refcnt:%d, cacheName:%s", ptNode->data, ref, pCacheObj->cacheName);
+  uDebug("cache:%s, data: %p acquired by data in cache, refcnt:%d", pCacheObj->name, ptNode->data, ref);
 
   // if the remained life span is less then the (*ptNode)->lifeSpan, add up one lifespan
   if (pCacheObj->extendLifespan) {
@@ -391,7 +386,7 @@ void *taosCacheAcquireByData(SCacheObj *pCacheObj, void *data) {
 
     if ((now - ptNode->addedTime) < ptNode->lifespan * ptNode->extendFactor) {
       ptNode->extendFactor += 1;
-      uDebug("%p extend life time to %" PRId64, ptNode->data,
+      uDebug("cache:%s, %p extend life time to %" PRId64, pCacheObj->name, ptNode->data,
              ptNode->lifespan * ptNode->extendFactor + ptNode->addedTime);
     }
   }
@@ -437,7 +432,7 @@ void taosCacheRelease(SCacheObj *pCacheObj, void **data, bool _remove) {
 
   *data = NULL;
   int16_t ref = T_REF_DEC(pNode);
-  uDebug("key:%p, %p is released, refcnt:%d, cacheName:%s", pNode->key, pNode->data, ref, pCacheObj->cacheName);
+  uDebug("cache:%s, key:%p, %p is released, refcnt:%d", pCacheObj->name, pNode->key, pNode->data, ref);
 
   if (_remove && (!pNode->inTrashCan)) {
     __cache_wr_lock(pCacheObj);
@@ -455,7 +450,7 @@ void taosCacheRelease(SCacheObj *pCacheObj, void **data, bool _remove) {
   }
 }
 
-void taosCacheEmpty(SCacheObj *pCacheObj) {
+void taosCacheEmpty(SCacheObj *pCacheObj, bool _remove) {
   SHashMutableIterator *pIter = taosHashCreateIter(pCacheObj->pHashTable);
   
   __cache_wr_lock(pCacheObj);
@@ -465,12 +460,16 @@ void taosCacheEmpty(SCacheObj *pCacheObj) {
     }
     
     SCacheDataNode *pNode = *(SCacheDataNode **)taosHashIterGet(pIter);
+    if (T_REF_VAL_GET(pNode) == 0 || _remove) {
+      taosCacheReleaseNode(pCacheObj, pNode);
+    } else {
       taosCacheMoveToTrash(pCacheObj, pNode);
     }
+  }
   __cache_unlock(pCacheObj);
   
   taosHashDestroyIter(pIter);
-  taosTrashCanEmpty(pCacheObj, false);
+  taosTrashCanEmpty(pCacheObj, _remove);
 }
 
 void taosCacheCleanup(SCacheObj *pCacheObj) {
@@ -481,7 +480,7 @@ void taosCacheCleanup(SCacheObj *pCacheObj) {
   pCacheObj->deleting = 1;
   pthread_join(pCacheObj->refreshWorker, NULL);
 
-  uInfo("cacheName:%p, will be cleanuped", pCacheObj->cacheName);
+  uInfo("cache:%s will be cleaned up", pCacheObj->name);
   doCleanupDataCache(pCacheObj);
 }
 
@@ -601,22 +600,25 @@ void doCleanupDataCache(SCacheObj *pCacheObj) {
   SHashMutableIterator *pIter = taosHashCreateIter(pCacheObj->pHashTable);
   while (taosHashIterNext(pIter)) {
     SCacheDataNode *pNode = *(SCacheDataNode **)taosHashIterGet(pIter);
-    // if (pNode->expiredTime <= expiredTime && T_REF_VAL_GET(pNode) <= 0) {
+
-    if (T_REF_VAL_GET(pNode) <= 0) {
+    int32_t c = T_REF_VAL_GET(pNode);
+    if (c <= 0) {
       taosCacheReleaseNode(pCacheObj, pNode);
     } else {
-      uDebug("key:%p, %p will not remove from cache, refcnt:%d, cacheName:%s", pNode->key, pNode->data,
+      uDebug("cache:%s key:%p, %p will not remove from cache, refcnt:%d", pCacheObj->name, pNode->key,
-             T_REF_VAL_GET(pNode), pCacheObj->cacheName);
+          pNode->data, T_REF_VAL_GET(pNode));
     }
   }
   taosHashDestroyIter(pIter);
 
+  // todo memory leak if there are object with refcount greater than 0 in hash table?
   taosHashCleanup(pCacheObj->pHashTable);
   __cache_unlock(pCacheObj);
 
   taosTrashCanEmpty(pCacheObj, true);
   __cache_lock_destroy(pCacheObj);
   
+  tfree(pCacheObj->name);
   memset(pCacheObj, 0, sizeof(SCacheObj));
   free(pCacheObj);
 }

src/vnode/inc/vnodeInt.h

@@ -53,7 +53,7 @@ typedef struct {
   STsdbCfg     tsdbCfg;
   SSyncCfg     syncCfg;
   SWalCfg      walCfg;
-  void        *qHandlePool;  // query handle pool
+  void        *qMgmt;
   char        *rootDir;
   char         db[TSDB_DB_NAME_LEN];
 } SVnodeObj;

src/vnode/src/vnodeMain.c

@@ -46,7 +46,6 @@ static uint32_t vnodeGetFileInfo(void *ahandle, char *name, uint32_t *index, uin
 static int      vnodeGetWalInfo(void *ahandle, char *name, uint32_t *index);
 static void     vnodeNotifyRole(void *ahandle, int8_t role);
 static void     vnodeNotifyFileSynced(void *ahandle, uint64_t fversion);
-static void     vnodeFreeqHandle(void* phandle);
 
 static pthread_once_t  vnodeModuleInit = PTHREAD_ONCE_INIT;
 
@@ -283,9 +282,7 @@ int32_t vnodeOpen(int32_t vnode, char *rootDir) {
   if (pVnode->role == TAOS_SYNC_ROLE_MASTER)
     cqStart(pVnode->cq);
 
-  const int32_t REFRESH_HANDLE_INTERVAL = 2; // every 2 seconds, rfresh handle pool
+  pVnode->qMgmt = qOpenQueryMgmt(pVnode->vgId);
-  pVnode->qHandlePool = taosCacheInit(TSDB_DATA_TYPE_BIGINT, REFRESH_HANDLE_INTERVAL, true,  vnodeFreeqHandle, "qhandle");
-
   pVnode->events = NULL;
   pVnode->status = TAOS_VN_STATUS_READY;
   vDebug("vgId:%d, vnode is opened in %s, pVnode:%p", pVnode->vgId, rootDir, pVnode);
@@ -328,6 +325,9 @@ void vnodeRelease(void *pVnodeRaw) {
     return;
   }
 
+  qCleanupQueryMgmt(pVnode->qMgmt);
+  pVnode->qMgmt = NULL;
+
   if (pVnode->tsdb)
     tsdbCloseRepo(pVnode->tsdb, 1);
   pVnode->tsdb = NULL;
@@ -393,6 +393,15 @@ void *vnodeAccquireVnode(int32_t vgId) {
   return pVnode;
 }
 
+void *vnodeAccquireRqueue(void *param) {
+  SVnodeObj *pVnode = param;
+  if (pVnode == NULL) return NULL;
+
+  atomic_add_fetch_32(&pVnode->refCount, 1);
+  vDebug("vgId:%d, get vnode rqueue, refCount:%d", pVnode->vgId, pVnode->refCount);
+  return ((SVnodeObj *)pVnode)->rqueue;
+}
+
 void *vnodeGetRqueue(void *pVnode) {
   return ((SVnodeObj *)pVnode)->rqueue;
 }
@@ -466,7 +475,7 @@ static void vnodeCleanUp(SVnodeObj *pVnode) {
   vTrace("vgId:%d, vnode will cleanup, refCount:%d", pVnode->vgId, pVnode->refCount);
 
   // release local resources only after cutting off outside connections
-  taosCacheCleanup(pVnode->qHandlePool);
+  qSetQueryMgmtClosed(pVnode->qMgmt);
   vnodeRelease(pVnode);
 }
 
@@ -872,12 +881,3 @@ PARSE_OVER:
   if(fp) fclose(fp);
   return terrno;
 }
-
-void vnodeFreeqHandle(void *qHandle) {
-  void** handle = qHandle;
-  if (handle == NULL || *handle == NULL) {
-    return;
-  }
-
-  qKillQuery(*handle);
-}

src/vnode/src/vnodeRead.c

@@ -14,6 +14,7 @@
  */
 
 #define _DEFAULT_SOURCE
+#include <dnode.h>
 #include "os.h"
 
 #include "tglobal.h"
@@ -73,18 +74,22 @@ static int32_t vnodeProcessQueryMsg(SVnodeObj *pVnode, SReadMsg *pReadMsg) {
     killQueryMsg->free = htons(killQueryMsg->free);
     killQueryMsg->qhandle = htobe64(killQueryMsg->qhandle);
 
-    vWarn("QInfo:%p connection %p broken, kill query", (void*)killQueryMsg->qhandle, pReadMsg->rpcMsg.handle);
+    void* handle = NULL;
+    if ((void**) killQueryMsg->qhandle != NULL) {
+      handle = *(void**) killQueryMsg->qhandle;
+    }
+
+    vWarn("QInfo:%p connection %p broken, kill query", handle, pReadMsg->rpcMsg.handle);
     assert(pReadMsg->rpcMsg.contLen > 0 && killQueryMsg->free == 1);
 
-    // this message arrived here by means of the *query* message, so release the vnode is necessary
+    void** qhandle = qAcquireQInfo(pVnode->qMgmt, (void**) killQueryMsg->qhandle);
-    void** qhandle = taosCacheAcquireByKey(pVnode->qHandlePool, (void*) &killQueryMsg->qhandle, sizeof(killQueryMsg->qhandle));
     if (qhandle == NULL || *qhandle == NULL) {
       vWarn("QInfo:%p invalid qhandle, no matched query handle, conn:%p", (void*) killQueryMsg->qhandle, pReadMsg->rpcMsg.handle);
     } else {
-      taosCacheRelease(pVnode->qHandlePool, (void**) &qhandle, true);
+      assert(qhandle == (void**) killQueryMsg->qhandle);
+      qReleaseQInfo(pVnode->qMgmt, (void**) &qhandle, true);
     }
 
-    vnodeRelease(pVnode);
     return TSDB_CODE_TSC_QUERY_CANCELLED;
   }
 
@@ -93,7 +98,7 @@ static int32_t vnodeProcessQueryMsg(SVnodeObj *pVnode, SReadMsg *pReadMsg) {
   void**  handle = NULL;
 
   if (contLen != 0) {
-    code = qCreateQueryInfo(pVnode->tsdb, pVnode->vgId, pQueryTableMsg, pVnode, vnodeRelease, &pQInfo);
+    code = qCreateQueryInfo(pVnode->tsdb, pVnode->vgId, pQueryTableMsg, pVnode, NULL, &pQInfo);
 
     SQueryTableRsp *pRsp = (SQueryTableRsp *) rpcMallocCont(sizeof(SQueryTableRsp));
     pRsp->code    = code;
@@ -105,25 +110,30 @@ static int32_t vnodeProcessQueryMsg(SVnodeObj *pVnode, SReadMsg *pReadMsg) {
 
     // current connect is broken
     if (code == TSDB_CODE_SUCCESS) {
-      if (vnodeNotifyCurrentQhandle(pReadMsg->rpcMsg.handle, pQInfo, pVnode->vgId) != TSDB_CODE_SUCCESS) {
+      // add lock here
-        vError("vgId:%d, QInfo:%p, dnode query discarded since link is broken, %p", pVnode->vgId, pQInfo,
+      handle = qRegisterQInfo(pVnode->qMgmt, pQInfo);
-               pReadMsg->rpcMsg.handle);
+      if (handle == NULL) {  // failed to register qhandle
-        pRsp->code = TSDB_CODE_RPC_NETWORK_UNAVAIL;
+        pRsp->code = TSDB_CODE_QRY_INVALID_QHANDLE;
 
-        // NOTE: there two refcount, needs to kill twice, todo refactor
-        // query has not been put into qhandle pool, kill it directly.
         qKillQuery(pQInfo);
         qKillQuery(pQInfo);
-
+      } else {
-        return pRsp->code;
-      }
-
-      handle = taosCachePut(pVnode->qHandlePool, pQInfo, sizeof(pQInfo), &pQInfo, sizeof(pQInfo), tsShellActivityTimer * 2);
         assert(*handle == pQInfo);
         pRsp->qhandle = htobe64((uint64_t) (handle));
+      }
+
+      if (handle != NULL && vnodeNotifyCurrentQhandle(pReadMsg->rpcMsg.handle, handle, pVnode->vgId) != TSDB_CODE_SUCCESS) {
+        vError("vgId:%d, QInfo:%p, query discarded since link is broken, %p", pVnode->vgId, pQInfo, pReadMsg->rpcMsg.handle);
+        pRsp->code = TSDB_CODE_RPC_NETWORK_UNAVAIL;
+
+        // NOTE: there two refcount, needs to kill twice
+        // query has not been put into qhandle pool, kill it directly.
+        qKillQuery(pQInfo);
+        qReleaseQInfo(pVnode->qMgmt, (void**) &handle, true);
+        return pRsp->code;
+      }
     } else {
       assert(pQInfo == NULL);
-      vnodeRelease(pVnode);
     }
 
     vDebug("vgId:%d, QInfo:%p, dnode query msg disposed", vgId, pQInfo);
@@ -138,9 +148,8 @@ static int32_t vnodeProcessQueryMsg(SVnodeObj *pVnode, SReadMsg *pReadMsg) {
 
   if (pQInfo != NULL) {
     qTableQuery(pQInfo); // do execute query
-
     assert(handle != NULL);
-    taosCacheRelease(pVnode->qHandlePool, (void**) &handle, false);
+    qReleaseQInfo(pVnode->qMgmt, (void**) &handle, false);
   }
 
   return code;
@@ -159,7 +168,7 @@ static int32_t vnodeProcessFetchMsg(SVnodeObj *pVnode, SReadMsg *pReadMsg) {
   memset(pRet, 0, sizeof(SRspRet));
   int32_t ret = 0;
 
-  void** handle = taosCacheAcquireByKey(pVnode->qHandlePool, pQInfo, sizeof(pQInfo));
+  void** handle = qAcquireQInfo(pVnode->qMgmt, pQInfo);
   if (handle == NULL || handle != pQInfo) {
     ret = TSDB_CODE_QRY_INVALID_QHANDLE;
   }
@@ -167,8 +176,8 @@ static int32_t vnodeProcessFetchMsg(SVnodeObj *pVnode, SReadMsg *pReadMsg) {
   if (pRetrieve->free == 1) {
     if (ret == TSDB_CODE_SUCCESS) {
       vDebug("vgId:%d, QInfo:%p, retrieve msg received to kill query and free qhandle", pVnode->vgId, pQInfo);
+      qReleaseQInfo(pVnode->qMgmt, (void**) &handle, true);
 
-      taosCacheRelease(pVnode->qHandlePool, (void**) &handle, true);
       pRet->rsp = (SRetrieveTableRsp *)rpcMallocCont(sizeof(SRetrieveTableRsp));
       pRet->len = sizeof(SRetrieveTableRsp);
 
@@ -178,27 +187,27 @@ static int32_t vnodeProcessFetchMsg(SVnodeObj *pVnode, SReadMsg *pReadMsg) {
       pRsp->completed = true;
       pRsp->useconds  = 0;
     } else { // todo handle error
-
+      qReleaseQInfo(pVnode->qMgmt, (void**) &handle, true);
     }
     return ret;
   }
 
-  vDebug("vgId:%d, QInfo:%p, retrieve msg is received", pVnode->vgId, *pQInfo);
-
   int32_t code = qRetrieveQueryResultInfo(*pQInfo);
-  if (code != TSDB_CODE_SUCCESS) {
+  if (code != TSDB_CODE_SUCCESS || ret != TSDB_CODE_SUCCESS) {
     //TODO
     pRet->rsp = (SRetrieveTableRsp *)rpcMallocCont(sizeof(SRetrieveTableRsp));
     memset(pRet->rsp, 0, sizeof(SRetrieveTableRsp));
+
   } else {
     // todo check code and handle error in build result set
     code = qDumpRetrieveResult(*pQInfo, (SRetrieveTableRsp **)&pRet->rsp, &pRet->len);
 
-    if (qHasMoreResultsToRetrieve(*pQInfo)) {
+    if (qHasMoreResultsToRetrieve(*handle)) {
+      dnodePutQhandleIntoReadQueue(pVnode, handle);
       pRet->qhandle = handle;
-      code = TSDB_CODE_VND_ACTION_NEED_REPROCESSED;
+      code = TSDB_CODE_SUCCESS;
     } else { // no further execution invoked, release the ref to vnode
-      taosCacheRelease(pVnode->qHandlePool, (void**) &handle, true);
+      qReleaseQInfo(pVnode->qMgmt, (void**) &handle, true);
     }
   }