/* * Copyright (c) 2019 TAOS Data, Inc. * * This program is free software: you can use, redistribute, and/or modify * it under the terms of the GNU Affero General Public License, version 3 * or later ("AGPL"), as published by the Free Software Foundation. * * This program is distributed in the hope that it will be useful, but WITHOUT * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or * FITNESS FOR A PARTICULAR PURPOSE. * * You should have received a copy of the GNU Affero General Public License * along with this program. If not, see . */ #include "trpc.h" #include "query.h" #include "tname.h" #include "catalogInt.h" #include "systable.h" SCtgOperation gCtgCacheOperation[CTG_OP_MAX] = { { CTG_OP_UPDATE_VGROUP, "update vgInfo", ctgOpUpdateVgroup }, { CTG_OP_UPDATE_TB_META, "update tbMeta", ctgOpUpdateTbMeta }, { CTG_OP_DROP_DB_CACHE, "drop DB", ctgOpDropDbCache }, { CTG_OP_DROP_DB_VGROUP, "drop DBVgroup", ctgOpDropDbVgroup }, { CTG_OP_DROP_STB_META, "drop stbMeta", ctgOpDropStbMeta }, { CTG_OP_DROP_TB_META, "drop tbMeta", ctgOpDropTbMeta }, { CTG_OP_UPDATE_USER, "update user", ctgOpUpdateUser }, { CTG_OP_UPDATE_VG_EPSET, "update epset", ctgOpUpdateEpset }, { CTG_OP_UPDATE_TB_INDEX, "update tbIndex", ctgOpUpdateTbIndex } }; int32_t ctgRLockVgInfo(SCatalog *pCtg, SCtgDBCache *dbCache, bool *inCache) { CTG_LOCK(CTG_READ, &dbCache->vgCache.vgLock); if (dbCache->deleted) { CTG_UNLOCK(CTG_READ, &dbCache->vgCache.vgLock); ctgDebug("db is dropping, dbId:%"PRIx64, dbCache->dbId); *inCache = false; return TSDB_CODE_SUCCESS; } if (NULL == dbCache->vgCache.vgInfo) { CTG_UNLOCK(CTG_READ, &dbCache->vgCache.vgLock); *inCache = false; ctgDebug("db vgInfo is empty, dbId:%"PRIx64, dbCache->dbId); return TSDB_CODE_SUCCESS; } *inCache = true; return TSDB_CODE_SUCCESS; } int32_t ctgWLockVgInfo(SCatalog *pCtg, SCtgDBCache *dbCache) { CTG_LOCK(CTG_WRITE, &dbCache->vgCache.vgLock); if (dbCache->deleted) { ctgDebug("db is dropping, dbId:%"PRIx64, dbCache->dbId); CTG_UNLOCK(CTG_WRITE, &dbCache->vgCache.vgLock); CTG_ERR_RET(TSDB_CODE_CTG_DB_DROPPED); } return TSDB_CODE_SUCCESS; } void ctgRUnlockVgInfo(SCtgDBCache *dbCache) { CTG_UNLOCK(CTG_READ, &dbCache->vgCache.vgLock); } void ctgWUnlockVgInfo(SCtgDBCache *dbCache) { CTG_UNLOCK(CTG_WRITE, &dbCache->vgCache.vgLock); } void ctgReleaseDBCache(SCatalog *pCtg, SCtgDBCache *dbCache) { CTG_UNLOCK(CTG_READ, &dbCache->dbLock); } int32_t ctgAcquireDBCacheImpl(SCatalog* pCtg, const char *dbFName, SCtgDBCache **pCache, bool acquire) { char *p = strchr(dbFName, '.'); if (p && CTG_IS_SYS_DBNAME(p + 1)) { dbFName = p + 1; } SCtgDBCache *dbCache = (SCtgDBCache *)taosHashGet(pCtg->dbCache, dbFName, strlen(dbFName)); if (NULL == dbCache) { *pCache = NULL; ctgDebug("db not in cache, dbFName:%s", dbFName); return TSDB_CODE_SUCCESS; } if (acquire) { CTG_LOCK(CTG_READ, &dbCache->dbLock); } if (dbCache->deleted) { if (acquire) { ctgReleaseDBCache(pCtg, dbCache); } *pCache = NULL; ctgDebug("db is removing from cache, dbFName:%s", dbFName); return TSDB_CODE_SUCCESS; } *pCache = dbCache; return TSDB_CODE_SUCCESS; } int32_t ctgAcquireDBCache(SCatalog* pCtg, const char *dbFName, SCtgDBCache **pCache) { CTG_RET(ctgAcquireDBCacheImpl(pCtg, dbFName, pCache, true)); } int32_t ctgGetDBCache(SCatalog* pCtg, const char *dbFName, SCtgDBCache **pCache) { CTG_RET(ctgAcquireDBCacheImpl(pCtg, dbFName, pCache, false)); } void ctgReleaseVgInfoToCache(SCatalog* pCtg, SCtgDBCache *dbCache) { ctgRUnlockVgInfo(dbCache); ctgReleaseDBCache(pCtg, dbCache); } void ctgReleaseTbMetaToCache(SCatalog* pCtg, SCtgDBCache *dbCache, SCtgTbCache* pCache) { if (pCache) { CTG_UNLOCK(CTG_READ, &pCache->metaLock); taosHashRelease(dbCache->tbCache, pCache); } if (dbCache) { ctgReleaseDBCache(pCtg, dbCache); } } void ctgReleaseTbIndexToCache(SCatalog* pCtg, SCtgDBCache *dbCache, SCtgTbCache* pCache) { if (pCache) { CTG_UNLOCK(CTG_READ, &pCache->indexLock); taosHashRelease(dbCache->tbCache, pCache); } if (dbCache) { ctgReleaseDBCache(pCtg, dbCache); } } int32_t ctgAcquireVgInfoFromCache(SCatalog* pCtg, const char *dbFName, SCtgDBCache **pCache) { SCtgDBCache *dbCache = NULL; ctgAcquireDBCache(pCtg, dbFName, &dbCache); if (NULL == dbCache) { ctgDebug("db %s not in cache", dbFName); goto _return; } bool inCache = false; ctgRLockVgInfo(pCtg, dbCache, &inCache); if (!inCache) { ctgDebug("vgInfo of db %s not in cache", dbFName); goto _return; } *pCache = dbCache; CTG_CACHE_STAT_INC(vgHitNum, 1); ctgDebug("Got db vgInfo from cache, dbFName:%s", dbFName); return TSDB_CODE_SUCCESS; _return: if (dbCache) { ctgReleaseDBCache(pCtg, dbCache); } *pCache = NULL; CTG_CACHE_STAT_INC(vgMissNum, 1); return TSDB_CODE_SUCCESS; } int32_t ctgAcquireTbMetaFromCache(SCatalog* pCtg, char *dbFName, char* tbName, SCtgDBCache **pDb, SCtgTbCache** pTb) { SCtgDBCache *dbCache = NULL; ctgAcquireDBCache(pCtg, dbFName, &dbCache); if (NULL == dbCache) { ctgDebug("db %s not in cache", dbFName); goto _return; } int32_t sz = 0; SCtgTbCache* pCache = taosHashAcquire(dbCache->tbCache, tbName, strlen(tbName)); if (NULL == pCache) { ctgDebug("tb %s not in cache, dbFName:%s", tbName, dbFName); goto _return; } CTG_LOCK(CTG_READ, &pCache->metaLock); if (NULL == pCache->pMeta) { ctgDebug("tb %s meta not in cache, dbFName:%s", tbName, dbFName); goto _return; } *pDb = dbCache; *pTb = pCache; ctgDebug("tb %s meta got in cache, dbFName:%s", tbName, dbFName); CTG_CACHE_STAT_INC(tbMetaHitNum, 1); return TSDB_CODE_SUCCESS; _return: ctgReleaseTbMetaToCache(pCtg, dbCache, pCache); CTG_CACHE_STAT_INC(tbMetaMissNum, 1); return TSDB_CODE_SUCCESS; } int32_t ctgAcquireTbIndexFromCache(SCatalog* pCtg, char *dbFName, char* tbName, SCtgDBCache **pDb, SCtgTbCache** pTb) { SCtgDBCache *dbCache = NULL; ctgAcquireDBCache(pCtg, dbFName, &dbCache); if (NULL == dbCache) { ctgDebug("db %s not in cache", dbFName); goto _return; } int32_t sz = 0; SCtgTbCache* pCache = taosHashAcquire(dbCache->tbCache, tbName, strlen(tbName)); if (NULL == pCache) { ctgDebug("tb %s not in cache, dbFName:%s", tbName, dbFName); goto _return; } CTG_LOCK(CTG_READ, &pCache->indexLock); if (NULL == pCache->pIndex) { ctgDebug("tb %s index not in cache, dbFName:%s", tbName, dbFName); goto _return; } *pDb = dbCache; *pTb = pCache; ctgDebug("tb %s index got in cache, dbFName:%s", tbName, dbFName); CTG_CACHE_STAT_INC(tbIndexHitNum, 1); return TSDB_CODE_SUCCESS; _return: ctgReleaseTbIndexToCache(pCtg, dbCache, pCache); CTG_CACHE_STAT_INC(tbIndexMissNum, 1); return TSDB_CODE_SUCCESS; } int32_t ctgTbMetaExistInCache(SCatalog* pCtg, char *dbFName, char* tbName, int32_t *exist) { SCtgDBCache *dbCache = NULL; ctgAcquireTbMetaFromCache(pCtg, dbFName, tbName, &dbCache, &tbCache); if (NULL == tbCache) { ctgReleaseTbMetaToCache(pCtg, dbCache, tbCache); *exist = 0; return TSDB_CODE_SUCCESS; } *exist = 1; ctgReleaseTbMetaToCache(pCtg, dbCache, tbCache); return TSDB_CODE_SUCCESS; } int32_t ctgReadTbMetaFromCache(SCatalog* pCtg, SCtgTbMetaCtx* ctx, STableMeta** pTableMeta) { int32_t code = 0; SCtgDBCache *dbCache = NULL; SCtgTbCache *tbCache = NULL; *pTableMeta = NULL; char dbFName[TSDB_DB_FNAME_LEN] = {0}; if (CTG_FLAG_IS_SYS_DB(ctx->flag)) { strcpy(dbFName, ctx->pName->dbname); } else { tNameGetFullDbName(ctx->pName, dbFName); } ctgAcquireTbMetaFromCache(pCtg, dbFName, ctx->pName->tname, &dbCache, &tbCache); if (NULL == tbCache) { ctgReleaseTbMetaToCache(pCtg, dbCache, tbCache); return TSDB_CODE_SUCCESS; } STableMeta* tbMeta = tbCache->pMeta; ctx->tbInfo.inCache = true; ctx->tbInfo.dbId = dbCache->dbId; ctx->tbInfo.suid = tbMeta->suid; ctx->tbInfo.tbType = tbMeta->tableType; if (tbMeta->tableType != TSDB_CHILD_TABLE) { ctgReleaseTbMetaToCache(pCtg, dbCache, tbCache); ctgDebug("Got tb %s meta from cache, type:%d, dbFName:%s", ctx->pName->tname, tbMeta->tableType, dbFName); return TSDB_CODE_SUCCESS; } STableMeta **stbMeta = taosHashGet(dbCache->stbCache, &tbMeta->suid, sizeof(tbMeta->suid)); if (NULL == stbMeta || NULL == *stbMeta) { ctgError("stb not in stbCache, suid:%"PRIx64, tbMeta->suid); goto _return; } if ((*stbMeta)->suid != tbMeta->suid) { ctgError("stb suid %" PRIx64 " in stbCache mis-match, expected suid:%"PRIx64 , (*stbMeta)->suid, tbMeta->suid); CTG_ERR_JRET(TSDB_CODE_CTG_INTERNAL_ERROR); } int32_t metaSize = CTG_META_SIZE(*stbMeta); *pTableMeta = taosMemoryRealloc(*pTableMeta, metaSize); if (NULL == *pTableMeta) { CTG_ERR_JRET(TSDB_CODE_CTG_MEM_ERROR); } memcpy(&(*pTableMeta)->sversion, &(*stbMeta)->sversion, metaSize - sizeof(SCTableMeta)); ctgReleaseTbMetaToCache(pCtg, dbCache, tbCache); ctgDebug("Got tb %s meta from cache, dbFName:%s", ctx->pName->tname, dbFName); return TSDB_CODE_SUCCESS; _return: ctgReleaseTbMetaToCache(pCtg, dbCache, tbCache); taosMemoryFreeClear(*pTableMeta); CTG_RET(code); } int32_t ctgReadTbVerFromCache(SCatalog *pCtg, const SName *pTableName, int32_t *sver, int32_t *tver, int32_t *tbType, uint64_t *suid, char *stbName) { *sver = -1; *tver = -1; SCtgDBCache *dbCache = NULL; SCtgTbCache *tbCache = NULL; char dbFName[TSDB_DB_FNAME_LEN] = {0}; tNameGetFullDbName(pTableName, dbFName); ctgAcquireTbMetaFromCache(pCtg, dbFName, pTableName->tname, &dbCache, &tbCache); if (NULL == tbCache) { ctgReleaseTbMetaToCache(pCtg, dbCache, tbCache); return TSDB_CODE_SUCCESS; } STableMeta* tbMeta = tbCache->pMeta; *tbType = tbMeta->tableType; *suid = tbMeta->suid; if (*tbType != TSDB_CHILD_TABLE) { *sver = tbMeta->sversion; *tver = tbMeta->tversion; ctgDebug("Got tb %s ver from cache, dbFName:%s, tbType:%d, sver:%d, tver:%d, suid:%" PRIx64, pTableName->tname, dbFName, *tbType, *sver, *tver, *suid); ctgReleaseTbMetaToCache(pCtg, dbCache, tbCache); return TSDB_CODE_SUCCESS; } STableMeta **stbMeta = taosHashGet(dbCache->stbCache, suid, sizeof(*suid)); if (NULL == stbMeta || NULL == *stbMeta) { ctgReleaseTbMetaToCache(pCtg, dbCache, tbCache); ctgDebug("stb not in stbCache, suid:%" PRIx64, *suid); return TSDB_CODE_SUCCESS; } if ((*stbMeta)->suid != *suid) { ctgReleaseTbMetaToCache(pCtg, dbCache, tbCache); ctgError("stb suid %" PRIx64 " in stbCache mis-match, expected suid:%" PRIx64 , (*stbMeta)->suid, *suid); CTG_ERR_RET(TSDB_CODE_CTG_INTERNAL_ERROR); } size_t nameLen = 0; char *name = taosHashGetKey(*stbMeta, &nameLen); strncpy(stbName, name, nameLen); stbName[nameLen] = 0; *sver = (*stbMeta)->sversion; *tver = (*stbMeta)->tversion; ctgReleaseTbMetaToCache(pCtg, dbCache, tbCache); ctgDebug("Got tb %s sver %d tver %d from cache, type:%d, dbFName:%s", pTableName->tname, *sver, *tver, *tbType, dbFName); return TSDB_CODE_SUCCESS; } int32_t ctgReadTbTypeFromCache(SCatalog* pCtg, const char* dbFName, const char *tableName, int32_t *tbType) { SCtgDBCache *dbCache = NULL; SCtgTbCache *tbCache = NULL; ctgAcquireTbMetaFromCache(pCtg, dbFName, tableName, &dbCache, &tbCache); if (NULL == tbCache) { ctgReleaseTbMetaToCache(pCtg, dbCache, tbCache); return TSDB_CODE_SUCCESS; } *tbType = tbCache->pMeta->tableType; ctgReleaseTbMetaToCache(pCtg, dbCache, tbCache); ctgDebug("Got tb %s tbType %d from cache, dbFName:%s", tableName, *tbType, dbFName); return TSDB_CODE_SUCCESS; } int32_t ctgReadTbIndexFromCache(SCatalog* pCtg, const SName* pTableName, SArray** pRes) { int32_t code = 0; SCtgDBCache *dbCache = NULL; SCtgTbCache *tbCache = NULL; char dbFName[TSDB_DB_FNAME_LEN] = {0}; tNameGetFullDbName(pTableName, dbFName); *pRes = NULL; ctgAcquireTbIndexFromCache(pCtg, dbFName, pTableName->tname, &dbCache, &tbCache); if (NULL == tbCache) { ctgReleaseTbIndexToCache(pCtg, dbCache, tbCache); return TSDB_CODE_SUCCESS; } CTG_ERR_JRET(ctgCloneTableIndex(tbCache->pIndex->pIndex, pRes)); _return: ctgReleaseTbIndexToCache(pCtg, dbCache, tbCache); CTG_RET(code); } int32_t ctgChkAuthFromCache(SCatalog* pCtg, const char* user, const char* dbFName, AUTH_TYPE type, bool *inCache, bool *pass) { if (NULL == pCtg->userCache) { ctgDebug("empty user auth cache, user:%s", user); goto _return; } SCtgUserAuth *pUser = (SCtgUserAuth *)taosHashGet(pCtg->userCache, user, strlen(user)); if (NULL == pUser) { ctgDebug("user not in cache, user:%s", user); goto _return; } *inCache = true; ctgDebug("Got user from cache, user:%s", user); CTG_CACHE_STAT_INC(userHitNum, 1); if (pUser->superUser) { *pass = true; return TSDB_CODE_SUCCESS; } CTG_LOCK(CTG_READ, &pUser->lock); if (pUser->createdDbs && taosHashGet(pUser->createdDbs, dbFName, strlen(dbFName))) { *pass = true; CTG_UNLOCK(CTG_READ, &pUser->lock); return TSDB_CODE_SUCCESS; } if (pUser->readDbs && taosHashGet(pUser->readDbs, dbFName, strlen(dbFName)) && type == AUTH_TYPE_READ) { *pass = true; } if (pUser->writeDbs && taosHashGet(pUser->writeDbs, dbFName, strlen(dbFName)) && type == AUTH_TYPE_WRITE) { *pass = true; } CTG_UNLOCK(CTG_READ, &pUser->lock); return TSDB_CODE_SUCCESS; _return: *inCache = false; CTG_CACHE_STAT_INC(userMissNum, 1); return TSDB_CODE_SUCCESS; } void ctgDequeue(SCtgCacheOperation **op) { SCtgQNode *orig = gCtgMgmt.queue.head; SCtgQNode *node = gCtgMgmt.queue.head->next; gCtgMgmt.queue.head = gCtgMgmt.queue.head->next; CTG_QUEUE_DEC(); taosMemoryFreeClear(orig); *op = node->op; } int32_t ctgEnqueue(SCatalog* pCtg, SCtgCacheOperation *operation) { SCtgQNode *node = taosMemoryCalloc(1, sizeof(SCtgQNode)); if (NULL == node) { qError("calloc %d failed", (int32_t)sizeof(SCtgQNode)); CTG_RET(TSDB_CODE_CTG_MEM_ERROR); } if (operation->syncOp) { tsem_init(&operation->rspSem, 0, 0); } node->op = operation; CTG_LOCK(CTG_WRITE, &gCtgMgmt.queue.qlock); gCtgMgmt.queue.tail->next = node; gCtgMgmt.queue.tail = node; CTG_UNLOCK(CTG_WRITE, &gCtgMgmt.queue.qlock); CTG_QUEUE_INC(); CTG_RT_STAT_INC(qNum, 1); tsem_post(&gCtgMgmt.queue.reqSem); ctgDebug("action [%s] added into queue", gCtgCacheOperation[operation->opId].name); if (operation->syncOp) { tsem_wait(&operation->rspSem); taosMemoryFree(operation); } return TSDB_CODE_SUCCESS; } int32_t ctgDropDbCacheEnqueue(SCatalog* pCtg, const char *dbFName, int64_t dbId) { int32_t code = 0; SCtgCacheOperation *op = taosMemoryCalloc(1, sizeof(SCtgCacheOperation)); op->opId = CTG_OP_DROP_DB_CACHE; SCtgDropDBMsg *msg = taosMemoryMalloc(sizeof(SCtgDropDBMsg)); if (NULL == msg) { ctgError("malloc %d failed", (int32_t)sizeof(SCtgDropDBMsg)); CTG_ERR_RET(TSDB_CODE_CTG_MEM_ERROR); } char *p = strchr(dbFName, '.'); if (p && CTG_IS_SYS_DBNAME(p + 1)) { dbFName = p + 1; } msg->pCtg = pCtg; strncpy(msg->dbFName, dbFName, sizeof(msg->dbFName)); msg->dbId = dbId; op->data = msg; CTG_ERR_JRET(ctgEnqueue(pCtg, op)); return TSDB_CODE_SUCCESS; _return: taosMemoryFreeClear(op->data); CTG_RET(code); } int32_t ctgDropDbVgroupEnqueue(SCatalog* pCtg, const char *dbFName, bool syncOp) { int32_t code = 0; SCtgCacheOperation *op = taosMemoryCalloc(1, sizeof(SCtgCacheOperation)); op->opId = CTG_OP_DROP_DB_VGROUP; op->syncOp = syncOp; SCtgDropDbVgroupMsg *msg = taosMemoryMalloc(sizeof(SCtgDropDbVgroupMsg)); if (NULL == msg) { ctgError("malloc %d failed", (int32_t)sizeof(SCtgDropDbVgroupMsg)); CTG_ERR_RET(TSDB_CODE_CTG_MEM_ERROR); } char *p = strchr(dbFName, '.'); if (p && CTG_IS_SYS_DBNAME(p + 1)) { dbFName = p + 1; } msg->pCtg = pCtg; strncpy(msg->dbFName, dbFName, sizeof(msg->dbFName)); op->data = msg; CTG_ERR_JRET(ctgEnqueue(pCtg, op)); return TSDB_CODE_SUCCESS; _return: taosMemoryFreeClear(op->data); CTG_RET(code); } int32_t ctgDropStbMetaEnqueue(SCatalog* pCtg, const char *dbFName, int64_t dbId, const char *stbName, uint64_t suid, bool syncOp) { int32_t code = 0; SCtgCacheOperation *op = taosMemoryCalloc(1, sizeof(SCtgCacheOperation)); op->opId = CTG_OP_DROP_STB_META; op->syncOp = syncOp; SCtgDropStbMetaMsg *msg = taosMemoryMalloc(sizeof(SCtgDropStbMetaMsg)); if (NULL == msg) { ctgError("malloc %d failed", (int32_t)sizeof(SCtgDropStbMetaMsg)); CTG_ERR_RET(TSDB_CODE_CTG_MEM_ERROR); } msg->pCtg = pCtg; strncpy(msg->dbFName, dbFName, sizeof(msg->dbFName)); strncpy(msg->stbName, stbName, sizeof(msg->stbName)); msg->dbId = dbId; msg->suid = suid; op->data = msg; CTG_ERR_JRET(ctgEnqueue(pCtg, op)); return TSDB_CODE_SUCCESS; _return: taosMemoryFreeClear(op->data); CTG_RET(code); } int32_t ctgDropTbMetaEnqueue(SCatalog* pCtg, const char *dbFName, int64_t dbId, const char *tbName, bool syncOp) { int32_t code = 0; SCtgCacheOperation *op = taosMemoryCalloc(1, sizeof(SCtgCacheOperation)); op->opId = CTG_OP_DROP_TB_META; op->syncOp = syncOp; SCtgDropTblMetaMsg *msg = taosMemoryMalloc(sizeof(SCtgDropTblMetaMsg)); if (NULL == msg) { ctgError("malloc %d failed", (int32_t)sizeof(SCtgDropTblMetaMsg)); CTG_ERR_RET(TSDB_CODE_CTG_MEM_ERROR); } msg->pCtg = pCtg; strncpy(msg->dbFName, dbFName, sizeof(msg->dbFName)); strncpy(msg->tbName, tbName, sizeof(msg->tbName)); msg->dbId = dbId; op->data = msg; CTG_ERR_JRET(ctgEnqueue(pCtg, op)); return TSDB_CODE_SUCCESS; _return: taosMemoryFreeClear(op->data); CTG_RET(code); } int32_t ctgUpdateVgroupEnqueue(SCatalog* pCtg, const char *dbFName, int64_t dbId, SDBVgInfo* dbInfo, bool syncOp) { int32_t code = 0; SCtgCacheOperation *op = taosMemoryCalloc(1, sizeof(SCtgCacheOperation)); op->opId = CTG_OP_UPDATE_VGROUP; op->syncOp = syncOp; SCtgUpdateVgMsg *msg = taosMemoryMalloc(sizeof(SCtgUpdateVgMsg)); if (NULL == msg) { ctgError("malloc %d failed", (int32_t)sizeof(SCtgUpdateVgMsg)); ctgFreeVgInfo(dbInfo); CTG_ERR_RET(TSDB_CODE_CTG_MEM_ERROR); } char *p = strchr(dbFName, '.'); if (p && CTG_IS_SYS_DBNAME(p + 1)) { dbFName = p + 1; } strncpy(msg->dbFName, dbFName, sizeof(msg->dbFName)); msg->pCtg = pCtg; msg->dbId = dbId; msg->dbInfo = dbInfo; op->data = msg; CTG_ERR_JRET(ctgEnqueue(pCtg, op)); return TSDB_CODE_SUCCESS; _return: ctgFreeVgInfo(dbInfo); taosMemoryFreeClear(op->data); CTG_RET(code); } int32_t ctgUpdateTbMetaEnqueue(SCatalog* pCtg, STableMetaOutput *output, bool syncOp) { int32_t code = 0; SCtgCacheOperation *op = taosMemoryCalloc(1, sizeof(SCtgCacheOperation)); op->opId = CTG_OP_UPDATE_TB_META; op->syncOp = syncOp; SCtgUpdateTbMetaMsg *msg = taosMemoryMalloc(sizeof(SCtgUpdateTbMetaMsg)); if (NULL == msg) { ctgError("malloc %d failed", (int32_t)sizeof(SCtgUpdateTbMetaMsg)); CTG_ERR_RET(TSDB_CODE_CTG_MEM_ERROR); } char *p = strchr(output->dbFName, '.'); if (p && CTG_IS_SYS_DBNAME(p + 1)) { memmove(output->dbFName, p + 1, strlen(p + 1)); } msg->pCtg = pCtg; msg->pMeta = output; op->data = msg; CTG_ERR_JRET(ctgEnqueue(pCtg, op)); return TSDB_CODE_SUCCESS; _return: taosMemoryFreeClear(msg); CTG_RET(code); } int32_t ctgUpdateVgEpsetEnqueue(SCatalog* pCtg, char *dbFName, int32_t vgId, SEpSet* pEpSet) { int32_t code = 0; SCtgCacheOperation *op = taosMemoryCalloc(1, sizeof(SCtgCacheOperation)); op->opId = CTG_OP_UPDATE_VG_EPSET; SCtgUpdateEpsetMsg *msg = taosMemoryMalloc(sizeof(SCtgUpdateEpsetMsg)); if (NULL == msg) { ctgError("malloc %d failed", (int32_t)sizeof(SCtgUpdateEpsetMsg)); CTG_ERR_RET(TSDB_CODE_CTG_MEM_ERROR); } msg->pCtg = pCtg; strcpy(msg->dbFName, dbFName); msg->vgId = vgId; msg->epSet = *pEpSet; op->data = msg; CTG_ERR_JRET(ctgEnqueue(pCtg, op)); return TSDB_CODE_SUCCESS; _return: taosMemoryFreeClear(msg); CTG_RET(code); } int32_t ctgUpdateUserEnqueue(SCatalog* pCtg, SGetUserAuthRsp *pAuth, bool syncOp) { int32_t code = 0; SCtgCacheOperation *op = taosMemoryCalloc(1, sizeof(SCtgCacheOperation)); op->opId = CTG_OP_UPDATE_USER; op->syncOp = syncOp; SCtgUpdateUserMsg *msg = taosMemoryMalloc(sizeof(SCtgUpdateUserMsg)); if (NULL == msg) { ctgError("malloc %d failed", (int32_t)sizeof(SCtgUpdateUserMsg)); CTG_ERR_RET(TSDB_CODE_CTG_MEM_ERROR); } msg->pCtg = pCtg; msg->userAuth = *pAuth; op->data = msg; CTG_ERR_JRET(ctgEnqueue(pCtg, op)); return TSDB_CODE_SUCCESS; _return: tFreeSGetUserAuthRsp(pAuth); taosMemoryFreeClear(msg); CTG_RET(code); } int32_t ctgUpdateTbIndexEnqueue(SCatalog* pCtg, SName* pName, STableIndex *pIndex, bool syncOp) { int32_t code = 0; SCtgCacheOperation *op = taosMemoryCalloc(1, sizeof(SCtgCacheOperation)); op->opId = CTG_OP_UPDATE_TB_INDEX; op->syncOp = syncOp; SCtgUpdateTbIndexMsg *msg = taosMemoryMalloc(sizeof(SCtgUpdateTbIndexMsg)); if (NULL == msg) { ctgError("malloc %d failed", (int32_t)sizeof(SCtgUpdateTbIndexMsg)); CTG_ERR_JRET(TSDB_CODE_CTG_MEM_ERROR); } msg->pCtg = pCtg; msg->pIndex = pIndex; tNameGetFullDbName(pName, msg->dbFName); strcpy(msg->tbName, pName->tname); op->data = msg; CTG_ERR_JRET(ctgEnqueue(pCtg, op)); return TSDB_CODE_SUCCESS; _return: taosArrayDestroyEx(pIndex, tFreeSTableIndexInfo); taosMemoryFreeClear(msg); CTG_RET(code); } int32_t ctgDropTbIndexEnqueue(SCatalog* pCtg, SName* pName, bool syncOp) { int32_t code = 0; SCtgCacheOperation *op = taosMemoryCalloc(1, sizeof(SCtgCacheOperation)); op->opId = CTG_OP_DROP_TB_INDEX; op->syncOp = syncOp; SCtgDropTbIndexMsg *msg = taosMemoryMalloc(sizeof(SCtgDropTbIndexMsg)); if (NULL == msg) { ctgError("malloc %d failed", (int32_t)sizeof(SCtgDropTbIndexMsg)); CTG_ERR_RET(TSDB_CODE_CTG_MEM_ERROR); } msg->pCtg = pCtg; tNameGetFullDbName(pName, msg->dbFName); strcpy(msg->tbName, pName->tname); op->data = msg; CTG_ERR_JRET(ctgEnqueue(pCtg, op)); return TSDB_CODE_SUCCESS; _return: taosArrayDestroyEx(pIndex, tFreeSTableIndexInfo); taosMemoryFreeClear(msg); CTG_RET(code); } int32_t ctgMetaRentInit(SCtgRentMgmt *mgmt, uint32_t rentSec, int8_t type) { mgmt->slotRIdx = 0; mgmt->slotNum = rentSec / CTG_RENT_SLOT_SECOND; mgmt->type = type; size_t msgSize = sizeof(SCtgRentSlot) * mgmt->slotNum; mgmt->slots = taosMemoryCalloc(1, msgSize); if (NULL == mgmt->slots) { qError("calloc %d failed", (int32_t)msgSize); CTG_ERR_RET(TSDB_CODE_CTG_MEM_ERROR); } qDebug("meta rent initialized, type:%d, slotNum:%d", type, mgmt->slotNum); return TSDB_CODE_SUCCESS; } int32_t ctgMetaRentAdd(SCtgRentMgmt *mgmt, void *meta, int64_t id, int32_t size) { int16_t widx = abs((int)(id % mgmt->slotNum)); SCtgRentSlot *slot = &mgmt->slots[widx]; int32_t code = 0; CTG_LOCK(CTG_WRITE, &slot->lock); if (NULL == slot->meta) { slot->meta = taosArrayInit(CTG_DEFAULT_RENT_SLOT_SIZE, size); if (NULL == slot->meta) { qError("taosArrayInit %d failed, id:%"PRIx64", slot idx:%d, type:%d", CTG_DEFAULT_RENT_SLOT_SIZE, id, widx, mgmt->type); CTG_ERR_JRET(TSDB_CODE_CTG_MEM_ERROR); } } if (NULL == taosArrayPush(slot->meta, meta)) { qError("taosArrayPush meta to rent failed, id:%"PRIx64", slot idx:%d, type:%d", id, widx, mgmt->type); CTG_ERR_JRET(TSDB_CODE_CTG_MEM_ERROR); } slot->needSort = true; qDebug("add meta to rent, id:%"PRIx64", slot idx:%d, type:%d", id, widx, mgmt->type); _return: CTG_UNLOCK(CTG_WRITE, &slot->lock); CTG_RET(code); } int32_t ctgMetaRentUpdate(SCtgRentMgmt *mgmt, void *meta, int64_t id, int32_t size, __compar_fn_t sortCompare, __compar_fn_t searchCompare) { int16_t widx = abs((int)(id % mgmt->slotNum)); SCtgRentSlot *slot = &mgmt->slots[widx]; int32_t code = 0; CTG_LOCK(CTG_WRITE, &slot->lock); if (NULL == slot->meta) { qError("empty meta slot, id:%"PRIx64", slot idx:%d, type:%d", id, widx, mgmt->type); CTG_ERR_JRET(TSDB_CODE_CTG_INTERNAL_ERROR); } if (slot->needSort) { qDebug("meta slot before sorte, slot idx:%d, type:%d, size:%d", widx, mgmt->type, (int32_t)taosArrayGetSize(slot->meta)); taosArraySort(slot->meta, sortCompare); slot->needSort = false; qDebug("meta slot sorted, slot idx:%d, type:%d, size:%d", widx, mgmt->type, (int32_t)taosArrayGetSize(slot->meta)); } void *orig = taosArraySearch(slot->meta, &id, searchCompare, TD_EQ); if (NULL == orig) { qError("meta not found in slot, id:%"PRIx64", slot idx:%d, type:%d, size:%d", id, widx, mgmt->type, (int32_t)taosArrayGetSize(slot->meta)); CTG_ERR_JRET(TSDB_CODE_CTG_INTERNAL_ERROR); } memcpy(orig, meta, size); qDebug("meta in rent updated, id:%"PRIx64", slot idx:%d, type:%d", id, widx, mgmt->type); _return: CTG_UNLOCK(CTG_WRITE, &slot->lock); if (code) { qWarn("meta in rent update failed, will try to add it, code:%x, id:%"PRIx64", slot idx:%d, type:%d", code, id, widx, mgmt->type); CTG_RET(ctgMetaRentAdd(mgmt, meta, id, size)); } CTG_RET(code); } int32_t ctgMetaRentRemove(SCtgRentMgmt *mgmt, int64_t id, __compar_fn_t sortCompare, __compar_fn_t searchCompare) { int16_t widx = abs((int)(id % mgmt->slotNum)); SCtgRentSlot *slot = &mgmt->slots[widx]; int32_t code = 0; CTG_LOCK(CTG_WRITE, &slot->lock); if (NULL == slot->meta) { qError("empty meta slot, id:%"PRIx64", slot idx:%d, type:%d", id, widx, mgmt->type); CTG_ERR_JRET(TSDB_CODE_CTG_INTERNAL_ERROR); } if (slot->needSort) { taosArraySort(slot->meta, sortCompare); slot->needSort = false; qDebug("meta slot sorted, slot idx:%d, type:%d", widx, mgmt->type); } int32_t idx = taosArraySearchIdx(slot->meta, &id, searchCompare, TD_EQ); if (idx < 0) { qError("meta not found in slot, id:%"PRIx64", slot idx:%d, type:%d", id, widx, mgmt->type); CTG_ERR_JRET(TSDB_CODE_CTG_INTERNAL_ERROR); } taosArrayRemove(slot->meta, idx); qDebug("meta in rent removed, id:%"PRIx64", slot idx:%d, type:%d", id, widx, mgmt->type); _return: CTG_UNLOCK(CTG_WRITE, &slot->lock); CTG_RET(code); } int32_t ctgMetaRentGetImpl(SCtgRentMgmt *mgmt, void **res, uint32_t *num, int32_t size) { int16_t ridx = atomic_add_fetch_16(&mgmt->slotRIdx, 1); if (ridx >= mgmt->slotNum) { ridx %= mgmt->slotNum; atomic_store_16(&mgmt->slotRIdx, ridx); } SCtgRentSlot *slot = &mgmt->slots[ridx]; int32_t code = 0; CTG_LOCK(CTG_READ, &slot->lock); if (NULL == slot->meta) { qDebug("empty meta in slot:%d, type:%d", ridx, mgmt->type); *num = 0; goto _return; } size_t metaNum = taosArrayGetSize(slot->meta); if (metaNum <= 0) { qDebug("no meta in slot:%d, type:%d", ridx, mgmt->type); *num = 0; goto _return; } size_t msize = metaNum * size; *res = taosMemoryMalloc(msize); if (NULL == *res) { qError("malloc %d failed", (int32_t)msize); CTG_ERR_JRET(TSDB_CODE_CTG_MEM_ERROR); } void *meta = taosArrayGet(slot->meta, 0); memcpy(*res, meta, msize); *num = (uint32_t)metaNum; qDebug("Got %d meta from rent, type:%d", (int32_t)metaNum, mgmt->type); _return: CTG_UNLOCK(CTG_READ, &slot->lock); CTG_RET(code); } int32_t ctgMetaRentGet(SCtgRentMgmt *mgmt, void **res, uint32_t *num, int32_t size) { while (true) { int64_t msec = taosGetTimestampMs(); int64_t lsec = atomic_load_64(&mgmt->lastReadMsec); if ((msec - lsec) < CTG_RENT_SLOT_SECOND * 1000) { *res = NULL; *num = 0; qDebug("too short time period to get expired meta, type:%d", mgmt->type); return TSDB_CODE_SUCCESS; } if (lsec != atomic_val_compare_exchange_64(&mgmt->lastReadMsec, lsec, msec)) { continue; } break; } CTG_ERR_RET(ctgMetaRentGetImpl(mgmt, res, num, size)); return TSDB_CODE_SUCCESS; } int32_t ctgAddNewDBCache(SCatalog *pCtg, const char *dbFName, uint64_t dbId) { int32_t code = 0; SCtgDBCache newDBCache = {0}; newDBCache.dbId = dbId; newDBCache.tbCache = taosHashInit(gCtgMgmt.cfg.maxTblCacheNum, taosGetDefaultHashFunction(TSDB_DATA_TYPE_BINARY), true, HASH_ENTRY_LOCK); if (NULL == newDBCache.tbCache) { ctgError("taosHashInit %d metaCache failed", gCtgMgmt.cfg.maxTblCacheNum); CTG_ERR_RET(TSDB_CODE_CTG_MEM_ERROR); } newDBCache.stbCache = taosHashInit(gCtgMgmt.cfg.maxTblCacheNum, taosGetDefaultHashFunction(TSDB_DATA_TYPE_UBIGINT), true, HASH_ENTRY_LOCK); if (NULL == newDBCache.stbCache) { ctgError("taosHashInit %d stbCache failed", gCtgMgmt.cfg.maxTblCacheNum); CTG_ERR_JRET(TSDB_CODE_CTG_MEM_ERROR); } code = taosHashPut(pCtg->dbCache, dbFName, strlen(dbFName), &newDBCache, sizeof(SCtgDBCache)); if (code) { if (HASH_NODE_EXIST(code)) { ctgDebug("db already in cache, dbFName:%s", dbFName); goto _return; } ctgError("taosHashPut db to cache failed, dbFName:%s", dbFName); CTG_ERR_JRET(TSDB_CODE_CTG_MEM_ERROR); } CTG_CACHE_STAT_INC(dbNum, 1); SDbVgVersion vgVersion = {.dbId = newDBCache.dbId, .vgVersion = -1}; strncpy(vgVersion.dbFName, dbFName, sizeof(vgVersion.dbFName)); ctgDebug("db added to cache, dbFName:%s, dbId:%"PRIx64, dbFName, dbId); CTG_ERR_RET(ctgMetaRentAdd(&pCtg->dbRent, &vgVersion, dbId, sizeof(SDbVgVersion))); ctgDebug("db added to rent, dbFName:%s, vgVersion:%d, dbId:%"PRIx64, dbFName, vgVersion.vgVersion, dbId); return TSDB_CODE_SUCCESS; _return: ctgFreeDbCache(&newDBCache); CTG_RET(code); } void ctgRemoveStbRent(SCatalog* pCtg, SCtgDBCache *dbCache) { if (NULL == dbCache->stbCache) { return; } void *pIter = taosHashIterate(dbCache->stbCache, NULL); while (pIter) { uint64_t *suid = NULL; suid = taosHashGetKey(pIter, NULL); if (TSDB_CODE_SUCCESS == ctgMetaRentRemove(&pCtg->stbRent, *suid, ctgStbVersionSortCompare, ctgStbVersionSearchCompare)) { ctgDebug("stb removed from rent, suid:%"PRIx64, *suid); } pIter = taosHashIterate(dbCache->stbCache, pIter); } } int32_t ctgRemoveDBFromCache(SCatalog* pCtg, SCtgDBCache *dbCache, const char* dbFName) { uint64_t dbId = dbCache->dbId; ctgInfo("start to remove db from cache, dbFName:%s, dbId:%"PRIx64, dbFName, dbCache->dbId); CTG_LOCK(CTG_WRITE, &dbCache->dbLock); atomic_store_8(&dbCache->deleted, 1); ctgRemoveStbRent(pCtg, &dbCache); ctgFreeDbCache(dbCache); CTG_UNLOCK(CTG_WRITE, &dbCache->dbLock); CTG_ERR_RET(ctgMetaRentRemove(&pCtg->dbRent, dbId, ctgDbVgVersionSortCompare, ctgDbVgVersionSearchCompare)); ctgDebug("db removed from rent, dbFName:%s, dbId:%"PRIx64, dbFName, dbId); if (taosHashRemove(pCtg->dbCache, dbFName, strlen(dbFName))) { ctgInfo("taosHashRemove from dbCache failed, may be removed, dbFName:%s", dbFName); CTG_ERR_RET(TSDB_CODE_CTG_DB_DROPPED); } CTG_CACHE_STAT_DEC(dbNum, 1); ctgInfo("db removed from cache, dbFName:%s, dbId:%"PRIx64, dbFName, dbId); return TSDB_CODE_SUCCESS; } int32_t ctgGetAddDBCache(SCatalog* pCtg, const char *dbFName, uint64_t dbId, SCtgDBCache **pCache) { int32_t code = 0; SCtgDBCache *dbCache = NULL; ctgGetDBCache(pCtg, dbFName, &dbCache); if (dbCache) { // TODO OPEN IT #if 0 if (dbCache->dbId == dbId) { *pCache = dbCache; return TSDB_CODE_SUCCESS; } #else if (0 == dbId) { *pCache = dbCache; return TSDB_CODE_SUCCESS; } if (dbId && (dbCache->dbId == 0)) { dbCache->dbId = dbId; *pCache = dbCache; return TSDB_CODE_SUCCESS; } if (dbCache->dbId == dbId) { *pCache = dbCache; return TSDB_CODE_SUCCESS; } #endif CTG_ERR_RET(ctgRemoveDBFromCache(pCtg, dbCache, dbFName)); } CTG_ERR_RET(ctgAddNewDBCache(pCtg, dbFName, dbId)); ctgGetDBCache(pCtg, dbFName, &dbCache); *pCache = dbCache; return TSDB_CODE_SUCCESS; } int32_t ctgWriteTbMetaToCache(SCatalog *pCtg, SCtgDBCache *dbCache, char *dbFName, uint64_t dbId, char *tbName, STableMeta *meta, int32_t metaSize) { if (NULL == dbCache->tbCache || NULL == dbCache->stbCache) { taosMemoryFree(meta); ctgError("db is dropping, dbId:%"PRIx64, dbCache->dbId); CTG_ERR_RET(TSDB_CODE_CTG_DB_DROPPED); } bool isStb = meta->tableType == TSDB_SUPER_TABLE; SCtgTbCache* pCache = taosHashGet(dbCache->tbCache, tbName, strlen(tbName)); STableMeta *orig = (pCache ? pCache->pMeta : NULL); int8_t origType = 0; uint64_t origSuid = 0; if (orig) { origType = orig->tableType; if (origType == meta->tableType && orig->uid == meta->uid && orig->sversion >= meta->sversion && orig->tversion >= meta->tversion) { taosMemoryFree(meta); ctgDebug("ignore table %s meta update", tbName); return TSDB_CODE_SUCCESS; } if (origType == TSDB_SUPER_TABLE) { if (taosHashRemove(dbCache->stbCache, &orig->suid, sizeof(orig->suid))) { ctgError("stb not exist in stbCache, dbFName:%s, stb:%s, suid:%"PRIx64, dbFName, tbName, orig->suid); } else { CTG_CACHE_STAT_DEC(stblNum, 1); ctgDebug("stb removed from stbCache, dbFName:%s, stb:%s, suid:%"PRIx64, dbFName, tbName, orig->suid); } ctgMetaRentRemove(&pCtg->stbRent, orig->suid, ctgStbVersionSortCompare, ctgStbVersionSearchCompare); origSuid = orig->suid; } } if (NULL == pCache) { SCtgTbCache cache = {0}; cache.pMeta = meta; if (taosHashPut(dbCache->tbCache, tbName, strlen(tbName), &cache, sizeof(SCtgTbCache)) != 0) { taosMemoryFree(meta); ctgError("taosHashPut new tbCache failed, dbFName:%s, tbName:%s, tbType:%d", dbFName, tbName, meta->tableType); CTG_ERR_RET(TSDB_CODE_CTG_MEM_ERROR); } pCache = taosHashGet(dbCache->tbCache, tbName, strlen(tbName)); } else { taosMemoryFree(pCache->pMeta); pCache->pMeta = meta; } if (NULL == orig) { CTG_CACHE_STAT_INC(tblNum, 1); } ctgDebug("tbmeta updated to cache, dbFName:%s, tbName:%s, tbType:%d", dbFName, tbName, meta->tableType); ctgdShowTableMeta(pCtg, tbName, meta); if (!isStb) { return TSDB_CODE_SUCCESS; } if (origSuid != meta->suid && taosHashPut(dbCache->stbCache, &meta->suid, sizeof(meta->suid), &pCache, POINTER_BYTES) != 0) { ctgError("taosHashPut to stable cache failed, suid:%"PRIx64, meta->suid); CTG_ERR_RET(TSDB_CODE_CTG_MEM_ERROR); } CTG_CACHE_STAT_INC(stblNum, 1); ctgDebug("stb updated to stbCache, dbFName:%s, tbName:%s, tbType:%d", dbFName, tbName, meta->tableType); SSTableMetaVersion metaRent = {.dbId = dbId, .suid = meta->suid, .sversion = meta->sversion, .tversion = meta->tversion}; strcpy(metaRent.dbFName, dbFName); strcpy(metaRent.stbName, tbName); CTG_ERR_RET(ctgMetaRentAdd(&pCtg->stbRent, &metaRent, metaRent.suid, sizeof(SSTableMetaVersion))); return TSDB_CODE_SUCCESS; } int32_t ctgWriteTbIndexToCache(SCatalog *pCtg, SCtgDBCache *dbCache, char *tbName, STableIndex **index) { if (NULL == dbCache->tbCache) { taosMemoryFreeClear(*index); ctgError("db is dropping, dbId:%"PRIx64, dbCache->dbId); CTG_ERR_RET(TSDB_CODE_CTG_DB_DROPPED); } STableIndex* pIndex = *index; SCtgTbCache* pCache = taosHashGet(dbCache->tbCache, tbName, strlen(tbName)); if (NULL == pCache) { SCtgTbCache cache = {0}; cache.pIndex = pIndex; if (taosHashPut(table->tbCache, tbName, strlen(tbName), &cache, sizeof(cache)) != 0) { taosMemoryFreeClear(*index); ctgError("taosHashPut new tbCache failed, tbName:%s", tbName); CTG_ERR_RET(TSDB_CODE_CTG_MEM_ERROR); } *index = NULL; ctgDebug("table %s index updated to cache, ver:%d, num:%d", tbName, pIndex->version, taosArrayGetSize(pIndex->pIndex)); return TSDB_CODE_SUCCESS; } if (pCache->pIndex) { taosArrayDestroyEx(pCache->pIndex->pIndex, tFreeSTableIndexInfo); taosMemoryFreeClear(pCache->pIndex); } pCache->pIndex = pIndex; *index = NULL; ctgDebug("table %s index updated to cache, ver:%d, num:%d", tbName, pIndex->version, taosArrayGetSize(pIndex->pIndex)); return TSDB_CODE_SUCCESS; } int32_t ctgUpdateTbMetaToCache(SCatalog* pCtg, STableMetaOutput* pOut, bool syncReq) { STableMetaOutput* pOutput = NULL; int32_t code = 0; CTG_ERR_RET(ctgCloneMetaOutput(pOut, &pOutput)); CTG_ERR_JRET(ctgUpdateTbMetaEnqueue(pCtg, pOutput, syncReq)); return TSDB_CODE_SUCCESS; _return: ctgFreeSTableMetaOutput(pOutput); CTG_RET(code); } int32_t ctgOpUpdateVgroup(SCtgCacheOperation *operation) { int32_t code = 0; SCtgUpdateVgMsg *msg = operation->data; SDBVgInfo* dbInfo = msg->dbInfo; char* dbFName = msg->dbFName; if (NULL == dbInfo->vgHash) { return TSDB_CODE_SUCCESS; } if (dbInfo->vgVersion < 0 || taosHashGetSize(dbInfo->vgHash) <= 0) { ctgError("invalid db vgInfo, dbFName:%s, vgHash:%p, vgVersion:%d, vgHashSize:%d", dbFName, dbInfo->vgHash, dbInfo->vgVersion, taosHashGetSize(dbInfo->vgHash)); CTG_ERR_RET(TSDB_CODE_CTG_MEM_ERROR); } bool newAdded = false; SDbVgVersion vgVersion = {.dbId = msg->dbId, .vgVersion = dbInfo->vgVersion, .numOfTable = dbInfo->numOfTable}; SCtgDBCache *dbCache = NULL; CTG_ERR_RET(ctgGetAddDBCache(msg->pCtg, dbFName, msg->dbId, &dbCache)); if (NULL == dbCache) { ctgInfo("conflict db update, ignore this update, dbFName:%s, dbId:%"PRIx64, dbFName, msg->dbId); CTG_ERR_RET(TSDB_CODE_CTG_INTERNAL_ERROR); } SCtgVgCache *vgCache = &dbCache->vgCache; CTG_ERR_RET(ctgWLockVgInfo(msg->pCtg, dbCache)); if (vgCache->vgInfo) { SDBVgInfo *vgInfo = vgCache->vgInfo; if (dbInfo->vgVersion < vgInfo->vgVersion) { ctgDebug("db vgVer is old, dbFName:%s, vgVer:%d, curVer:%d", dbFName, dbInfo->vgVersion, vgInfo->vgVersion); ctgWUnlockVgInfo(dbCache); return TSDB_CODE_SUCCESS; } if (dbInfo->vgVersion == vgInfo->vgVersion && dbInfo->numOfTable == vgInfo->numOfTable) { ctgDebug("no new db vgVer or numOfTable, dbFName:%s, vgVer:%d, numOfTable:%d", dbFName, dbInfo->vgVersion, dbInfo->numOfTable); ctgWUnlockVgInfo(dbCache); return TSDB_CODE_SUCCESS; } ctgFreeVgInfo(vgInfo); } vgCache->vgInfo = dbInfo; msg->dbInfo = NULL; ctgDebug("db vgInfo updated, dbFName:%s, vgVer:%d, dbId:%"PRIx64, dbFName, vgVersion.vgVersion, vgVersion.dbId); ctgWUnlockVgInfo(dbCache); dbCache = NULL; strncpy(vgVersion.dbFName, dbFName, sizeof(vgVersion.dbFName)); CTG_ERR_RET(ctgMetaRentUpdate(&msg->pCtg->dbRent, &vgVersion, vgVersion.dbId, sizeof(SDbVgVersion), ctgDbVgVersionSortCompare, ctgDbVgVersionSearchCompare)); _return: ctgFreeVgInfo(msg->dbInfo); taosMemoryFreeClear(msg); CTG_RET(code); } int32_t ctgOpDropDbCache(SCtgCacheOperation *operation) { int32_t code = 0; SCtgDropDBMsg *msg = operation->data; SCatalog* pCtg = msg->pCtg; SCtgDBCache *dbCache = NULL; ctgGetDBCache(msg->pCtg, msg->dbFName, &dbCache); if (NULL == dbCache) { goto _return; } if (dbCache->dbId != msg->dbId) { ctgInfo("dbId already updated, dbFName:%s, dbId:%"PRIx64 ", targetId:%"PRIx64, msg->dbFName, dbCache->dbId, msg->dbId); goto _return; } CTG_ERR_JRET(ctgRemoveDBFromCache(pCtg, dbCache, msg->dbFName)); _return: taosMemoryFreeClear(msg); CTG_RET(code); } int32_t ctgOpDropDbVgroup(SCtgCacheOperation *operation) { int32_t code = 0; SCtgDropDbVgroupMsg *msg = operation->data; SCatalog* pCtg = msg->pCtg; SCtgDBCache *dbCache = NULL; ctgGetDBCache(msg->pCtg, msg->dbFName, &dbCache); if (NULL == dbCache) { goto _return; } CTG_ERR_RET(ctgWLockVgInfo(pCtg, dbCache)); ctgFreeVgInfo(dbCache->vgCache.vgInfo); dbCache->vgCache.vgInfo = NULL; ctgDebug("db vgInfo removed, dbFName:%s", msg->dbFName); ctgWUnlockVgInfo(dbCache); _return: taosMemoryFreeClear(msg); CTG_RET(code); } int32_t ctgOpUpdateTbMeta(SCtgCacheOperation *operation) { int32_t code = 0; SCtgUpdateTbMetaMsg *msg = operation->data; SCatalog* pCtg = msg->pCtg; STableMetaOutput* pMeta = msg->pMeta; SCtgDBCache *dbCache = NULL; if ((!CTG_IS_META_CTABLE(pMeta->metaType)) && NULL == pMeta->tbMeta) { ctgError("no valid tbmeta got from meta rsp, dbFName:%s, tbName:%s", pMeta->dbFName, pMeta->tbName); CTG_ERR_JRET(TSDB_CODE_CTG_INTERNAL_ERROR); } if (CTG_IS_META_BOTH(pMeta->metaType) && TSDB_SUPER_TABLE != pMeta->tbMeta->tableType) { ctgError("table type error, expected:%d, actual:%d", TSDB_SUPER_TABLE, pMeta->tbMeta->tableType); CTG_ERR_JRET(TSDB_CODE_CTG_INTERNAL_ERROR); } CTG_ERR_JRET(ctgGetAddDBCache(pCtg, pMeta->dbFName, pMeta->dbId, &dbCache)); if (NULL == dbCache) { ctgInfo("conflict db update, ignore this update, dbFName:%s, dbId:%" PRIx64, pMeta->dbFName, pMeta->dbId); CTG_ERR_JRET(TSDB_CODE_CTG_INTERNAL_ERROR); } if (CTG_IS_META_TABLE(pMeta->metaType) || CTG_IS_META_BOTH(pMeta->metaType)) { int32_t metaSize = CTG_META_SIZE(pMeta->tbMeta); CTG_ERR_JRET(ctgWriteTbMetaToCache(pCtg, dbCache, pMeta->dbFName, pMeta->dbId, pMeta->tbName, pMeta->tbMeta, metaSize)); pMeta->tbMeta = NULL; } if (CTG_IS_META_CTABLE(pMeta->metaType) || CTG_IS_META_BOTH(pMeta->metaType)) { SCTableMeta* ctbMeta = taosMemoryMalloc(sizeof(SCTableMeta)); if (NULL == ctbMeta) { CTG_ERR_JRET(TSDB_CODE_OUT_OF_MEMORY); } memcpy(ctbMeta, &pMeta->ctbMeta, sizeof(SCTableMeta)); CTG_ERR_JRET(ctgWriteTbMetaToCache(pCtg, dbCache, pMeta->dbFName, pMeta->dbId, pMeta->ctbName, (STableMeta *)ctbMeta, sizeof(SCTableMeta))); } _return: if (pMeta) { taosMemoryFreeClear(pMeta->tbMeta); taosMemoryFreeClear(pMeta); } taosMemoryFreeClear(msg); CTG_RET(code); } int32_t ctgOpDropStbMeta(SCtgCacheOperation *operation) { int32_t code = 0; SCtgDropStbMetaMsg *msg = operation->data; SCatalog* pCtg = msg->pCtg; SCtgDBCache *dbCache = NULL; ctgGetDBCache(pCtg, msg->dbFName, &dbCache); if (NULL == dbCache) { return TSDB_CODE_SUCCESS; } if (msg->dbId && (dbCache->dbId != msg->dbId)) { ctgDebug("dbId already modified, dbFName:%s, current:%"PRIx64", dbId:%"PRIx64", stb:%s, suid:%"PRIx64, msg->dbFName, dbCache->dbId, msg->dbId, msg->stbName, msg->suid); return TSDB_CODE_SUCCESS; } if (taosHashRemove(dbCache->stbCache, &msg->suid, sizeof(msg->suid))) { ctgDebug("stb not exist in stbCache, may be removed, dbFName:%s, stb:%s, suid:%"PRIx64, msg->dbFName, msg->stbName, msg->suid); } else { CTG_CACHE_STAT_DEC(stblNum, 1); } if (taosHashRemove(dbCache->tbCache, msg->stbName, strlen(msg->stbName))) { ctgError("stb not exist in cache, dbFName:%s, stb:%s, suid:%"PRIx64, msg->dbFName, msg->stbName, msg->suid); } else { CTG_CACHE_STAT_DEC(tblNum, 1); } ctgInfo("stb removed from cache, dbFName:%s, stbName:%s, suid:%"PRIx64, msg->dbFName, msg->stbName, msg->suid); CTG_ERR_JRET(ctgMetaRentRemove(&msg->pCtg->stbRent, msg->suid, ctgStbVersionSortCompare, ctgStbVersionSearchCompare)); ctgDebug("stb removed from rent, dbFName:%s, stbName:%s, suid:%"PRIx64, msg->dbFName, msg->stbName, msg->suid); _return: taosMemoryFreeClear(msg); CTG_RET(code); } int32_t ctgOpDropTbMeta(SCtgCacheOperation *operation) { int32_t code = 0; SCtgDropTblMetaMsg *msg = operation->data; SCatalog* pCtg = msg->pCtg; SCtgDBCache *dbCache = NULL; ctgGetDBCache(pCtg, msg->dbFName, &dbCache); if (NULL == dbCache) { return TSDB_CODE_SUCCESS; } if (dbCache->dbId != msg->dbId) { ctgDebug("dbId %" PRIx64 " not match with curId %"PRIx64", dbFName:%s, tbName:%s"msg->dbId, dbCache->dbId, msg->dbFName, msg->tbName); return TSDB_CODE_SUCCESS; } if (taosHashRemove(dbCache->tbCache, msg->tbName, strlen(msg->tbName))) { ctgError("tb %s not exist in cache, dbFName:%s", msg->tbName, msg->dbFName); CTG_ERR_JRET(TSDB_CODE_CTG_INTERNAL_ERROR); } else { CTG_CACHE_STAT_DEC(tblNum, 1); } ctgDebug("table %s removed from cache, dbFName:%s", msg->tbName, msg->dbFName); _return: taosMemoryFreeClear(msg); CTG_RET(code); } int32_t ctgOpUpdateUser(SCtgCacheOperation *operation) { int32_t code = 0; SCtgUpdateUserMsg *msg = operation->data; SCatalog* pCtg = msg->pCtg; if (NULL == pCtg->userCache) { pCtg->userCache = taosHashInit(gCtgMgmt.cfg.maxUserCacheNum, taosGetDefaultHashFunction(TSDB_DATA_TYPE_BINARY), false, HASH_ENTRY_LOCK); if (NULL == pCtg->userCache) { ctgError("taosHashInit %d user cache failed", gCtgMgmt.cfg.maxUserCacheNum); CTG_ERR_JRET(TSDB_CODE_OUT_OF_MEMORY); } } SCtgUserAuth *pUser = (SCtgUserAuth *)taosHashGet(pCtg->userCache, msg->userAuth.user, strlen(msg->userAuth.user)); if (NULL == pUser) { SCtgUserAuth userAuth = {0}; userAuth.version = msg->userAuth.version; userAuth.superUser = msg->userAuth.superAuth; userAuth.createdDbs = msg->userAuth.createdDbs; userAuth.readDbs = msg->userAuth.readDbs; userAuth.writeDbs = msg->userAuth.writeDbs; if (taosHashPut(pCtg->userCache, msg->userAuth.user, strlen(msg->userAuth.user), &userAuth, sizeof(userAuth))) { ctgError("taosHashPut user %s to cache failed", msg->userAuth.user); CTG_ERR_JRET(TSDB_CODE_OUT_OF_MEMORY); } taosMemoryFreeClear(msg); return TSDB_CODE_SUCCESS; } pUser->version = msg->userAuth.version; CTG_LOCK(CTG_WRITE, &pUser->lock); taosHashCleanup(pUser->createdDbs); pUser->createdDbs = msg->userAuth.createdDbs; msg->userAuth.createdDbs = NULL; taosHashCleanup(pUser->readDbs); pUser->readDbs = msg->userAuth.readDbs; msg->userAuth.readDbs = NULL; taosHashCleanup(pUser->writeDbs); pUser->writeDbs = msg->userAuth.writeDbs; msg->userAuth.writeDbs = NULL; CTG_UNLOCK(CTG_WRITE, &pUser->lock); _return: taosHashCleanup(msg->userAuth.createdDbs); taosHashCleanup(msg->userAuth.readDbs); taosHashCleanup(msg->userAuth.writeDbs); taosMemoryFreeClear(msg); CTG_RET(code); } int32_t ctgOpUpdateEpset(SCtgCacheOperation *operation) { int32_t code = 0; SCtgUpdateEpsetMsg *msg = operation->data; SCatalog* pCtg = msg->pCtg; SCtgDBCache *dbCache = NULL; CTG_ERR_JRET(ctgGetDBCache(pCtg, msg->dbFName, &dbCache)); if (NULL == dbCache) { ctgDebug("db %s not exist, ignore epset update", msg->dbFName); goto _return; } CTG_ERR_JRET(ctgWLockVgInfo(pCtg, dbCache)); SDBVgInfo *vgInfo = dbCache->vgCache.vgInfo; if (NULL == vgInfo) { ctgDebug("vgroup in db %s not cached, ignore epset update", msg->dbFName); goto _return; } SVgroupInfo* pInfo = taosHashGet(vgInfo->vgHash, &msg->vgId, sizeof(msg->vgId)); if (NULL == pInfo) { ctgDebug("no vgroup %d in db %s, ignore epset update", msg->vgId, msg->dbFName); goto _return; } SEp* pOrigEp = &pInfo->epSet.eps[pInfo->epSet.inUse]; SEp* pNewEp = &msg->epSet.eps[msg->epSet.inUse]; ctgDebug("vgroup %d epset updated from %d/%d=>%s:%d to %d/%d=>%s:%d, dbFName:%s in ctg", pInfo->vgId, pInfo->epSet.inUse, pInfo->epSet.numOfEps, pOrigEp->fqdn, pOrigEp->port, msg->epSet.inUse, msg->epSet.numOfEps, pNewEp->fqdn, pNewEp->port, msg->dbFName); pInfo->epSet = msg->epSet; _return: if (dbCache) { ctgWUnlockVgInfo(dbCache); } taosMemoryFreeClear(msg); CTG_RET(code); } int32_t ctgOpUpdateTbIndex(SCtgCacheOperation *operation) { int32_t code = 0; SCtgUpdateTbIndexMsg *msg = operation->data; SCatalog* pCtg = msg->pCtg; STableIndex* pIndex = msg->pIndex; SCtgDBCache *dbCache = NULL; CTG_ERR_JRET(ctgGetAddDBCache(pCtg, msg->dbFName, 0, &dbCache)); if (NULL == dbCache) { CTG_ERR_JRET(code); } CTG_ERR_JRET(ctgWriteTbIndexToCache(pCtg, dbCache, msg->tbName, &pIndex)); _return: if (pIndex) { taosArrayDestroyEx(pIndex->pIndex, tFreeSTableIndexInfo); taosMemoryFreeClear(pIndex); } taosMemoryFreeClear(msg); CTG_RET(code); } void ctgUpdateThreadUnexpectedStopped(void) { if (CTG_IS_LOCKED(&gCtgMgmt.lock) > 0) CTG_UNLOCK(CTG_READ, &gCtgMgmt.lock); } void ctgCleanupCacheQueue(void) { SCtgQNode *node = NULL; SCtgQNode *nodeNext = NULL; while (true) { node = gCtgMgmt.queue.head->next; while (node) { if (node->op) { taosMemoryFree(node->op->data); if (node->op->syncOp) { tsem_post(&node->op->rspSem); } else { taosMemoryFree(node->op); } } nodeNext = node->next; taosMemoryFree(node); node = nodeNext; } if (CTG_IS_LOCKED(&gCtgMgmt.lock)) { taosUsleep(1); } else { break; } } taosMemoryFreeClear(gCtgMgmt.queue.head); gCtgMgmt.queue.tail = NULL; } void* ctgUpdateThreadFunc(void* param) { setThreadName("catalog"); #ifdef WINDOWS atexit(ctgUpdateThreadUnexpectedStopped); #endif qInfo("catalog update thread started"); CTG_LOCK(CTG_READ, &gCtgMgmt.lock); while (true) { if (tsem_wait(&gCtgMgmt.queue.reqSem)) { qError("ctg tsem_wait failed, error:%s", tstrerror(TAOS_SYSTEM_ERROR(errno))); } if (atomic_load_8((int8_t*)&gCtgMgmt.exit)) { CTG_UNLOCK(CTG_READ, &gCtgMgmt.lock); ctgCleanupCacheQueue(); break; } SCtgCacheOperation *operation = NULL; ctgDequeue(&operation); SCatalog *pCtg = ((SCtgUpdateMsgHeader *)operation->data)->pCtg; ctgDebug("process [%s] operation", gCtgCacheOperation[operation->opId].name); (*gCtgCacheOperation[operation->opId].func)(operation); if (operation->syncOp) { tsem_post(&operation->rspSem); } CTG_RT_STAT_INC(qDoneNum, 1); ctgdShowClusterCache(pCtg); } if (CTG_IS_LOCKED(&gCtgMgmt.lock)) CTG_UNLOCK(CTG_READ, &gCtgMgmt.lock); qInfo("catalog update thread stopped"); return NULL; } int32_t ctgStartUpdateThread() { TdThreadAttr thAttr; taosThreadAttrInit(&thAttr); taosThreadAttrSetDetachState(&thAttr, PTHREAD_CREATE_JOINABLE); if (taosThreadCreate(&gCtgMgmt.updateThread, &thAttr, ctgUpdateThreadFunc, NULL) != 0) { terrno = TAOS_SYSTEM_ERROR(errno); CTG_ERR_RET(terrno); } taosThreadAttrDestroy(&thAttr); return TSDB_CODE_SUCCESS; }