diff --git a/docs/en/29-roadmap/index.md b/docs/en/29-roadmap/index.md new file mode 100644 index 0000000000..e79565ff40 --- /dev/null +++ b/docs/en/29-roadmap/index.md @@ -0,0 +1,14 @@ +--- +title: Product Roadmap +--- + +The 2025 roadmap for TDengine OSS is described in the following table. + +| Quarter | Feature | +| :----- | :----- | +| 2025Q1 |
  1. Virtual tables
  2. Query engine: conditional expressions in REGEXP, GREATEST, LEAST, and CAST functions; improvements in single-row selection functions; time range interpolation with INTERP
  3. Storage engine: support for writing query results into supertables; KEEP parameter for supertables; performance improvements for the parameter binding interface
  4. Stream processing: support for virtual tables; decreased compute resource usage; new mechanism for event notification; faster stream creation
  5. Data types: support for the decimal data type
  6. High availability: faster recovery from downtime; improved client failover
  7. Stability: LTS release TDengine 3.3.6.x
  8. JDBC driver: more efficient data ingestion
  9. Ecosystem: integration with Microsoft Excel
| +| 2025Q2 |
  1. Query engine: relaxed restrictions on JOIN queries; support for all mathematical functions in MySQL; integral, integral average, and continuous variance functions; optimization of the CSUM function; support for COUNT(DISTINCT) syntax; enhancements to event windows; faster filtering by tag; faster INTERP queries
  2. Storage engine: decreased compute resource usage for TSMAs; improved write jitter
  3. Stream processing: high availability of snodes
  4. Data types: support for the blob data type
  5. Data subscription: support for the MQTT protocol
  6. High availability: faster replica configuration changes; faster recovery from downtime for clusters; improved data recovery after power outage
  7. Observability: diagnostic tool for data ingestion
| +| 2025Q3 |
  1. Query engine: more subqueries; support for all operators in MySQL; support for all time functions in MySQL; improved window calculation; reduced jitter in query performance; support for specifying columns in count windows
  2. Storage engine: faster ingestion in SQL mode
  3. Observability: diagnostic tool for queries; improved EXPLAIN output; monitoring of long-running tasks
| +| 2025Q4 |
  1. Query engine: window functions (i.e. the OVER clause); support for all string, aggregation, and conditional functions in MySQL; sorting within groups for partition queries; controls for query resource usage; faster aggregate queries on subtables; time range interpolation in INTERVAL windows
  2. Data types: support for variable-length strings
  3. Caching: faster row-oriented caching
  4. Observability: more insight into operations and maintenance
| + +For more information, see [TDengine Public Roadmap](https://github.com/orgs/taosdata/projects/4). diff --git a/docs/zh/05-basic/02-insert.md b/docs/zh/05-basic/02-insert.md index b129fdbff1..c5d3096aab 100644 --- a/docs/zh/05-basic/02-insert.md +++ b/docs/zh/05-basic/02-insert.md @@ -111,7 +111,7 @@ TDengine 还支持直接向超级表写入数据。需要注意的是，超级 ```sql insert into meters (tbname, ts, current, voltage, phase, location, group_id) -values( "d1001, "2018-10-03 14:38:05", 10.2, 220, 0.23, "California.SanFrancisco", 2) +values("d1001", "2018-10-03 14:38:05", 10.2, 220, 0.23, "California.SanFrancisco", 2) ``` ### 零代码写入 diff --git a/docs/zh/14-reference/03-taos-sql/02-database.md b/docs/zh/14-reference/03-taos-sql/02-database.md index 35cb99a3dd..3742b7c571 100644 --- a/docs/zh/14-reference/03-taos-sql/02-database.md +++ b/docs/zh/14-reference/03-taos-sql/02-database.md @@ -67,7 +67,7 @@ database_option: { - DURATION：数据文件存储数据的时间跨度。可以使用加单位的表示形式，如 DURATION 100h、DURATION 10d 等，支持 m（分钟）、h（小时）和 d（天）三个单位。不加时间单位时默认单位为天，如 DURATION 50 表示 50 天。 - MAXROWS：文件块中记录的最大条数，默认为 4096 条。 - MINROWS：文件块中记录的最小条数，默认为 100 条。 -- KEEP：表示数据文件保存的天数，缺省值为 3650，取值范围 [1, 365000]，且必须大于或等于 3 倍的 DURATION 参数值。数据库会自动删除保存时间超过 KEEP 值的数据从而释放存储空间。KEEP 可以使用加单位的表示形式，如 KEEP 100h、KEEP 10d 等，支持 m（分钟）、h（小时）和 d（天）三个单位。也可以不写单位，如 KEEP 50，此时默认单位为天。企业版支持[多级存储](../../operation/planning/#%E5%A4%9A%E7%BA%A7%E5%AD%98%E5%82%A8)功能, 因此, 可以设置多个保存时间（多个以英文逗号分隔，最多 3 个，满足 keep 0 \<= keep 1 \<= keep 2，如 KEEP 100h,100d,3650d）; 社区版不支持多级存储功能（即使配置了多个保存时间, 也不会生效, KEEP 会取最大的保存时间）。了解更多，请点击 [关于主键时间戳](https://docs.taosdata.com/reference/taos-sql/insert/) +- KEEP：表示数据文件保存的天数，缺省值为 3650，取值范围 [1, 365000]，且必须大于或等于 3 倍的 DURATION 参数值。数据库会自动删除保存时间超过 KEEP 值的数据从而释放存储空间。KEEP 可以使用加单位的表示形式，如 KEEP 100h、KEEP 10d 等，支持 m（分钟）、h（小时）和 d（天）三个单位。也可以不写单位，如 KEEP 50，此时默认单位为天。企业版支持[多级存储](https://docs.taosdata.com/operation/planning/#%E5%A4%9A%E7%BA%A7%E5%AD%98%E5%82%A8)功能, 因此, 可以设置多个保存时间（多个以英文逗号分隔，最多 3 个，满足 keep 0 \<= keep 1 \<= keep 2，如 KEEP 100h,100d,3650d）; 社区版不支持多级存储功能（即使配置了多个保存时间, 也不会生效, KEEP 会取最大的保存时间）。了解更多，请点击 [关于主键时间戳](https://docs.taosdata.com/reference/taos-sql/insert/) - KEEP_TIME_OFFSET：自 3.2.0.0 版本生效。删除或迁移保存时间超过 KEEP 值的数据的延迟执行时间，默认值为 0 (小时)。在数据文件保存时间超过 KEEP 后，删除或迁移操作不会立即执行，而会额外等待本参数指定的时间间隔，以实现与业务高峰期错开的目的。 - STT_TRIGGER：表示落盘文件触发文件合并的个数。对于少表高频写入场景，此参数建议使用默认配置；而对于多表低频写入场景，此参数建议配置较大的值。 diff --git a/docs/zh/29-roadmap/index.md b/docs/zh/29-roadmap/index.md new file mode 100644 index 0000000000..a68e97a410 --- /dev/null +++ b/docs/zh/29-roadmap/index.md @@ -0,0 +1,14 @@ +--- +title: 产品路线图 +--- + +TDengine OSS 之 2025 年年度路线图如下表所示。 + +| 季度 | 功能 | +| :----- | :----- | +| 2025Q1 |
  1. 虚拟表
  2. 查询能力：REGEXP、GREATEST、LEAST、CAST 函数支持判断表达式、单行选择函数的其他列值、INTERP 支持插值时间范围
  3. 存储能力：支持将查询结果写入超级表、超级表支持 KEEP 参数、STMT 写入性能提升
  4. 流计算：支持虚拟表、计算资源优化、事件通知机制、创建时间优化
  5. 数据类型：Decimal
  6. 高可用：加快宕机恢复时间、优化客户端 Failover 机制
  7. 稳定性：开始维护新的稳定版本 3.3.6.x
  8. JDBC：高效写入
  9. 生态工具：对接 Tableau
  10. 生态工具：对接 Excel
| +| 2025Q2 |
  1. 查询能力：大幅放宽关联查询限制、支持 MySQL 所有数学函数、支持积分/积分平均/连续方差函数、CSUM 函数优化、COUNT(DISTINCT) 语法、事件窗口功能增强、提升标签过滤性能、提升 INTERP 查询性能
  2. 存储能力：TSMA 计算资源优化、写入抖动优化
  3. 流计算：节点高可用
  4. 数据类型：BLOB
  5. 数据订阅：支持 MQTT 协议
  6. 高可用：提高副本变更速度、提高集群宕机恢复速度、优化断电数据恢复机制
  7. 可观测性：写入诊断工具
  8. 生态工具：对接帆软 FineBI
| +| 2025Q3 |
  1. 查询能力：支持更多子查询类型、支持 MySQL 运算符、支持 MySQL 所有时间函数、窗口计算逻辑优化、查询性能抖动、计数窗口允许指定列
  2. 存储能力：提高 SQL 模式写入速度
  3. 可观测性：查询诊断工具、优化 EXPLAIN 输出、长任务观测
| +| 2025Q4 |
  1. 查询能力：窗口函数（OVER 子句）、支持 MySQL 所有字符串/聚合/条件函数、Partition 支持组内排序、控制查询资源占用、提高子表聚合查询性能、INTERVAL 窗口支持插值时间范围
  2. 数据类型：支持不定长度字符串数据类型
  3. 数据缓存：提升按行缓存性能
  4. 可观测性：增强运维可观测性
| + +欲了解更多信息，请参见 [TDengine Public Roadmap](https://github.com/orgs/taosdata/projects/4) 。 diff --git a/include/common/tmsgcb.h b/include/common/tmsgcb.h index c934cb6961..2847f4278a 100644 --- a/include/common/tmsgcb.h +++ b/include/common/tmsgcb.h @@ -38,6 +38,7 @@ typedef enum { STREAM_QUEUE, ARB_QUEUE, STREAM_CTRL_QUEUE, + STREAM_LONG_EXEC_QUEUE, QUEUE_MAX, } EQueueType; diff --git a/include/libs/executor/executor.h b/include/libs/executor/executor.h index 883c5f7b99..988a650d6c 100644 --- a/include/libs/executor/executor.h +++ b/include/libs/executor/executor.h @@ -180,7 +180,7 @@ void qCleanExecTaskBlockBuf(qTaskInfo_t tinfo); */ int32_t qAsyncKillTask(qTaskInfo_t tinfo, int32_t rspCode); -int32_t qKillTask(qTaskInfo_t tinfo, int32_t rspCode); +int32_t qKillTask(qTaskInfo_t tinfo, int32_t rspCode, int64_t waitDuration); bool qTaskIsExecuting(qTaskInfo_t qinfo); diff --git a/include/libs/stream/tstream.h b/include/libs/stream/tstream.h index a4d89dcdcc..c4c0aaf742 100644 --- a/include/libs/stream/tstream.h +++ b/include/libs/stream/tstream.h @@ -58,6 +58,7 @@ extern "C" { #define STREAM_EXEC_T_STOP_ALL_TASKS (-5) #define STREAM_EXEC_T_RESUME_TASK (-6) #define STREAM_EXEC_T_ADD_FAILED_TASK (-7) +#define STREAM_EXEC_T_STOP_ONE_TASK (-8) typedef struct SStreamTask SStreamTask; typedef struct SStreamQueue SStreamQueue; @@ -483,8 +484,10 @@ typedef struct STaskUpdateInfo { } STaskUpdateInfo; typedef struct SScanWalInfo { - int32_t scanCounter; + int32_t scanSentinel; tmr_h scanTimer; + int64_t lastScanTs; + int32_t tickCounter; } SScanWalInfo; typedef struct SFatalErrInfo { @@ -752,15 +755,19 @@ void streamMetaCleanup(); int32_t streamMetaOpen(const char* path, void* ahandle, FTaskBuild expandFunc, FTaskExpand expandTaskFn, int32_t vgId, int64_t stage, startComplete_fn_t fn, SStreamMeta** pMeta); void streamMetaClose(SStreamMeta* streamMeta); -int32_t streamMetaSaveTask(SStreamMeta* pMeta, SStreamTask* pTask); // save to stream meta store -int32_t streamMetaRemoveTask(SStreamMeta* pMeta, STaskId* pKey); + +int32_t streamMetaSaveTaskInMeta(SStreamMeta* pMeta, SStreamTask* pTask); // save to stream meta store +int32_t streamMetaRemoveTaskInMeta(SStreamMeta* pMeta, STaskId* pKey); + int32_t streamMetaRegisterTask(SStreamMeta* pMeta, int64_t ver, SStreamTask* pTask, bool* pAdded); int32_t streamMetaUnregisterTask(SStreamMeta* pMeta, int64_t streamId, int32_t taskId); int32_t streamMetaGetNumOfTasks(SStreamMeta* pMeta); + int32_t streamMetaAcquireTaskNoLock(SStreamMeta* pMeta, int64_t streamId, int32_t taskId, SStreamTask** pTask); int32_t streamMetaAcquireTaskUnsafe(SStreamMeta* pMeta, STaskId* pId, SStreamTask** pTask); int32_t streamMetaAcquireTask(SStreamMeta* pMeta, int64_t streamId, int32_t taskId, SStreamTask** pTask); void streamMetaReleaseTask(SStreamMeta* pMeta, SStreamTask* pTask); + void streamMetaClear(SStreamMeta* pMeta); void streamMetaInitBackend(SStreamMeta* pMeta); int32_t streamMetaCommit(SStreamMeta* pMeta); @@ -794,6 +801,7 @@ void streamMetaLoadAllTasks(SStreamMeta* pMeta); int32_t streamMetaStartAllTasks(SStreamMeta* pMeta); int32_t streamMetaStopAllTasks(SStreamMeta* pMeta); int32_t streamMetaStartOneTask(SStreamMeta* pMeta, int64_t streamId, int32_t taskId); +int32_t streamMetaStopOneTask(SStreamMeta* pMeta, int64_t streamId, int32_t taskId); bool streamMetaAllTasksReady(const SStreamMeta* pMeta); int32_t streamTaskSendNegotiateChkptIdMsg(SStreamTask* pTask); int32_t streamTaskCheckIfReqConsenChkptId(SStreamTask* pTask, int64_t ts); diff --git a/include/util/tworker.h b/include/util/tworker.h index a3ba7dba6d..bc0dde1a37 100644 --- a/include/util/tworker.h +++ b/include/util/tworker.h @@ -76,7 +76,7 @@ void tQWorkerFreeQueue(SQWorkerPool *pool, STaosQueue *queue); int32_t tAutoQWorkerInit(SAutoQWorkerPool *pool); void tAutoQWorkerCleanup(SAutoQWorkerPool *pool); -STaosQueue *tAutoQWorkerAllocQueue(SAutoQWorkerPool *pool, void *ahandle, FItem fp); +STaosQueue *tAutoQWorkerAllocQueue(SAutoQWorkerPool *pool, void *ahandle, FItem fp, int32_t minNum); void tAutoQWorkerFreeQueue(SAutoQWorkerPool *pool, STaosQueue *queue); int32_t tWWorkerInit(SWWorkerPool *pool); diff --git a/source/dnode/mgmt/mgmt_vnode/inc/vmInt.h b/source/dnode/mgmt/mgmt_vnode/inc/vmInt.h index 84f5149624..9b4c11d6ae 100644 --- a/source/dnode/mgmt/mgmt_vnode/inc/vmInt.h +++ b/source/dnode/mgmt/mgmt_vnode/inc/vmInt.h @@ -32,6 +32,7 @@ typedef struct SVnodeMgmt { const char *name; SQueryAutoQWorkerPool queryPool; SAutoQWorkerPool streamPool; + SAutoQWorkerPool streamLongExecPool; SWWorkerPool streamCtrlPool; SWWorkerPool fetchPool; SSingleWorker mgmtWorker; @@ -75,6 +76,7 @@ typedef struct { STaosQueue *pQueryQ; STaosQueue *pStreamQ; STaosQueue *pStreamCtrlQ; + STaosQueue *pStreamLongExecQ; STaosQueue *pFetchQ; STaosQueue *pMultiMgmQ; } SVnodeObj; @@ -137,6 +139,8 @@ int32_t vmPutMsgToQueryQueue(SVnodeMgmt *pMgmt, SRpcMsg *pMsg); int32_t vmPutMsgToFetchQueue(SVnodeMgmt *pMgmt, SRpcMsg *pMsg); int32_t vmPutMsgToStreamQueue(SVnodeMgmt *pMgmt, SRpcMsg *pMsg); int32_t vmPutMsgToStreamCtrlQueue(SVnodeMgmt *pMgmt, SRpcMsg *pMsg); +int32_t vmPutMsgToStreamLongExecQueue(SVnodeMgmt *pMgmt, SRpcMsg *pMsg); + int32_t vmPutMsgToMergeQueue(SVnodeMgmt *pMgmt, SRpcMsg *pMsg); int32_t vmPutMsgToMgmtQueue(SVnodeMgmt *pMgmt, SRpcMsg *pMsg); int32_t vmPutMsgToMultiMgmtQueue(SVnodeMgmt *pMgmt, SRpcMsg *pMsg); diff --git a/source/dnode/mgmt/mgmt_vnode/src/vmHandle.c b/source/dnode/mgmt/mgmt_vnode/src/vmHandle.c index 234d4f41e1..1dea7d3cad 100644 --- a/source/dnode/mgmt/mgmt_vnode/src/vmHandle.c +++ b/source/dnode/mgmt/mgmt_vnode/src/vmHandle.c @@ -1008,27 +1008,29 @@ SArray *vmGetMsgHandles() { if (dmSetMgmtHandle(pArray, TDMT_STREAM_TASK_RUN, vmPutMsgToStreamQueue, 0) == NULL) goto _OVER; if (dmSetMgmtHandle(pArray, TDMT_STREAM_TASK_DISPATCH, vmPutMsgToStreamCtrlQueue, 0) == NULL) goto _OVER; if (dmSetMgmtHandle(pArray, TDMT_STREAM_TASK_DISPATCH_RSP, vmPutMsgToStreamCtrlQueue, 0) == NULL) goto _OVER; - if (dmSetMgmtHandle(pArray, TDMT_STREAM_RETRIEVE, vmPutMsgToStreamQueue, 0) == NULL) goto _OVER; - if (dmSetMgmtHandle(pArray, TDMT_STREAM_RETRIEVE_RSP, vmPutMsgToStreamQueue, 0) == NULL) goto _OVER; if (dmSetMgmtHandle(pArray, TDMT_VND_STREAM_TASK_CHECK, vmPutMsgToStreamCtrlQueue, 0) == NULL) goto _OVER; if (dmSetMgmtHandle(pArray, TDMT_VND_STREAM_TASK_CHECK_RSP, vmPutMsgToStreamCtrlQueue, 0) == NULL) goto _OVER; - if (dmSetMgmtHandle(pArray, TDMT_STREAM_TASK_PAUSE, vmPutMsgToWriteQueue, 0) == NULL) goto _OVER; - if (dmSetMgmtHandle(pArray, TDMT_STREAM_TASK_RESUME, vmPutMsgToWriteQueue, 0) == NULL) goto _OVER; - if (dmSetMgmtHandle(pArray, TDMT_STREAM_TASK_STOP, vmPutMsgToWriteQueue, 0) == NULL) goto _OVER; - if (dmSetMgmtHandle(pArray, TDMT_VND_STREAM_CHECK_POINT_SOURCE, vmPutMsgToWriteQueue, 0) == NULL) goto _OVER; if (dmSetMgmtHandle(pArray, TDMT_STREAM_TASK_CHECKPOINT_READY, vmPutMsgToStreamCtrlQueue, 0) == NULL) goto _OVER; if (dmSetMgmtHandle(pArray, TDMT_STREAM_TASK_CHECKPOINT_READY_RSP, vmPutMsgToStreamCtrlQueue, 0) == NULL) goto _OVER; if (dmSetMgmtHandle(pArray, TDMT_STREAM_RETRIEVE_TRIGGER, vmPutMsgToStreamCtrlQueue, 0) == NULL) goto _OVER; if (dmSetMgmtHandle(pArray, TDMT_STREAM_RETRIEVE_TRIGGER_RSP, vmPutMsgToStreamCtrlQueue, 0) == NULL) goto _OVER; - if (dmSetMgmtHandle(pArray, TDMT_VND_STREAM_TASK_UPDATE, vmPutMsgToWriteQueue, 0) == NULL) goto _OVER; - if (dmSetMgmtHandle(pArray, TDMT_VND_STREAM_TASK_RESET, vmPutMsgToWriteQueue, 0) == NULL) goto _OVER; if (dmSetMgmtHandle(pArray, TDMT_MND_STREAM_HEARTBEAT_RSP, vmPutMsgToStreamCtrlQueue, 0) == NULL) goto _OVER; if (dmSetMgmtHandle(pArray, TDMT_MND_STREAM_REQ_CHKPT_RSP, vmPutMsgToStreamCtrlQueue, 0) == NULL) goto _OVER; if (dmSetMgmtHandle(pArray, TDMT_MND_STREAM_CHKPT_REPORT_RSP, vmPutMsgToStreamCtrlQueue, 0) == NULL) goto _OVER; + if (dmSetMgmtHandle(pArray, TDMT_VND_STREAM_SCAN_HISTORY, vmPutMsgToStreamLongExecQueue, 0) == NULL) goto _OVER; + if (dmSetMgmtHandle(pArray, TDMT_VND_GET_STREAM_PROGRESS, vmPutMsgToStreamQueue, 0) == NULL) goto _OVER; + if (dmSetMgmtHandle(pArray, TDMT_STREAM_RETRIEVE, vmPutMsgToStreamQueue, 0) == NULL) goto _OVER; + if (dmSetMgmtHandle(pArray, TDMT_STREAM_RETRIEVE_RSP, vmPutMsgToStreamQueue, 0) == NULL) goto _OVER; if (dmSetMgmtHandle(pArray, TDMT_STREAM_TASK_UPDATE_CHKPT, vmPutMsgToWriteQueue, 0) == NULL) goto _OVER; if (dmSetMgmtHandle(pArray, TDMT_STREAM_CONSEN_CHKPT, vmPutMsgToWriteQueue, 0) == NULL) goto _OVER; + if (dmSetMgmtHandle(pArray, TDMT_STREAM_TASK_PAUSE, vmPutMsgToWriteQueue, 0) == NULL) goto _OVER; + if (dmSetMgmtHandle(pArray, TDMT_STREAM_TASK_RESUME, vmPutMsgToWriteQueue, 0) == NULL) goto _OVER; + if (dmSetMgmtHandle(pArray, TDMT_STREAM_TASK_STOP, vmPutMsgToWriteQueue, 0) == NULL) goto _OVER; + if (dmSetMgmtHandle(pArray, TDMT_VND_STREAM_CHECK_POINT_SOURCE, vmPutMsgToWriteQueue, 0) == NULL) goto _OVER; + if (dmSetMgmtHandle(pArray, TDMT_VND_STREAM_TASK_UPDATE, vmPutMsgToWriteQueue, 0) == NULL) goto _OVER; + if (dmSetMgmtHandle(pArray, TDMT_VND_STREAM_TASK_RESET, vmPutMsgToWriteQueue, 0) == NULL) goto _OVER; if (dmSetMgmtHandle(pArray, TDMT_VND_ALTER_REPLICA, vmPutMsgToMgmtQueue, 0) == NULL) goto _OVER; if (dmSetMgmtHandle(pArray, TDMT_VND_ALTER_CONFIG, vmPutMsgToWriteQueue, 0) == NULL) goto _OVER; diff --git a/source/dnode/mgmt/mgmt_vnode/src/vmInt.c b/source/dnode/mgmt/mgmt_vnode/src/vmInt.c index d71e0b02c4..6f30977e10 100644 --- a/source/dnode/mgmt/mgmt_vnode/src/vmInt.c +++ b/source/dnode/mgmt/mgmt_vnode/src/vmInt.c @@ -398,10 +398,14 @@ void vmCloseVnode(SVnodeMgmt *pMgmt, SVnodeObj *pVnode, bool commitAndRemoveWal, dInfo("vgId:%d, wait for vnode stream queue:%p is empty, %d remains", pVnode->vgId, pVnode->pStreamQ, taosQueueItemSize(pVnode->pStreamQ)); - while (!taosQueueEmpty(pVnode->pStreamQ)) taosMsleep(10); + while (!taosQueueEmpty(pVnode->pStreamQ)) taosMsleep(50); dInfo("vgId:%d, wait for vnode stream ctrl queue:%p is empty", pVnode->vgId, pVnode->pStreamCtrlQ); - while (!taosQueueEmpty(pVnode->pStreamCtrlQ)) taosMsleep(10); + while (!taosQueueEmpty(pVnode->pStreamCtrlQ)) taosMsleep(50); + + dInfo("vgId:%d, wait for vnode stream long-exec queue:%p is empty, %d remains", pVnode->vgId, + pVnode->pStreamLongExecQ, taosQueueItemSize(pVnode->pStreamLongExecQ)); + while (!taosQueueEmpty(pVnode->pStreamLongExecQ)) taosMsleep(50); dInfo("vgId:%d, all vnode queues is empty", pVnode->vgId); diff --git a/source/dnode/mgmt/mgmt_vnode/src/vmWorker.c b/source/dnode/mgmt/mgmt_vnode/src/vmWorker.c index b398bdf242..5acd06bbda 100644 --- a/source/dnode/mgmt/mgmt_vnode/src/vmWorker.c +++ b/source/dnode/mgmt/mgmt_vnode/src/vmWorker.c @@ -150,7 +150,7 @@ static void vmProcessStreamCtrlQueue(SQueueInfo *pInfo, STaosQall* pQall, int32_ SRpcMsg *pMsg = pItem; const STraceId *trace = &pMsg->info.traceId; - dGTrace("vgId:%d, msg:%p get from vnode-ctrl-stream queue", pVnode->vgId, pMsg); + dGTrace("vgId:%d, msg:%p get from vnode-stream-ctrl queue", pVnode->vgId, pMsg); code = vnodeProcessStreamCtrlMsg(pVnode->pImpl, pMsg, pInfo); if (code != 0) { terrno = code; @@ -165,6 +165,26 @@ static void vmProcessStreamCtrlQueue(SQueueInfo *pInfo, STaosQall* pQall, int32_ } } +static void vmProcessStreamLongExecQueue(SQueueInfo *pInfo, SRpcMsg *pMsg) { + SVnodeObj *pVnode = pInfo->ahandle; + const STraceId *trace = &pMsg->info.traceId; + int32_t code = 0; + + dGTrace("vgId:%d, msg:%p get from vnode-stream long-exec queue", pVnode->vgId, pMsg); + + code = vnodeProcessStreamLongExecMsg(pVnode->pImpl, pMsg, pInfo); + if (code != 0) { + terrno = code; + dGError("vgId:%d, msg:%p failed to process stream msg %s since %s", pVnode->vgId, pMsg, TMSG_INFO(pMsg->msgType), + tstrerror(code)); + vmSendRsp(pMsg, code); + } + + dGTrace("vgId:%d, msg:%p is freed, code:0x%x", pVnode->vgId, pMsg, code); + rpcFreeCont(pMsg->pCont); + taosFreeQitem(pMsg); +} + static void vmProcessFetchQueue(SQueueInfo *pInfo, STaosQall *qall, int32_t numOfMsgs) { SVnodeObj *pVnode = pInfo->ahandle; SRpcMsg *pMsg = NULL; @@ -274,9 +294,13 @@ static int32_t vmPutMsgToQueue(SVnodeMgmt *pMgmt, SRpcMsg *pMsg, EQueueType qtyp code = taosWriteQitem(pVnode->pStreamQ, pMsg); break; case STREAM_CTRL_QUEUE: - dGTrace("vgId:%d, msg:%p put into vnode-ctrl-stream queue", pVnode->vgId, pMsg); + dGTrace("vgId:%d, msg:%p put into vnode-stream-ctrl queue", pVnode->vgId, pMsg); code = taosWriteQitem(pVnode->pStreamCtrlQ, pMsg); break; + case STREAM_LONG_EXEC_QUEUE: + dGTrace("vgId:%d, msg:%p put into vnode-stream-long-exec queue", pVnode->vgId, pMsg); + code = taosWriteQitem(pVnode->pStreamLongExecQ, pMsg); + break; case FETCH_QUEUE: dGTrace("vgId:%d, msg:%p put into vnode-fetch queue", pVnode->vgId, pMsg); code = taosWriteQitem(pVnode->pFetchQ, pMsg); @@ -335,6 +359,8 @@ int32_t vmPutMsgToStreamQueue(SVnodeMgmt *pMgmt, SRpcMsg *pMsg) { return vmPutMs int32_t vmPutMsgToStreamCtrlQueue(SVnodeMgmt *pMgmt, SRpcMsg *pMsg) { return vmPutMsgToQueue(pMgmt, pMsg, STREAM_CTRL_QUEUE); } +int32_t vmPutMsgToStreamLongExecQueue(SVnodeMgmt *pMgmt, SRpcMsg *pMsg) { return vmPutMsgToQueue(pMgmt, pMsg, STREAM_LONG_EXEC_QUEUE); } + int32_t vmPutMsgToMultiMgmtQueue(SVnodeMgmt *pMgmt, SRpcMsg *pMsg) { const STraceId *trace = &pMsg->info.traceId; dGTrace("msg:%p, put into vnode-multi-mgmt queue", pMsg); @@ -409,6 +435,10 @@ int32_t vmGetQueueSize(SVnodeMgmt *pMgmt, int32_t vgId, EQueueType qtype) { break; case STREAM_CTRL_QUEUE: size = taosQueueItemSize(pVnode->pStreamCtrlQ); + break; + case STREAM_LONG_EXEC_QUEUE: + size = taosQueueItemSize(pVnode->pStreamLongExecQ); + break; default: break; } @@ -451,13 +481,16 @@ int32_t vmAllocQueue(SVnodeMgmt *pMgmt, SVnodeObj *pVnode) { } pVnode->pQueryQ = tQueryAutoQWorkerAllocQueue(&pMgmt->queryPool, pVnode, (FItem)vmProcessQueryQueue); - pVnode->pStreamQ = tAutoQWorkerAllocQueue(&pMgmt->streamPool, pVnode, (FItem)vmProcessStreamQueue); pVnode->pFetchQ = tWWorkerAllocQueue(&pMgmt->fetchPool, pVnode, (FItems)vmProcessFetchQueue); + + // init stream msg processing queue family + pVnode->pStreamQ = tAutoQWorkerAllocQueue(&pMgmt->streamPool, pVnode, (FItem)vmProcessStreamQueue, 2); pVnode->pStreamCtrlQ = tWWorkerAllocQueue(&pMgmt->streamCtrlPool, pVnode, (FItems)vmProcessStreamCtrlQueue); + pVnode->pStreamLongExecQ = tAutoQWorkerAllocQueue(&pMgmt->streamLongExecPool, pVnode, (FItem)vmProcessStreamLongExecQueue, 1); if (pVnode->pWriteW.queue == NULL || pVnode->pSyncW.queue == NULL || pVnode->pSyncRdW.queue == NULL || pVnode->pApplyW.queue == NULL || pVnode->pQueryQ == NULL || pVnode->pStreamQ == NULL || pVnode->pFetchQ == NULL - || pVnode->pStreamCtrlQ == NULL) { + || pVnode->pStreamCtrlQ == NULL || pVnode->pStreamLongExecQ == NULL) { return TSDB_CODE_OUT_OF_MEMORY; } @@ -473,6 +506,7 @@ int32_t vmAllocQueue(SVnodeMgmt *pMgmt, SVnodeObj *pVnode) { dInfo("vgId:%d, fetch-queue:%p is alloced, thread:%08" PRId64, pVnode->vgId, pVnode->pFetchQ, taosQueueGetThreadId(pVnode->pFetchQ)); dInfo("vgId:%d, stream-queue:%p is alloced", pVnode->vgId, pVnode->pStreamQ); + dInfo("vgId:%d, stream-long-exec-queue:%p is alloced", pVnode->vgId, pVnode->pStreamLongExecQ); dInfo("vgId:%d, stream-ctrl-queue:%p is alloced, thread:%08" PRId64, pVnode->vgId, pVnode->pStreamCtrlQ, taosQueueGetThreadId(pVnode->pStreamCtrlQ)); return 0; @@ -481,17 +515,22 @@ int32_t vmAllocQueue(SVnodeMgmt *pMgmt, SVnodeObj *pVnode) { void vmFreeQueue(SVnodeMgmt *pMgmt, SVnodeObj *pVnode) { tQueryAutoQWorkerFreeQueue(&pMgmt->queryPool, pVnode->pQueryQ); tAutoQWorkerFreeQueue(&pMgmt->streamPool, pVnode->pStreamQ); + tAutoQWorkerFreeQueue(&pMgmt->streamLongExecPool, pVnode->pStreamLongExecQ); tWWorkerFreeQueue(&pMgmt->streamCtrlPool, pVnode->pStreamCtrlQ); tWWorkerFreeQueue(&pMgmt->fetchPool, pVnode->pFetchQ); pVnode->pQueryQ = NULL; + pVnode->pFetchQ = NULL; + pVnode->pStreamQ = NULL; pVnode->pStreamCtrlQ = NULL; - pVnode->pFetchQ = NULL; + pVnode->pStreamLongExecQ = NULL; + dDebug("vgId:%d, queue is freed", pVnode->vgId); } int32_t vmStartWorker(SVnodeMgmt *pMgmt) { - int32_t code = 0; + int32_t code = 0; + SQueryAutoQWorkerPool *pQPool = &pMgmt->queryPool; pQPool->name = "vnode-query"; pQPool->min = tsNumOfVnodeQueryThreads; @@ -505,8 +544,13 @@ int32_t vmStartWorker(SVnodeMgmt *pMgmt) { pStreamPool->ratio = tsRatioOfVnodeStreamThreads; if ((code = tAutoQWorkerInit(pStreamPool)) != 0) return code; + SAutoQWorkerPool *pLongExecPool = &pMgmt->streamLongExecPool; + pLongExecPool->name = "vnode-stream-long-exec"; + pLongExecPool->ratio = tsRatioOfVnodeStreamThreads/3; + if ((code = tAutoQWorkerInit(pLongExecPool)) != 0) return code; + SWWorkerPool *pStreamCtrlPool = &pMgmt->streamCtrlPool; - pStreamCtrlPool->name = "vnode-ctrl-stream"; + pStreamCtrlPool->name = "vnode-stream-ctrl"; pStreamCtrlPool->max = 1; if ((code = tWWorkerInit(pStreamCtrlPool)) != 0) return code; @@ -541,6 +585,7 @@ int32_t vmStartWorker(SVnodeMgmt *pMgmt) { void vmStopWorker(SVnodeMgmt *pMgmt) { tQueryAutoQWorkerCleanup(&pMgmt->queryPool); tAutoQWorkerCleanup(&pMgmt->streamPool); + tAutoQWorkerCleanup(&pMgmt->streamLongExecPool); tWWorkerCleanup(&pMgmt->streamCtrlPool); tWWorkerCleanup(&pMgmt->fetchPool); dDebug("vnode workers are closed"); diff --git a/source/dnode/mnode/impl/src/mndStream.c b/source/dnode/mnode/impl/src/mndStream.c index 220319122c..009cd1a108 100644 --- a/source/dnode/mnode/impl/src/mndStream.c +++ b/source/dnode/mnode/impl/src/mndStream.c @@ -2663,7 +2663,7 @@ int32_t mndProcessConsensusInTmr(SRpcMsg *pMsg) { int64_t now = taosGetTimestampMs(); bool allReady = true; SArray *pNodeSnapshot = NULL; - int32_t maxAllowedTrans = 50; + int32_t maxAllowedTrans = 20; int32_t numOfTrans = 0; int32_t code = 0; void *pIter = NULL; @@ -2750,6 +2750,7 @@ int32_t mndProcessConsensusInTmr(SRpcMsg *pMsg) { return TSDB_CODE_FAILED; } + // todo: check for redundant consensus-checkpoint trans, if this kinds of trans repeatly failed. code = mndCreateSetConsensusChkptIdTrans(pMnode, pStream, pe->req.taskId, chkId, pe->req.startTs); if (code != TSDB_CODE_SUCCESS && code != TSDB_CODE_ACTION_IN_PROGRESS) { mError("failed to create consensus-checkpoint trans, stream:0x%" PRIx64, pStream->uid); diff --git a/source/dnode/vnode/inc/vnode.h b/source/dnode/vnode/inc/vnode.h index 50d75c4838..871a8c06e1 100644 --- a/source/dnode/vnode/inc/vnode.h +++ b/source/dnode/vnode/inc/vnode.h @@ -113,6 +113,7 @@ int32_t vnodeProcessQueryMsg(SVnode *pVnode, SRpcMsg *pMsg, SQueueInfo *pInfo); int32_t vnodeProcessFetchMsg(SVnode *pVnode, SRpcMsg *pMsg, SQueueInfo *pInfo); int32_t vnodeProcessStreamMsg(SVnode *pVnode, SRpcMsg *pMsg, SQueueInfo *pInfo); int32_t vnodeProcessStreamCtrlMsg(SVnode *pVnode, SRpcMsg *pMsg, SQueueInfo *pInfo); +int32_t vnodeProcessStreamLongExecMsg(SVnode *pVnode, SRpcMsg *pMsg, SQueueInfo *pInfo); void vnodeProposeWriteMsg(SQueueInfo *pInfo, STaosQall *qall, int32_t numOfMsgs); void vnodeApplyWriteMsg(SQueueInfo *pInfo, STaosQall *qall, int32_t numOfMsgs); void vnodeProposeCommitOnNeed(SVnode *pVnode, bool atExit); diff --git a/source/dnode/vnode/src/inc/vnodeInt.h b/source/dnode/vnode/src/inc/vnodeInt.h index 5bf0a9b199..6b3cf47d13 100644 --- a/source/dnode/vnode/src/inc/vnodeInt.h +++ b/source/dnode/vnode/src/inc/vnodeInt.h @@ -239,7 +239,7 @@ void tqClose(STQ*); int tqPushMsg(STQ*, tmsg_t msgType); int tqRegisterPushHandle(STQ* pTq, void* handle, SRpcMsg* pMsg); void tqUnregisterPushHandle(STQ* pTq, void* pHandle); -int tqScanWalAsync(STQ* pTq, bool ckPause); +void tqScanWalAsync(STQ* pTq); int32_t tqStopStreamTasksAsync(STQ* pTq); int32_t tqProcessTaskCheckPointSourceReq(STQ* pTq, SRpcMsg* pMsg, SRpcMsg* pRsp); int32_t tqProcessTaskCheckpointReadyMsg(STQ* pTq, SRpcMsg* pMsg); diff --git a/source/dnode/vnode/src/sma/smaRollup.c b/source/dnode/vnode/src/sma/smaRollup.c index 7d83dbcf84..1f755f816e 100644 --- a/source/dnode/vnode/src/sma/smaRollup.c +++ b/source/dnode/vnode/src/sma/smaRollup.c @@ -1302,7 +1302,7 @@ _checkpoint: } streamMetaWLock(pMeta); - if ((code = streamMetaSaveTask(pMeta, pTask)) != 0) { + if ((code = streamMetaSaveTaskInMeta(pMeta, pTask)) != 0) { streamMetaWUnLock(pMeta); taosHashCancelIterate(pInfoHash, infoHash); TSDB_CHECK_CODE(code, lino, _exit); diff --git a/source/dnode/vnode/src/tq/tq.c b/source/dnode/vnode/src/tq/tq.c index 119edb47bc..63727e5c45 100644 --- a/source/dnode/vnode/src/tq/tq.c +++ b/source/dnode/vnode/src/tq/tq.c @@ -920,12 +920,12 @@ static void doStartFillhistoryStep2(SStreamTask* pTask, SStreamTask* pStreamTask // now the fill-history task starts to scan data from wal files. code = streamTaskHandleEvent(pTask->status.pSM, TASK_EVENT_SCANHIST_DONE); - if (code == TSDB_CODE_SUCCESS) { - code = tqScanWalAsync(pTq, false); - if (code) { - tqError("vgId:%d failed to start scan wal file, code:%s", vgId, tstrerror(code)); - } - } +// if (code == TSDB_CODE_SUCCESS) { +// code = tqScanWalAsync(pTq, false); +// if (code) { +// tqError("vgId:%d failed to start scan wal file, code:%s", vgId, tstrerror(code)); +// } +// } } } @@ -1113,23 +1113,14 @@ int32_t tqProcessTaskRunReq(STQ* pTq, SRpcMsg* pMsg) { // extracted submit data from wal files for all tasks if (req.reqType == STREAM_EXEC_T_EXTRACT_WAL_DATA) { return tqScanWal(pTq); - } + } else { + code = tqStreamTaskProcessRunReq(pTq->pStreamMeta, pMsg, vnodeIsRoleLeader(pTq->pVnode)); + if (code) { + tqError("vgId:%d failed to create task run req, code:%s", TD_VID(pTq->pVnode), tstrerror(code)); + } - code = tqStreamTaskProcessRunReq(pTq->pStreamMeta, pMsg, vnodeIsRoleLeader(pTq->pVnode)); - if (code) { - tqError("vgId:%d failed to create task run req, code:%s", TD_VID(pTq->pVnode), tstrerror(code)); return code; } - - // let's continue scan data in the wal files - if (req.reqType >= 0 || req.reqType == STREAM_EXEC_T_RESUME_TASK) { - code = tqScanWalAsync(pTq, false); // it's ok to failed - if (code) { - tqError("vgId:%d failed to start scan wal file, code:%s", pTq->pStreamMeta->vgId, tstrerror(code)); - } - } - - return code; } int32_t tqProcessTaskDispatchReq(STQ* pTq, SRpcMsg* pMsg) { diff --git a/source/dnode/vnode/src/tq/tqPush.c b/source/dnode/vnode/src/tq/tqPush.c index 2b2667773a..fc83343c99 100644 --- a/source/dnode/vnode/src/tq/tqPush.c +++ b/source/dnode/vnode/src/tq/tqPush.c @@ -49,20 +49,6 @@ int32_t tqPushMsg(STQ* pTq, tmsg_t msgType) { } } - streamMetaRLock(pTq->pStreamMeta); - int32_t numOfTasks = streamMetaGetNumOfTasks(pTq->pStreamMeta); - streamMetaRUnLock(pTq->pStreamMeta); - -// tqTrace("vgId:%d handle submit, restore:%d, numOfTasks:%d", TD_VID(pTq->pVnode), pTq->pVnode->restored, numOfTasks); - - // push data for stream processing: - // 1. the vnode has already been restored. - // 2. the vnode should be the leader. - // 3. the stream is not suspended yet. - if ((!tsDisableStream) && (numOfTasks > 0)) { - code = tqScanWalAsync(pTq, true); - } - return code; } diff --git a/source/dnode/vnode/src/tq/tqStreamTask.c b/source/dnode/vnode/src/tq/tqStreamTask.c index 9ea84830f1..b34ea78f64 100644 --- a/source/dnode/vnode/src/tq/tqStreamTask.c +++ b/source/dnode/vnode/src/tq/tqStreamTask.c @@ -16,23 +16,19 @@ #include "tq.h" #include "vnd.h" -#define MAX_REPEAT_SCAN_THRESHOLD 3 -#define SCAN_WAL_IDLE_DURATION 500 // idle for 500ms to do next wal scan +#define SCAN_WAL_IDLE_DURATION 250 // idle for 500ms to do next wal scan +#define SCAN_WAL_WAIT_COUNT 2 typedef struct SBuildScanWalMsgParam { int64_t metaId; - int32_t numOfTasks; - int8_t restored; SMsgCb msgCb; } SBuildScanWalMsgParam; -static int32_t doScanWalForAllTasks(SStreamMeta* pStreamMeta); +static int32_t doScanWalForAllTasks(SStreamMeta* pStreamMeta, int32_t* pNumOfTasks); static int32_t setWalReaderStartOffset(SStreamTask* pTask, int32_t vgId); static bool handleFillhistoryScanComplete(SStreamTask* pTask, int64_t ver); static bool taskReadyForDataFromWal(SStreamTask* pTask); static int32_t doPutDataIntoInputQ(SStreamTask* pTask, int64_t maxVer, int32_t* numOfItems, bool* pSucc); -static int32_t tqScanWalInFuture(STQ* pTq, int32_t numOfTasks, int32_t idleDuration); -static int32_t doScanWalAsync(STQ* pTq, bool ckPause); // extract data blocks(submit/delete) from WAL, and add them into the input queue for all the sources tasks. int32_t tqScanWal(STQ* pTq) { @@ -40,44 +36,55 @@ int32_t tqScanWal(STQ* pTq) { int32_t vgId = pMeta->vgId; int64_t st = taosGetTimestampMs(); int32_t numOfTasks = 0; + int64_t el = 0; + int32_t code = 0; - tqDebug("vgId:%d continue to check if data in wal are available, scanCounter:%d", vgId, pMeta->scanInfo.scanCounter); - - // check all tasks - int32_t code = doScanWalForAllTasks(pMeta); - if (code) { - tqError("vgId:%d failed to start all tasks, try next time, code:%s", vgId, tstrerror(code)); + int32_t old = atomic_val_compare_exchange_32(&pMeta->scanInfo.scanSentinel, 0, 1); + if (old == 0) { + tqDebug("vgId:%d try to scan wal to extract data", vgId); + } else { + tqDebug("vgId:%d already in wal scan, abort", vgId); return code; } - streamMetaWLock(pMeta); - int32_t times = (--pMeta->scanInfo.scanCounter); - if (times < 0) { - tqError("vgId:%d invalid scan counter:%d, reset to 0", vgId, times); - times = 0; + // the scan wal interval less than 200, not scan, actually. + if ((pMeta->scanInfo.lastScanTs > 0) && (st - pMeta->scanInfo.lastScanTs < 200)) { + tqDebug("vgId:%d scan wal less than 200ms, do nothing", vgId); + atomic_store_32(&pMeta->scanInfo.scanSentinel, 0); + return code; } - numOfTasks = taosArrayGetSize(pMeta->pTaskList); - streamMetaWUnLock(pMeta); + // check all tasks + code = doScanWalForAllTasks(pMeta, &numOfTasks); - int64_t el = (taosGetTimestampMs() - st); - tqDebug("vgId:%d scan wal for stream tasks completed, elapsed time:%" PRId64 " ms", vgId, el); + pMeta->scanInfo.lastScanTs = taosGetTimestampMs(); + el = (pMeta->scanInfo.lastScanTs - st); - if (times > 0) { - tqDebug("vgId:%d scan wal for stream tasks for %d times in %dms", vgId, times, SCAN_WAL_IDLE_DURATION); - code = tqScanWalInFuture(pTq, numOfTasks, SCAN_WAL_IDLE_DURATION); - if (code) { - tqError("vgId:%d sched scan wal in %dms failed, ignore this failure", vgId, SCAN_WAL_IDLE_DURATION); - } + if (code) { + tqError("vgId:%d failed to scan wal for all tasks, try next time, elapsed time:%" PRId64 "ms code:%s", vgId, el, + tstrerror(code)); + } else { + tqDebug("vgId:%d scan wal for stream tasks completed, elapsed time:%" PRId64 "ms", vgId, el); } + atomic_store_32(&pMeta->scanInfo.scanSentinel, 0); return code; } -static void doStartScanWal(void* param, void* tmrId) { - int32_t vgId = 0; - int32_t code = 0; +static bool waitEnoughDuration(SStreamMeta* pMeta) { + if ((++pMeta->scanInfo.tickCounter) >= SCAN_WAL_WAIT_COUNT) { + pMeta->scanInfo.tickCounter = 0; + return true; + } + return false; +} + +static void doStartScanWal(void* param, void* tmrId) { + int32_t vgId = 0; + int32_t code = 0; + int32_t numOfTasks = 0; + tmr_h pTimer = NULL; SBuildScanWalMsgParam* pParam = (SBuildScanWalMsgParam*)param; SStreamMeta* pMeta = taosAcquireRef(streamMetaRefPool, pParam->metaId); @@ -87,10 +94,18 @@ static void doStartScanWal(void* param, void* tmrId) { return; } + vgId = pMeta->vgId; + code = streamTimerGetInstance(&pTimer); + if (code) { + tqFatal("vgId:%d failed to get tmr ctrl during sched scan wal, not scan wal, code:%s", vgId, tstrerror(code)); + taosMemoryFree(pParam); + return; + } + if (pMeta->closeFlag) { code = taosReleaseRef(streamMetaRefPool, pParam->metaId); if (code == TSDB_CODE_SUCCESS) { - tqDebug("vgId:%d jump out of scan wal timer since closed", vgId); + tqInfo("vgId:%d jump out of scan wal timer since closed", vgId); } else { tqError("vgId:%d failed to release ref for streamMeta, rid:%" PRId64 " code:%s", vgId, pParam->metaId, tstrerror(code)); @@ -100,71 +115,100 @@ static void doStartScanWal(void* param, void* tmrId) { return; } - vgId = pMeta->vgId; + if (pMeta->role != NODE_ROLE_LEADER) { + tqDebug("vgId:%d not leader, role:%d not scan wal anymore", vgId, pMeta->role); - tqDebug("vgId:%d create msg to start wal scan, numOfTasks:%d, vnd restored:%d", vgId, pParam->numOfTasks, - pParam->restored); -#if 0 - // wait for the vnode is freed, and invalid read may occur. + code = taosReleaseRef(streamMetaRefPool, pParam->metaId); + if (code == TSDB_CODE_SUCCESS) { + tqDebug("vgId:%d jump out of scan wal timer since not leader", vgId); + } else { + tqError("vgId:%d failed to release ref for streamMeta, rid:%" PRId64 " code:%s", vgId, pParam->metaId, + tstrerror(code)); + } + + taosMemFree(pParam); + return; + } + + if (pMeta->startInfo.startAllTasks) { + tqTrace("vgId:%d in restart procedure, not ready to scan wal", vgId); + goto _end; + } + + if (!waitEnoughDuration(pMeta)) { + streamTmrStart(doStartScanWal, SCAN_WAL_IDLE_DURATION, pParam, pTimer, &pMeta->scanInfo.scanTimer, vgId, + "scan-wal"); + code = taosReleaseRef(streamMetaRefPool, pParam->metaId); + if (code) { + tqError("vgId:%d failed to release ref for streamMeta, rid:%" PRId64 " code:%s", vgId, pParam->metaId, + tstrerror(code)); + } + return; + } + + streamMetaRLock(pMeta); + numOfTasks = taosArrayGetSize(pMeta->pTaskList); + streamMetaRUnLock(pMeta); + + if (numOfTasks == 0) { + goto _end; + } + + tqTrace("vgId:%d create msg to start wal scan, numOfTasks:%d", vgId, numOfTasks); + + #if 0 + // wait for the vnode is freed, and invalid read may occur. taosMsleep(10000); -#endif + #endif code = streamTaskSchedTask(&pParam->msgCb, vgId, 0, 0, STREAM_EXEC_T_EXTRACT_WAL_DATA); if (code) { tqError("vgId:%d failed sched task to scan wal, code:%s", vgId, tstrerror(code)); } +_end: + + streamTmrStart(doStartScanWal, SCAN_WAL_IDLE_DURATION, pParam, pTimer, &pMeta->scanInfo.scanTimer, vgId, "scan-wal"); + tqDebug("vgId:%d scan-wal will start in %dms", vgId, SCAN_WAL_IDLE_DURATION*SCAN_WAL_WAIT_COUNT); + code = taosReleaseRef(streamMetaRefPool, pParam->metaId); if (code) { tqError("vgId:%d failed to release ref for streamMeta, rid:%" PRId64 " code:%s", vgId, pParam->metaId, tstrerror(code)); } - - taosMemoryFree(pParam); } -int32_t tqScanWalInFuture(STQ* pTq, int32_t numOfTasks, int32_t idleDuration) { +void tqScanWalAsync(STQ* pTq) { SStreamMeta* pMeta = pTq->pStreamMeta; int32_t code = 0; int32_t vgId = TD_VID(pTq->pVnode); tmr_h pTimer = NULL; SBuildScanWalMsgParam* pParam = NULL; + // 1. the vnode should be the leader. + // 2. the stream isn't disabled + if ((pMeta->role == NODE_ROLE_FOLLOWER) || tsDisableStream) { + tqInfo("vgId:%d follower node or stream disabled, not scan wal", vgId); + return; + } + pParam = taosMemoryMalloc(sizeof(SBuildScanWalMsgParam)); if (pParam == NULL) { - return terrno; + tqError("vgId:%d failed to start scan wal, stream not executes, code:%s", vgId, tstrerror(code)); + return; } pParam->metaId = pMeta->rid; - pParam->numOfTasks = numOfTasks; - pParam->restored = pTq->pVnode->restored; pParam->msgCb = pTq->pVnode->msgCb; code = streamTimerGetInstance(&pTimer); if (code) { - tqError("vgId:%d failed to get tmr ctrl during sched scan wal", vgId); + tqFatal("vgId:%d failed to get tmr ctrl during sched scan wal", vgId); taosMemoryFree(pParam); } else { - streamTmrStart(doStartScanWal, idleDuration, pParam, pTimer, &pMeta->scanInfo.scanTimer, vgId, "scan-wal-fut"); + streamTmrStart(doStartScanWal, SCAN_WAL_IDLE_DURATION, pParam, pTimer, &pMeta->scanInfo.scanTimer, vgId, + "scan-wal"); } - - return code; -} - -int32_t tqScanWalAsync(STQ* pTq, bool ckPause) { - SStreamMeta* pMeta = pTq->pStreamMeta; - bool alreadyRestored = pTq->pVnode->restored; - int32_t code = 0; - - // do not launch the stream tasks, if it is a follower or not restored vnode. - if (!(vnodeIsRoleLeader(pTq->pVnode) && alreadyRestored)) { - return TSDB_CODE_SUCCESS; - } - - streamMetaWLock(pMeta); - code = doScanWalAsync(pTq, ckPause); - streamMetaWUnLock(pMeta); - return code; } int32_t tqStopStreamTasksAsync(STQ* pTq) { @@ -347,13 +391,10 @@ int32_t doPutDataIntoInputQ(SStreamTask* pTask, int64_t maxVer, int32_t* numOfIt return code; } -int32_t doScanWalForAllTasks(SStreamMeta* pStreamMeta) { +int32_t doScanWalForAllTasks(SStreamMeta* pStreamMeta, int32_t* pNumOfTasks) { int32_t vgId = pStreamMeta->vgId; SArray* pTaskList = NULL; - int32_t numOfTasks = taosArrayGetSize(pStreamMeta->pTaskList); - if (numOfTasks == 0) { - return TSDB_CODE_SUCCESS; - } + int32_t numOfTasks = 0; // clone the task list, to avoid the task update during scan wal files streamMetaWLock(pStreamMeta); @@ -364,10 +405,13 @@ int32_t doScanWalForAllTasks(SStreamMeta* pStreamMeta) { return terrno; } - tqDebug("vgId:%d start to check wal to extract new submit block for %d tasks", vgId, numOfTasks); - // update the new task number numOfTasks = taosArrayGetSize(pTaskList); + if (pNumOfTasks != NULL) { + *pNumOfTasks = numOfTasks; + } + + tqDebug("vgId:%d start to check wal to extract new submit block for %d tasks", vgId, numOfTasks); for (int32_t i = 0; i < numOfTasks; ++i) { STaskId* pTaskId = taosArrayGet(pTaskList, i); @@ -426,51 +470,9 @@ int32_t doScanWalForAllTasks(SStreamMeta* pStreamMeta) { return TSDB_CODE_SUCCESS; } -int32_t doScanWalAsync(STQ* pTq, bool ckPause) { - SStreamMeta* pMeta = pTq->pStreamMeta; - bool alreadyRestored = pTq->pVnode->restored; - int32_t vgId = pMeta->vgId; - int32_t numOfTasks = taosArrayGetSize(pMeta->pTaskList); - - if (numOfTasks == 0) { - tqDebug("vgId:%d no stream tasks existed to run", vgId); - return 0; - } - - if (pMeta->startInfo.startAllTasks) { - tqTrace("vgId:%d in restart procedure, not scan wal", vgId); - return 0; - } - - pMeta->scanInfo.scanCounter += 1; - if (pMeta->scanInfo.scanCounter > MAX_REPEAT_SCAN_THRESHOLD) { - pMeta->scanInfo.scanCounter = MAX_REPEAT_SCAN_THRESHOLD; - } - - if (pMeta->scanInfo.scanCounter > 1) { - tqDebug("vgId:%d wal read task has been launched, remain scan times:%d", vgId, pMeta->scanInfo.scanCounter); - return 0; - } - - int32_t numOfPauseTasks = pMeta->numOfPausedTasks; - if (ckPause && numOfTasks == numOfPauseTasks) { - tqDebug("vgId:%d ignore all submit, all streams had been paused, reset the walScanCounter", vgId); - - // reset the counter value, since we do not launch the scan wal operation. - pMeta->scanInfo.scanCounter = 0; - return 0; - } - - tqDebug("vgId:%d create msg to start wal scan to launch stream tasks, numOfTasks:%d, vnd restored:%d", vgId, - numOfTasks, alreadyRestored); - - return streamTaskSchedTask(&pTq->pVnode->msgCb, vgId, 0, 0, STREAM_EXEC_T_EXTRACT_WAL_DATA); -} - void streamMetaFreeTQDuringScanWalError(STQ* pTq) { SBuildScanWalMsgParam* p = taosMemoryCalloc(1, sizeof(SBuildScanWalMsgParam)); p->metaId = pTq->pStreamMeta->rid; - p->numOfTasks = 0; doStartScanWal(p, 0); } \ No newline at end of file diff --git a/source/dnode/vnode/src/tq/tqUtil.c b/source/dnode/vnode/src/tq/tqUtil.c index 197a45cdb9..6f001981fb 100644 --- a/source/dnode/vnode/src/tq/tqUtil.c +++ b/source/dnode/vnode/src/tq/tqUtil.c @@ -47,6 +47,10 @@ END: void tqUpdateNodeStage(STQ* pTq, bool isLeader) { SSyncState state = syncGetState(pTq->pVnode->sync); streamMetaUpdateStageRole(pTq->pStreamMeta, state.term, isLeader); + + if (isLeader) { + tqScanWalAsync(pTq); + } } static int32_t tqInitTaosxRsp(SMqDataRsp* pRsp, STqOffsetVal pOffset) { diff --git a/source/dnode/vnode/src/tqCommon/tqCommon.c b/source/dnode/vnode/src/tqCommon/tqCommon.c index 77632d328e..2c48ada0fa 100644 --- a/source/dnode/vnode/src/tqCommon/tqCommon.c +++ b/source/dnode/vnode/src/tqCommon/tqCommon.c @@ -260,13 +260,13 @@ int32_t tqStreamTaskProcessUpdateReq(SStreamMeta* pMeta, SMsgCb* cb, SRpcMsg* pM // stream do update the nodeEp info, write it into stream meta. if (updated) { tqDebug("s-task:%s vgId:%d save task after update epset, and stop task", idstr, vgId); - code = streamMetaSaveTask(pMeta, pTask); + code = streamMetaSaveTaskInMeta(pMeta, pTask); if (code) { tqError("s-task:%s vgId:%d failed to save task, code:%s", idstr, vgId, tstrerror(code)); } if (pHTask != NULL) { - code = streamMetaSaveTask(pMeta, pHTask); + code = streamMetaSaveTaskInMeta(pMeta, pHTask); if (code) { tqError("s-task:%s vgId:%d failed to save related history task, code:%s", idstr, vgId, tstrerror(code)); } @@ -743,6 +743,8 @@ int32_t tqStreamTaskProcessDropReq(SStreamMeta* pMeta, char* msg, int32_t msgLen } streamMetaWUnLock(pMeta); + tqDebug("vgId:%d process drop task:0x%x completed", vgId, pReq->taskId); + return 0; // always return success } @@ -857,6 +859,9 @@ int32_t tqStreamTaskProcessRunReq(SStreamMeta* pMeta, SRpcMsg* pMsg, bool isLead } else if (type == STREAM_EXEC_T_ADD_FAILED_TASK) { code = streamMetaAddFailedTask(pMeta, req.streamId, req.taskId); return code; + } else if (type == STREAM_EXEC_T_STOP_ONE_TASK) { + code = streamMetaStopOneTask(pMeta, req.streamId, req.taskId); + return code; } else if (type == STREAM_EXEC_T_RESUME_TASK) { // task resume to run after idle for a while SStreamTask* pTask = NULL; code = streamMetaAcquireTask(pMeta, req.streamId, req.taskId, &pTask); @@ -938,10 +943,10 @@ int32_t tqStartTaskCompleteCallback(SStreamMeta* pMeta) { streamMetaWUnLock(pMeta); - if (scanWal && (vgId != SNODE_HANDLE)) { - tqDebug("vgId:%d start scan wal for executing tasks", vgId); - code = tqScanWalAsync(pMeta->ahandle, true); - } +// if (scanWal && (vgId != SNODE_HANDLE)) { +// tqDebug("vgId:%d start scan wal for executing tasks", vgId); +// code = tqScanWalAsync(pMeta->ahandle, true); +// } return code; } @@ -1170,7 +1175,7 @@ static int32_t tqProcessTaskResumeImpl(void* handle, SStreamTask* pTask, int64_t pTask->hTaskInfo.operatorOpen = false; code = streamStartScanHistoryAsync(pTask, igUntreated); } else if (level == TASK_LEVEL__SOURCE && (streamQueueGetNumOfItems(pTask->inputq.queue) == 0)) { - code = tqScanWalAsync((STQ*)handle, false); +// code = tqScanWalAsync((STQ*)handle, false); } else { code = streamTrySchedExec(pTask); } diff --git a/source/dnode/vnode/src/vnd/vnodeSvr.c b/source/dnode/vnode/src/vnd/vnodeSvr.c index 6190f4b0a7..6562504988 100644 --- a/source/dnode/vnode/src/vnd/vnodeSvr.c +++ b/source/dnode/vnode/src/vnd/vnodeSvr.c @@ -945,8 +945,6 @@ int32_t vnodeProcessStreamMsg(SVnode *pVnode, SRpcMsg *pMsg, SQueueInfo *pInfo) return tqProcessTaskRetrieveReq(pVnode->pTq, pMsg); case TDMT_STREAM_RETRIEVE_RSP: return tqProcessTaskRetrieveRsp(pVnode->pTq, pMsg); - case TDMT_VND_STREAM_SCAN_HISTORY: - return tqProcessTaskScanHistory(pVnode->pTq, pMsg); case TDMT_VND_GET_STREAM_PROGRESS: return tqStreamProgressRetrieveReq(pVnode->pTq, pMsg); default: @@ -993,6 +991,22 @@ int32_t vnodeProcessStreamCtrlMsg(SVnode *pVnode, SRpcMsg *pMsg, SQueueInfo *pIn } } +int32_t vnodeProcessStreamLongExecMsg(SVnode *pVnode, SRpcMsg *pMsg, SQueueInfo *pInfo) { + vTrace("vgId:%d, msg:%p in stream long exec queue is processing", pVnode->config.vgId, pMsg); + if (!syncIsReadyForRead(pVnode->sync)) { + vnodeRedirectRpcMsg(pVnode, pMsg, terrno); + return 0; + } + + switch (pMsg->msgType) { + case TDMT_VND_STREAM_SCAN_HISTORY: + return tqProcessTaskScanHistory(pVnode->pTq, pMsg); + default: + vError("unknown msg type:%d in stream long exec queue", pMsg->msgType); + return TSDB_CODE_APP_ERROR; + } +} + void smaHandleRes(void *pVnode, int64_t smaId, const SArray *data) { int32_t code = tdProcessTSmaInsert(((SVnode *)pVnode)->pSma, smaId, (const char *)data); if (code) { diff --git a/source/libs/executor/src/executor.c b/source/libs/executor/src/executor.c index 364ab02c80..b6044eaacf 100644 --- a/source/libs/executor/src/executor.c +++ b/source/libs/executor/src/executor.c @@ -972,26 +972,43 @@ int32_t qAsyncKillTask(qTaskInfo_t qinfo, int32_t rspCode) { return TSDB_CODE_SUCCESS; } -int32_t qKillTask(qTaskInfo_t tinfo, int32_t rspCode) { +int32_t qKillTask(qTaskInfo_t tinfo, int32_t rspCode, int64_t waitDuration) { + int64_t st = taosGetTimestampMs(); SExecTaskInfo* pTaskInfo = (SExecTaskInfo*)tinfo; if (pTaskInfo == NULL) { return TSDB_CODE_QRY_INVALID_QHANDLE; } - qDebug("%s sync killed execTask", GET_TASKID(pTaskInfo)); + if (waitDuration > 0) { + qDebug("%s sync killed execTask, and waiting for %.2fs", GET_TASKID(pTaskInfo), waitDuration/1000.0); + } else { + qDebug("%s async killed execTask", GET_TASKID(pTaskInfo)); + } + setTaskKilled(pTaskInfo, TSDB_CODE_TSC_QUERY_KILLED); - while (1) { - taosWLockLatch(&pTaskInfo->lock); - if (qTaskIsExecuting(pTaskInfo)) { // let's wait for 100 ms and try again - taosWUnLockLatch(&pTaskInfo->lock); - taosMsleep(100); - } else { // not running now - pTaskInfo->code = rspCode; - taosWUnLockLatch(&pTaskInfo->lock); - return TSDB_CODE_SUCCESS; + if (waitDuration > 0) { + while (1) { + taosWLockLatch(&pTaskInfo->lock); + if (qTaskIsExecuting(pTaskInfo)) { // let's wait for 100 ms and try again + taosWUnLockLatch(&pTaskInfo->lock); + + taosMsleep(200); + + int64_t d = taosGetTimestampMs() - st; + if (d >= waitDuration && waitDuration >= 0) { + qWarn("%s waiting more than %.2fs, not wait anymore", GET_TASKID(pTaskInfo), waitDuration / 1000.0); + return TSDB_CODE_SUCCESS; + } + } else { // not running now + pTaskInfo->code = rspCode; + taosWUnLockLatch(&pTaskInfo->lock); + return TSDB_CODE_SUCCESS; + } } } + + return TSDB_CODE_SUCCESS; } bool qTaskIsExecuting(qTaskInfo_t qinfo) { diff --git a/source/libs/executor/src/projectoperator.c b/source/libs/executor/src/projectoperator.c index eb448a13f5..b0f78c1aad 100644 --- a/source/libs/executor/src/projectoperator.c +++ b/source/libs/executor/src/projectoperator.c @@ -42,9 +42,7 @@ typedef struct SIndefOperatorInfo { } SIndefOperatorInfo; static int32_t doGenerateSourceData(SOperatorInfo* pOperator); -static SSDataBlock* doProjectOperation1(SOperatorInfo* pOperator); static int32_t doProjectOperation(SOperatorInfo* pOperator, SSDataBlock** pResBlock); -static SSDataBlock* doApplyIndefinitFunction1(SOperatorInfo* pOperator); static int32_t doApplyIndefinitFunction(SOperatorInfo* pOperator, SSDataBlock** pResBlock); static int32_t setRowTsColumnOutputInfo(SqlFunctionCtx* pCtx, int32_t numOfCols, SArray** pResList); static int32_t setFunctionResultOutput(SOperatorInfo* pOperator, SOptrBasicInfo* pInfo, SAggSupporter* pSup, @@ -557,12 +555,6 @@ static void doHandleDataBlock(SOperatorInfo* pOperator, SSDataBlock* pBlock, SOp } } -SSDataBlock* doApplyIndefinitFunction1(SOperatorInfo* pOperator) { - SSDataBlock* pResBlock = NULL; - pOperator->pTaskInfo->code = doApplyIndefinitFunction(pOperator, &pResBlock); - return pResBlock; -} - int32_t doApplyIndefinitFunction(SOperatorInfo* pOperator, SSDataBlock** pResBlock) { QRY_PARAM_CHECK(pResBlock); SIndefOperatorInfo* pIndefInfo = pOperator->info; diff --git a/source/libs/executor/src/streameventwindowoperator.c b/source/libs/executor/src/streameventwindowoperator.c index fa6008eba7..a2304a2e6c 100644 --- a/source/libs/executor/src/streameventwindowoperator.c +++ b/source/libs/executor/src/streameventwindowoperator.c @@ -234,6 +234,11 @@ int32_t updateEventWindowInfo(SStreamAggSupporter* pAggSup, SEventWindowInfo* pW pWin->ekey = pTsData[i]; pWinInfo->pWinFlag->endFlag = ends[i]; } else if (pWin->ekey == pTsData[i]) { + if (pWinInfo->pWinFlag->endFlag == true && ends[i] == false) { + (*pWinRow) = i + 1 - start; + *pRebuild = true; + goto _end; + } pWinInfo->pWinFlag->endFlag |= ends[i]; } else if (ends[i] && !pWinInfo->pWinFlag->endFlag) { *pRebuild = true; diff --git a/source/libs/executor/src/streamtimewindowoperator.c b/source/libs/executor/src/streamtimewindowoperator.c index 5cab26b9d3..3b6c1963dc 100644 --- a/source/libs/executor/src/streamtimewindowoperator.c +++ b/source/libs/executor/src/streamtimewindowoperator.c @@ -479,14 +479,12 @@ void destroyFlusedppPos(void* ppRes) { } void clearGroupResInfo(SGroupResInfo* pGroupResInfo) { - if (pGroupResInfo->freeItem) { - int32_t size = taosArrayGetSize(pGroupResInfo->pRows); - for (int32_t i = pGroupResInfo->index; i < size; i++) { - void* pPos = taosArrayGetP(pGroupResInfo->pRows, i); - destroyFlusedPos(pPos); - } - pGroupResInfo->freeItem = false; + int32_t size = taosArrayGetSize(pGroupResInfo->pRows); + for (int32_t i = pGroupResInfo->index; i < size; i++) { + void* pPos = taosArrayGetP(pGroupResInfo->pRows, i); + destroyFlusedPos(pPos); } + pGroupResInfo->freeItem = false; taosArrayDestroy(pGroupResInfo->pRows); pGroupResInfo->pRows = NULL; pGroupResInfo->index = 0; diff --git a/source/libs/stream/src/streamCheckpoint.c b/source/libs/stream/src/streamCheckpoint.c index 41e3d40542..387010d0a6 100644 --- a/source/libs/stream/src/streamCheckpoint.c +++ b/source/libs/stream/src/streamCheckpoint.c @@ -734,7 +734,7 @@ int32_t streamTaskUpdateTaskCheckpointInfo(SStreamTask* pTask, bool restored, SV pTask->status.taskStatus = TASK_STATUS__READY; - code = streamMetaSaveTask(pMeta, pTask); + code = streamMetaSaveTaskInMeta(pMeta, pTask); streamMutexUnlock(&pTask->lock); if (code != TSDB_CODE_SUCCESS) { @@ -1584,6 +1584,11 @@ int32_t streamTaskSendNegotiateChkptIdMsg(SStreamTask* pTask) { streamFreeTaskState(pTask, p); pTask->pBackend = NULL; } + + if (pTask->exec.pExecutor != NULL) { + qDestroyTask(pTask->exec.pExecutor); + pTask->exec.pExecutor = NULL; + } return 0; } diff --git a/source/libs/stream/src/streamExec.c b/source/libs/stream/src/streamExec.c index 999a6b5a4c..054f88ec3d 100644 --- a/source/libs/stream/src/streamExec.c +++ b/source/libs/stream/src/streamExec.c @@ -880,7 +880,7 @@ static int32_t doStreamExecTask(SStreamTask* pTask) { } double el = (taosGetTimestampMs() - st) / 1000.0; - if (el > 5.0) { // elapsed more than 5 sec, not occupy the CPU anymore + if (el > 2.0) { // elapsed more than 5 sec, not occupy the CPU anymore stDebug("s-task:%s occupy more than 5.0s, release the exec threads and idle for 500ms", id); streamTaskSetIdleInfo(pTask, 500); return code; diff --git a/source/libs/stream/src/streamHb.c b/source/libs/stream/src/streamHb.c index d973d7c05f..312ed85764 100644 --- a/source/libs/stream/src/streamHb.c +++ b/source/libs/stream/src/streamHb.c @@ -327,7 +327,7 @@ void streamMetaHbToMnode(void* param, void* tmrId) { pMeta->pHbInfo->hbStart = 0; code = taosReleaseRef(streamMetaRefPool, rid); if (code == TSDB_CODE_SUCCESS) { - stDebug("vgId:%d jump out of meta timer", vgId); + stInfo("vgId:%d jump out of meta timer since closed", vgId); } else { stError("vgId:%d jump out of meta timer, failed to release the meta rid:%" PRId64, vgId, rid); } @@ -341,7 +341,7 @@ void streamMetaHbToMnode(void* param, void* tmrId) { if (code == TSDB_CODE_SUCCESS) { stInfo("vgId:%d role:%d not leader not send hb to mnode", vgId, role); } else { - stError("vgId:%d role:%d not leader not send hb to mnodefailed to release the meta rid:%" PRId64, vgId, role, rid); + stError("vgId:%d role:%d not leader not send hb to mnode, failed to release meta rid:%" PRId64, vgId, role, rid); } // taosMemoryFree(param); return; @@ -417,7 +417,7 @@ void destroyMetaHbInfo(SMetaHbInfo* pInfo) { void streamMetaWaitForHbTmrQuit(SStreamMeta* pMeta) { // wait for the stream meta hb function stopping if (pMeta->role == NODE_ROLE_LEADER) { - taosMsleep(2 * META_HB_CHECK_INTERVAL); + taosMsleep(3 * META_HB_CHECK_INTERVAL); stDebug("vgId:%d wait for meta to stop timer", pMeta->vgId); } } diff --git a/source/libs/stream/src/streamMeta.c b/source/libs/stream/src/streamMeta.c index ddd0201460..d5561ebe76 100644 --- a/source/libs/stream/src/streamMeta.c +++ b/source/libs/stream/src/streamMeta.c @@ -427,7 +427,10 @@ int32_t streamMetaOpen(const char* path, void* ahandle, FTaskBuild buildTaskFn, pMeta->pTaskList = taosArrayInit(4, sizeof(SStreamTaskId)); TSDB_CHECK_NULL(pMeta->pTaskList, code, lino, _err, terrno); - pMeta->scanInfo.scanCounter = 0; + pMeta->scanInfo.scanSentinel = 0; + pMeta->scanInfo.lastScanTs = 0; + pMeta->scanInfo.tickCounter = 0; + pMeta->vgId = vgId; pMeta->ahandle = ahandle; pMeta->buildTaskFn = buildTaskFn; @@ -632,7 +635,7 @@ void streamMetaCloseImpl(void* arg) { } // todo let's check the status for each task -int32_t streamMetaSaveTask(SStreamMeta* pMeta, SStreamTask* pTask) { +int32_t streamMetaSaveTaskInMeta(SStreamMeta* pMeta, SStreamTask* pTask) { int32_t vgId = pTask->pMeta->vgId; void* buf = NULL; int32_t len; @@ -682,7 +685,7 @@ int32_t streamMetaSaveTask(SStreamMeta* pMeta, SStreamTask* pTask) { return code; } -int32_t streamMetaRemoveTask(SStreamMeta* pMeta, STaskId* pTaskId) { +int32_t streamMetaRemoveTaskInMeta(SStreamMeta* pMeta, STaskId* pTaskId) { int64_t key[2] = {pTaskId->streamId, pTaskId->taskId}; int32_t code = tdbTbDelete(pMeta->pTaskDb, key, STREAM_TASK_KEY_LEN, pMeta->txn); if (code != 0) { @@ -705,7 +708,7 @@ int32_t streamMetaRegisterTask(SStreamMeta* pMeta, int64_t ver, SStreamTask* pTa void* p = taosHashGet(pMeta->pTasksMap, &id, sizeof(id)); if (p != NULL) { - stDebug("s-task:%" PRIx64 " already exist in meta, no need to register", id.taskId); + stDebug("s-task:0x%" PRIx64 " already exist in meta, no need to register", id.taskId); tFreeStreamTask(pTask); return code; } @@ -735,7 +738,7 @@ int32_t streamMetaRegisterTask(SStreamMeta* pMeta, int64_t ver, SStreamTask* pTa return code; } - if ((code = streamMetaSaveTask(pMeta, pTask)) != 0) { + if ((code = streamMetaSaveTaskInMeta(pMeta, pTask)) != 0) { int32_t unused = taosHashRemove(pMeta->pTasksMap, &id, sizeof(id)); void* pUnused = taosArrayPop(pMeta->pTaskList); @@ -885,6 +888,7 @@ static void doRemoveIdFromList(SArray* pTaskList, int32_t num, SStreamTaskId* id static int32_t streamTaskSendTransSuccessMsg(SStreamTask* pTask, void* param) { int32_t code = 0; + if (pTask->info.taskLevel == TASK_LEVEL__SOURCE) { code = streamTaskSendCheckpointSourceRsp(pTask); if (code) { @@ -895,7 +899,7 @@ static int32_t streamTaskSendTransSuccessMsg(SStreamTask* pTask, void* param) { // let's kill the query procedure within stream, to end it ASAP. if (pTask->info.taskLevel != TASK_LEVEL__SINK && pTask->exec.pExecutor != NULL) { - code = qKillTask(pTask->exec.pExecutor, TSDB_CODE_SUCCESS); + code = qKillTask(pTask->exec.pExecutor, TSDB_CODE_SUCCESS, -1); if (code != TSDB_CODE_SUCCESS) { stError("s-task:%s failed to kill task related query handle, code:%s", pTask->id.idStr, tstrerror(code)); } @@ -932,7 +936,7 @@ int32_t streamMetaUnregisterTask(SStreamMeta* pMeta, int64_t streamId, int32_t t code = taosHashRemove(pMeta->pTasksMap, &id, sizeof(id)); doRemoveIdFromList(pMeta->pTaskList, (int32_t)taosArrayGetSize(pMeta->pTaskList), &pTask->id); - code = streamMetaRemoveTask(pMeta, &id); + code = streamMetaRemoveTaskInMeta(pMeta, &id); if (code) { stError("vgId:%d failed to remove task:0x%" PRIx64 ", code:%s", pMeta->vgId, id.taskId, tstrerror(code)); } @@ -963,6 +967,32 @@ int32_t streamMetaUnregisterTask(SStreamMeta* pMeta, int64_t streamId, int32_t t return 0; } +int32_t streamMetaStopOneTask(SStreamMeta* pMeta, int64_t streamId, int32_t taskId) { + SStreamTask* pTask = NULL; + int32_t code = 0; + int32_t vgId = pMeta->vgId; + int32_t numOfTasks = 0; + + streamMetaWLock(pMeta); + +// code = streamMetaUnregisterTask(pMeta, streamId, taskId); +// numOfTasks = streamMetaGetNumOfTasks(pMeta); +// if (code) { +// stError("vgId:%d failed to drop task:0x%x, code:%s", vgId, taskId, tstrerror(code)); +// } +// +// code = streamMetaCommit(pMeta); +// if (code) { +// stError("vgId:%d failed to commit after drop task:0x%x, code:%s", vgId, taskId, tstrerror(code)); +// } else { +// stDebug("s-task:0x%"PRIx64"-0x%x vgId:%d dropped, remain tasks:%d", streamId, taskId, pMeta->vgId, numOfTasks); +// } + + streamMetaWUnLock(pMeta); + + return code; +} + int32_t streamMetaBegin(SStreamMeta* pMeta) { streamMetaWLock(pMeta); int32_t code = tdbBegin(pMeta->db, &pMeta->txn, tdbDefaultMalloc, tdbDefaultFree, NULL, @@ -1185,7 +1215,7 @@ void streamMetaLoadAllTasks(SStreamMeta* pMeta) { if (taosArrayGetSize(pRecycleList) > 0) { for (int32_t i = 0; i < taosArrayGetSize(pRecycleList); ++i) { STaskId* pId = taosArrayGet(pRecycleList, i); - code = streamMetaRemoveTask(pMeta, pId); + code = streamMetaRemoveTaskInMeta(pMeta, pId); if (code) { stError("s-task:0x%" PRIx64 " failed to remove task, code:%s", pId->taskId, tstrerror(code)); } @@ -1213,8 +1243,8 @@ void streamMetaNotifyClose(SStreamMeta* pMeta) { vgId, (pMeta->role == NODE_ROLE_LEADER), startTs, sendCount); // wait for the stream meta hb function stopping - streamMetaWaitForHbTmrQuit(pMeta); pMeta->closeFlag = true; + streamMetaWaitForHbTmrQuit(pMeta); stDebug("vgId:%d start to check all tasks for closing", vgId); int64_t st = taosGetTimestampMs(); @@ -1253,6 +1283,12 @@ void streamMetaNotifyClose(SStreamMeta* pMeta) { double el = (taosGetTimestampMs() - st) / 1000.0; stDebug("vgId:%d stop all %d task(s) completed, elapsed time:%.2f Sec.", pMeta->vgId, numOfTasks, el); + + if (pMeta->scanInfo.scanTimer != NULL) { + streamTmrStop(pMeta->scanInfo.scanTimer); + pMeta->scanInfo.scanTimer = NULL; + } + streamMetaRUnLock(pMeta); } @@ -1320,7 +1356,7 @@ void streamMetaUpdateStageRole(SStreamMeta* pMeta, int64_t stage, bool isLeader) // mark the sign to send msg before close all tasks // 1. for leader vnode, always send msg before closing - // 2. for follower vnode, if it's is changed from leader, also sending msg before closing. + // 2. for follower vnode, if it's changed from leader, also sending msg before closing. if (pMeta->role == NODE_ROLE_LEADER) { pMeta->sendMsgBeforeClosing = true; } @@ -1330,11 +1366,11 @@ void streamMetaUpdateStageRole(SStreamMeta* pMeta, int64_t stage, bool isLeader) if (isLeader) { stInfo("vgId:%d update meta stage:%" PRId64 ", prev:%" PRId64 " leader:%d, start to send Hb, rid:%" PRId64, - pMeta->vgId, prevStage, stage, isLeader, pMeta->rid); + pMeta->vgId, stage, prevStage, isLeader, pMeta->rid); streamMetaStartHb(pMeta); } else { stInfo("vgId:%d update meta stage:%" PRId64 " prev:%" PRId64 " leader:%d sendMsg beforeClosing:%d", pMeta->vgId, - prevStage, stage, isLeader, pMeta->sendMsgBeforeClosing); + stage, prevStage, isLeader, pMeta->sendMsgBeforeClosing); } } diff --git a/source/libs/stream/src/streamStartHistory.c b/source/libs/stream/src/streamStartHistory.c index 54a8929123..f8b1b5ecbc 100644 --- a/source/libs/stream/src/streamStartHistory.c +++ b/source/libs/stream/src/streamStartHistory.c @@ -76,7 +76,7 @@ int32_t streamStartScanHistoryAsync(SStreamTask* pTask, int8_t igUntreated) { memcpy(serializedReq, &req, len); SRpcMsg rpcMsg = {.contLen = len, .pCont = serializedReq, .msgType = TDMT_VND_STREAM_SCAN_HISTORY}; - return tmsgPutToQueue(pTask->pMsgCb, STREAM_QUEUE, &rpcMsg); + return tmsgPutToQueue(pTask->pMsgCb, STREAM_LONG_EXEC_QUEUE, &rpcMsg); } void streamExecScanHistoryInFuture(SStreamTask* pTask, int32_t idleDuration) { diff --git a/source/libs/stream/src/streamStartTask.c b/source/libs/stream/src/streamStartTask.c index c40d5ef928..60c1694dda 100644 --- a/source/libs/stream/src/streamStartTask.c +++ b/source/libs/stream/src/streamStartTask.c @@ -451,7 +451,6 @@ int32_t streamMetaStopAllTasks(SStreamMeta* pMeta) { continue; } - int64_t refId = pTask->id.refId; int32_t ret = streamTaskStop(pTask); if (ret) { stError("s-task:0x%x failed to stop task, code:%s", pTaskId->taskId, tstrerror(ret)); diff --git a/source/libs/stream/src/streamTask.c b/source/libs/stream/src/streamTask.c index d27ed520c6..e4e8a37b37 100644 --- a/source/libs/stream/src/streamTask.c +++ b/source/libs/stream/src/streamTask.c @@ -703,7 +703,7 @@ int32_t streamTaskStop(SStreamTask* pTask) { } if (pTask->info.taskLevel != TASK_LEVEL__SINK && pTask->exec.pExecutor != NULL) { - code = qKillTask(pTask->exec.pExecutor, TSDB_CODE_SUCCESS); + code = qKillTask(pTask->exec.pExecutor, TSDB_CODE_SUCCESS, 5000); if (code != TSDB_CODE_SUCCESS) { stError("s-task:%s failed to kill task related query handle, code:%s", id, tstrerror(code)); } @@ -862,7 +862,7 @@ int32_t streamTaskClearHTaskAttr(SStreamTask* pTask, int32_t resetRelHalt) { pStreamTask->status.taskStatus = TASK_STATUS__READY; } - code = streamMetaSaveTask(pMeta, pStreamTask); + code = streamMetaSaveTaskInMeta(pMeta, pStreamTask); streamMutexUnlock(&(pStreamTask->lock)); streamMetaReleaseTask(pMeta, pStreamTask); @@ -1025,7 +1025,7 @@ static int32_t taskPauseCallback(SStreamTask* pTask, void* param) { // in case of fill-history task, stop the tsdb file scan operation. if (pTask->info.fillHistory == 1) { void* pExecutor = pTask->exec.pExecutor; - code = qKillTask(pExecutor, TSDB_CODE_SUCCESS); + code = qKillTask(pExecutor, TSDB_CODE_SUCCESS, 10000); } stDebug("vgId:%d s-task:%s set pause flag and pause task", pMeta->vgId, pTask->id.idStr); @@ -1287,6 +1287,8 @@ const char* streamTaskGetExecType(int32_t type) { return "resume-task-from-idle"; case STREAM_EXEC_T_ADD_FAILED_TASK: return "record-start-failed-task"; + case STREAM_EXEC_T_STOP_ONE_TASK: + return "stop-one-task"; case 0: return "exec-all-tasks"; default: diff --git a/source/util/src/tworker.c b/source/util/src/tworker.c index dbd8cb159e..469f98fcf0 100644 --- a/source/util/src/tworker.c +++ b/source/util/src/tworker.c @@ -256,7 +256,7 @@ static void *tAutoQWorkerThreadFp(SQueueWorker *worker) { return NULL; } -STaosQueue *tAutoQWorkerAllocQueue(SAutoQWorkerPool *pool, void *ahandle, FItem fp) { +STaosQueue *tAutoQWorkerAllocQueue(SAutoQWorkerPool *pool, void *ahandle, FItem fp, int32_t minNum) { int32_t code; STaosQueue *queue; @@ -280,7 +280,10 @@ STaosQueue *tAutoQWorkerAllocQueue(SAutoQWorkerPool *pool, void *ahandle, FItem int32_t queueNum = taosGetQueueNumber(pool->qset); int32_t curWorkerNum = taosArrayGetSize(pool->workers); int32_t dstWorkerNum = ceilf(queueNum * pool->ratio); - if (dstWorkerNum < 2) dstWorkerNum = 2; + + if (dstWorkerNum < minNum) { + dstWorkerNum = minNum; + } // spawn a thread to process queue while (curWorkerNum < dstWorkerNum) { diff --git a/tests/ci/func.txt b/tests/ci/func.txt index 45d4fb1c11..c724568537 100644 --- a/tests/ci/func.txt +++ b/tests/ci/func.txt @@ -79,7 +79,7 @@ (void)streamMetaAddFailedTask (void)streamMetaAddTaskLaunchResult (void)streamMetaCommit -(void)streamMetaRemoveTask +(void)streamMetaRemoveTaskInMeta (void)streamMetaSendHbHelper (void)streamMetaStartAllTasks (void)streamMetaStartOneTask diff --git a/tests/script/tsim/stream/tag.sim b/tests/script/tsim/stream/tag.sim index f293f4ac05..9f4c62e747 100644 --- a/tests/script/tsim/stream/tag.sim +++ b/tests/script/tsim/stream/tag.sim @@ -26,7 +26,7 @@ sql insert into t1 values(1648791223000,0,1,1,1.0); sql insert into t1 values(1648791223001,9,2,2,1.1); sql insert into t1 values(1648791223009,0,3,3,1.0); -sleep 300 +sleep 1000 sql select * from streamt; if $data01 != 3 then