Merge pull request #14109 from taosdata/fix/dnode

feat: drop dnode

source/dnode/mnode/impl/src/mndDnode.c

@@ -508,7 +508,7 @@ static int32_t mndProcessCreateDnodeReq(SRpcMsg *pReq) {
     goto _OVER;
   }
 
-  mDebug("dnode:%s:%d, start to create", createReq.fqdn, createReq.port);
+  mInfo("dnode:%s:%d, start to create", createReq.fqdn, createReq.port);
 
   if (createReq.fqdn[0] == 0 || createReq.port <= 0 || createReq.port > UINT16_MAX) {
     terrno = TSDB_CODE_MND_INVALID_DNODE_EP;
@@ -546,7 +546,7 @@ static int32_t mndDropDnode(SMnode *pMnode, SRpcMsg *pReq, SDnodeObj *pDnode, SM
   pTrans = mndTransCreate(pMnode, TRN_POLICY_RETRY, TRN_CONFLICT_GLOBAL, pReq);
   if (pTrans == NULL) goto _OVER;
   mndTransSetSerial(pTrans);
-  mDebug("trans:%d, used to drop dnode:%d", pTrans->id, pDnode->id);
+  mInfo("trans:%d, used to drop dnode:%d", pTrans->id, pDnode->id);
 
   pRaw = mndDnodeActionEncode(pDnode);
   if (pRaw == NULL || mndTransAppendRedolog(pTrans, pRaw) != 0) goto _OVER;
@@ -559,11 +559,11 @@ static int32_t mndDropDnode(SMnode *pMnode, SRpcMsg *pReq, SDnodeObj *pDnode, SM
   pRaw = NULL;
 
   if (pMObj != NULL) {
-    mDebug("trans:%d, mnode on dnode:%d will be dropped", pTrans->id, pDnode->id);
+    mInfo("trans:%d, mnode on dnode:%d will be dropped", pTrans->id, pDnode->id);
     if (mndSetDropMnodeInfoToTrans(pMnode, pTrans, pMObj) != 0) goto _OVER;
   }
   if (numOfVnodes > 0) {
-    mDebug("trans:%d, %d vnodes on dnode:%d will be dropped", pTrans->id, numOfVnodes, pDnode->id);
+    mInfo("trans:%d, %d vnodes on dnode:%d will be dropped", pTrans->id, numOfVnodes, pDnode->id);
     if (mndSetMoveVgroupsInfoToTrans(pMnode, pTrans, pDnode->id) != 0) goto _OVER;
   }
   if (mndTransPrepare(pMnode, pTrans) != 0) goto _OVER;
@@ -588,7 +588,7 @@ static int32_t mndProcessDropDnodeReq(SRpcMsg *pReq) {
     goto _OVER;
   }
 
-  mDebug("dnode:%d, start to drop", dropReq.dnodeId);
+  mInfo("dnode:%d, start to drop", dropReq.dnodeId);
 
   if (dropReq.dnodeId <= 0) {
     terrno = TSDB_CODE_MND_INVALID_DNODE_ID;

source/dnode/mnode/impl/src/mndProfile.c

@@ -447,7 +447,7 @@ static int32_t mndUpdateAppInfo(SMnode *pMnode, SClientHbReq *pHbReq, SRpcConnIn
       mError("failed to create new app %" PRIx64 " since %s", pReq->appId, terrstr());
       return -1;
     } else {
-      mDebug("a new app %" PRIx64 "created", pReq->appId);
+      mDebug("a new app %" PRIx64 " is created", pReq->appId);
       mndReleaseApp(pMnode, pApp);
       return TSDB_CODE_SUCCESS;
     }

source/dnode/mnode/impl/src/mndSync.c

@@ -75,7 +75,7 @@ void mndSyncCommitMsg(struct SSyncFSM *pFsm, const SRpcMsg *pMsg, SFsmCbMeta cbM
 }
 
 int32_t mndSyncGetSnapshot(struct SSyncFSM *pFsm, SSnapshot *pSnapshot, void *pReaderParam, void **ppReader) {
-  mInfo("start to read snapshot from sdb in atomic way");
+  mDebug("start to read snapshot from sdb in atomic way");
   SMnode *pMnode = pFsm->data;
   return sdbStartRead(pMnode->pSdb, (SSdbIter **)ppReader, &pSnapshot->lastApplyIndex, &pSnapshot->lastApplyTerm,
                       &pSnapshot->lastConfigIndex);
@@ -96,7 +96,7 @@ void mndRestoreFinish(struct SSyncFSM *pFsm) {
     mndTransPullup(pMnode);
     mndSetRestore(pMnode, true);
   } else {
-    mInfo("mnode sync restore finished, and will set ready after first deploy");
+    mInfo("mnode sync restore finished");
   }
 }
 
@@ -118,13 +118,13 @@ void mndReConfig(struct SSyncFSM *pFsm, const SRpcMsg *pMsg, SReConfigCbMeta cbM
 }
 
 int32_t mndSnapshotStartRead(struct SSyncFSM *pFsm, void **ppReader) {
-  mInfo("start to read snapshot from sdb");
+  mDebug("start to read snapshot from sdb");
   SMnode *pMnode = pFsm->data;
   return sdbStartRead(pMnode->pSdb, (SSdbIter **)ppReader, NULL, NULL, NULL);
 }
 
 int32_t mndSnapshotStopRead(struct SSyncFSM *pFsm, void *pReader) {
-  mInfo("stop to read snapshot from sdb");
+  mDebug("stop to read snapshot from sdb");
   SMnode *pMnode = pFsm->data;
   return sdbStopRead(pMnode->pSdb, pReader);
 }

source/dnode/mnode/impl/src/mndTrans.c

@@ -432,11 +432,11 @@ static const char *mndTransStr(ETrnStage stage) {
 }
 
 static void mndTransTestStartFunc(SMnode *pMnode, void *param, int32_t paramLen) {
-  mInfo("test trans start, param:%s, len:%d", (char *)param, paramLen);
+  mDebug("test trans start, param:%s, len:%d", (char *)param, paramLen);
 }
 
 static void mndTransTestStopFunc(SMnode *pMnode, void *param, int32_t paramLen) {
-  mInfo("test trans stop, param:%s, len:%d", (char *)param, paramLen);
+  mDebug("test trans stop, param:%s, len:%d", (char *)param, paramLen);
 }
 
 static TransCbFp mndTransGetCbFp(ETrnFunc ftype) {

source/dnode/mnode/impl/src/mndVgroup.c

@@ -1017,6 +1017,11 @@ int32_t mndSetMoveVgroupInfoToTrans(SMnode *pMnode, STrans *pTrans, SDbObj *pDb,
   if (mndAddDropVnodeAction(pMnode, pTrans, pDb, &newVg, &del, true) != 0) return -1;
   if (mndAddAlterVnodeConfirmAction(pMnode, pTrans, pDb, &newVg) != 0) return -1;
 
+  SSdbRaw *pRaw = mndVgroupActionEncode(&newVg);
+  if (pRaw == NULL || mndTransAppendCommitlog(pTrans, pRaw) != 0) return -1;
+  sdbSetRawStatus(pRaw, SDB_STATUS_READY);
+  pRaw = NULL;
+
   mInfo("vgId:%d, vgroup info after move, replica:%d", newVg.vgId, newVg.replica);
   for (int32_t i = 0; i < newVg.replica; ++i) {
     mInfo("vgId:%d, vnode:%d dnode:%d", newVg.vgId, i, newVg.vnodeGid[i].dnodeId);
@@ -1025,6 +1030,7 @@ int32_t mndSetMoveVgroupInfoToTrans(SMnode *pMnode, STrans *pTrans, SDbObj *pDb,
 }
 
 int32_t mndSetMoveVgroupsInfoToTrans(SMnode *pMnode, STrans *pTrans, int32_t delDnodeId) {
+  int32_t code = 0;
   SArray *pArray = mndBuildDnodesArray(pMnode, delDnodeId);
   if (pArray == NULL) return -1;
 
@@ -1042,18 +1048,24 @@ int32_t mndSetMoveVgroupsInfoToTrans(SMnode *pMnode, STrans *pTrans, int32_t del
       }
     }
 
+    code = 0;
     if (vnIndex != -1) {
       mInfo("vgId:%d, vnode:%d will be removed from dnode:%d", pVgroup->vgId, vnIndex, delDnodeId);
       SDbObj *pDb = mndAcquireDb(pMnode, pVgroup->dbName);
-      mndSetMoveVgroupInfoToTrans(pMnode, pTrans, pDb, pVgroup, vnIndex, pArray);
+      code = mndSetMoveVgroupInfoToTrans(pMnode, pTrans, pDb, pVgroup, vnIndex, pArray);
       mndReleaseDb(pMnode, pDb);
     }
 
     sdbRelease(pMnode->pSdb, pVgroup);
+
+    if (code != 0) {
+      sdbCancelFetch(pMnode->pSdb, pIter);
+      break;
+    }
   }
 
   taosArrayDestroy(pArray);
-  return 0;
+  return code;
 }
 
 static int32_t mndAddIncVgroupReplicaToTrans(SMnode *pMnode, STrans *pTrans, SDbObj *pDb, SVgObj *pVgroup,

source/dnode/mnode/sdb/src/sdbFile.c

@@ -519,7 +519,7 @@ static void sdbCloseIter(SSdbIter *pIter) {
     pIter->name = NULL;
   }
 
-  mInfo("sdbiter:%p, is closed, total:%" PRId64, pIter, pIter->total);
+  mDebug("sdbiter:%p, is closed, total:%" PRId64, pIter, pIter->total);
   taosMemoryFree(pIter);
 }
 
@@ -556,7 +556,7 @@ int32_t sdbStartRead(SSdb *pSdb, SSdbIter **ppIter, int64_t *index, int64_t *ter
   if (term != NULL) *term = commitTerm;
   if (config != NULL) *config = commitConfig;
 
-  mInfo("sdbiter:%p, is created to read snapshot, commit index:%" PRId64 " term:%" PRId64 " config:%" PRId64 " file:%s",
+  mDebug("sdbiter:%p, is created to read snapshot, commit index:%" PRId64 " term:%" PRId64 " config:%" PRId64 " file:%s",
         pIter, commitIndex, commitTerm, commitConfig, pIter->name);
   return 0;
 }
@@ -583,14 +583,14 @@ int32_t sdbDoRead(SSdb *pSdb, SSdbIter *pIter, void **ppBuf, int32_t *len) {
     taosMemoryFree(pBuf);
     return -1;
   } else if (readlen == 0) {
-    mInfo("sdbiter:%p, read snapshot to the end, total:%" PRId64, pIter, pIter->total);
+    mDebug("sdbiter:%p, read snapshot to the end, total:%" PRId64, pIter, pIter->total);
     *ppBuf = NULL;
     *len = 0;
     taosMemoryFree(pBuf);
     return 0;
   } else {  // (readlen <= maxlen)
     pIter->total += readlen;
-    mInfo("sdbiter:%p, read:%d bytes from snapshot, total:%" PRId64, pIter, readlen, pIter->total);
+    mDebug("sdbiter:%p, read:%d bytes from snapshot, total:%" PRId64, pIter, readlen, pIter->total);
     *ppBuf = pBuf;
     *len = readlen;
     return 0;
@@ -609,7 +609,7 @@ int32_t sdbStartWrite(SSdb *pSdb, SSdbIter **ppIter) {
   }
 
   *ppIter = pIter;
-  mInfo("sdbiter:%p, is created to write snapshot, file:%s", pIter, pIter->name);
+  mDebug("sdbiter:%p, is created to write snapshot, file:%s", pIter, pIter->name);
   return 0;
 }
 
@@ -618,7 +618,7 @@ int32_t sdbStopWrite(SSdb *pSdb, SSdbIter *pIter, bool isApply) {
 
   if (!isApply) {
     sdbCloseIter(pIter);
-    mInfo("sdbiter:%p, not apply to sdb", pIter);
+    mDebug("sdbiter:%p, not apply to sdb", pIter);
     return 0;
   }
 
@@ -641,7 +641,7 @@ int32_t sdbStopWrite(SSdb *pSdb, SSdbIter *pIter, bool isApply) {
     return -1;
   }
 
-  mInfo("sdbiter:%p, successfully applyed to sdb", pIter);
+  mDebug("sdbiter:%p, successfully applyed to sdb", pIter);
   return 0;
 }
 
@@ -654,6 +654,6 @@ int32_t sdbDoWrite(SSdb *pSdb, SSdbIter *pIter, void *pBuf, int32_t len) {
   }
 
   pIter->total += writelen;
-  mInfo("sdbiter:%p, write:%d bytes to snapshot, total:%" PRId64, pIter, writelen, pIter->total);
+  mDebug("sdbiter:%p, write:%d bytes to snapshot, total:%" PRId64, pIter, writelen, pIter->total);
   return 0;
 }
\ No newline at end of file

tests/script/jenkins/basic.txt

@@ -23,6 +23,7 @@
 # ---- dnode
 ./test.sh -f tsim/dnode/create_dnode.sim
 ./test.sh -f tsim/dnode/drop_dnode_has_mnode.sim
+./test.sh -f tsim/dnode/drop_dnode_has_vnode_replica1.sim
 
 # ---- insert
 ./test.sh -f tsim/insert/basic0.sim

tests/script/tsim/dnode/drop_dnode_has_mnode.sim

@@ -70,6 +70,7 @@ if $data(2)[2] != follower then
 endi
 
 print =============== step4: drop dnode 2
+sql_error drop dnode 1
 sql drop dnode 2
 
 print show dnodes; 
@@ -83,7 +84,7 @@ if $data00 != 1 then
   return -1
 endi
 
-print show dnodes; 
+print show mnodes; 
 sql show mnodes
 print $data[0][0] $data[0][1] $data[0][2] $data[0][3] $data[0][4] 
 print $data[1][0] $data[1][1] $data[1][2] $data[1][3] $data[1][4] 

tests/script/tsim/dnode/drop_dnode_has_vnode_replica1.sim

+system sh/stop_dnodes.sh
+system sh/deploy.sh -n dnode1 -i 1
+system sh/deploy.sh -n dnode2 -i 2
+system sh/deploy.sh -n dnode3 -i 3
+system sh/cfg.sh -n dnode1 -c supportVnodes -v 0
+system sh/exec.sh -n dnode1 -s start
+system sh/exec.sh -n dnode2 -s start
+sql connect
+
+print =============== step1 create dnode2
+sql create dnode $hostname port 7200
+sql create dnode $hostname port 7300
+
+$x = 0
+step1:
+	$ = $x + 1
+	sleep 1000
+	if $x == 10 then
+	  print ====> dnode not online!
+		return -1
+	endi
+sql show dnodes
+print ===> $data00 $data01 $data02 $data03 $data04 $data05
+print ===> $data10 $data11 $data12 $data13 $data14 $data15
+if $rows != 3 then
+  return -1
+endi
+if $data(1)[4] != ready then
+  goto step1
+endi
+if $data(2)[4] != ready then
+  goto step1
+endi
+if $data(3)[4] != offline then
+  goto step1
+endi
+
+print =============== step2 drop dnode which not exist
+sql_error drop dnode 1
+sql_error drop dnode 4
+
+print =============== step3 create database
+sql create database d1 vgroups 1
+sql use d1
+sql create table d1.st (ts timestamp, i int) tags (j int)
+sql create table d1.c1 using st tags(1)
+sql show d1.tables
+if $rows != 1 then
+  return -1
+endi
+
+sql show d1.vgroups
+print $data[0][0] $data[0][1] $data[0][2] $data[0][3] $data[0][4] 
+if $rows != 1 then
+  return -1
+endi
+if $data(2)[3] != 2 then
+  return -1
+endi
+
+print =============== step4: drop dnode 2
+print cant drop since no enough vnodes
+sql_error drop dnode 2
+
+system sh/exec.sh -n dnode2 -s stop
+system sh/exec.sh -n dnode3 -s start
+$x = 0
+step4:
+	$ = $x + 1
+	sleep 1000
+	if $x == 10 then
+	  print ====> dnode not online!
+		return -1
+	endi
+sql show dnodes
+print ===> $data00 $data01 $data02 $data03 $data04 $data05
+print ===> $data10 $data11 $data12 $data13 $data14 $data15
+if $rows != 3 then
+  return -1
+endi
+if $data(1)[4] != ready then
+  goto step4
+endi
+if $data(2)[4] != offline then
+  goto step4
+endi
+if $data(3)[4] != ready then
+  goto step4
+endi
+
+# cant drop since offline
+sql_error drop dnode 2
+
+print =============== step5: start dnode2
+system sh/exec.sh -n dnode2 -s start
+
+$x = 0
+step5:
+	$ = $x + 1
+	sleep 1000
+	if $x == 10 then
+	  print ====> dnode not online!
+		return -1
+	endi
+sql show dnodes
+print ===> $data00 $data01 $data02 $data03 $data04 $data05
+print ===> $data10 $data11 $data12 $data13 $data14 $data15
+if $rows != 3 then
+  return -1
+endi
+if $data(1)[4] != ready then
+  goto step5
+endi
+if $data(2)[4] != ready then
+  goto step5
+endi
+if $data(3)[4] != ready then
+  goto step5
+endi
+
+print =============== step6: drop dnode2
+sql drop dnode 2
+
+print show dnodes; 
+sql show dnodes;
+print $data[0][0] $data[0][1] $data[0][2] $data[0][3] $data[0][4] 
+print $data[1][0] $data[1][1] $data[1][2] $data[1][3] $data[1][4] 
+if $rows != 2 then 
+  return -1
+endi
+
+print show d1.vgroups
+sql show d1.vgroups
+print $data[0][0] $data[0][1] $data[0][2] $data[0][3] $data[0][4] 
+if $rows != 1 then
+  return -1
+endi
+if $data(2)[3] != 3 then
+  return -1
+endi
+
+print =============== step7: select data
+sql show d1.tables
+if $rows != 1 then
+  return -1
+endi
+
+return
+system sh/exec.sh -n dnode1 -s stop -x SIGINT
+system sh/exec.sh -n dnode2 -s stop -x SIGINT
+system sh/exec.sh -n dnode2 -s stop -x SIGINT