提交 ae9e11bb 编写于 作者: S Shengliang Guan

refactor: add sync sem for vnode

上级 1e2fdd57
......@@ -52,10 +52,9 @@ typedef struct {
typedef struct {
int32_t vgId;
int32_t refCount;
int32_t vgVersion;
int32_t refCount;
int8_t dropped;
int8_t accessState;
char *path;
SVnode *pImpl;
STaosQueue *pWriteQ;
......
......@@ -323,7 +323,6 @@ SArray *vmGetMsgHandles() {
if (dmSetMgmtHandle(pArray, TDMT_MON_VM_INFO, vmPutMsgToMonitorQueue, 0) == NULL) goto _OVER;
if (dmSetMgmtHandle(pArray, TDMT_MON_VM_LOAD, vmPutMsgToMonitorQueue, 0) == NULL) goto _OVER;
// Requests handled by VNODE
if (dmSetMgmtHandle(pArray, TDMT_VND_SUBMIT, vmPutMsgToWriteQueue, 0) == NULL) goto _OVER;
if (dmSetMgmtHandle(pArray, TDMT_VND_QUERY, vmPutMsgToQueryQueue, 0) == NULL) goto _OVER;
if (dmSetMgmtHandle(pArray, TDMT_VND_QUERY_CONTINUE, vmPutMsgToQueryQueue, 0) == NULL) goto _OVER;
......
......@@ -49,10 +49,9 @@ int32_t vmOpenVnode(SVnodeMgmt *pMgmt, SWrapperCfg *pCfg, SVnode *pImpl) {
}
pVnode->vgId = pCfg->vgId;
pVnode->refCount = 0;
pVnode->vgVersion = pCfg->vgVersion;
pVnode->refCount = 0;
pVnode->dropped = 0;
pVnode->accessState = TSDB_VN_ALL_ACCCESS;
pVnode->path = tstrdup(pCfg->path);
pVnode->pImpl = pImpl;
......@@ -96,7 +95,7 @@ void vmCloseVnode(SVnodeMgmt *pMgmt, SVnodeObj *pVnode) {
dDebug("vgId:%d, vnode is closed", pVnode->vgId);
if (pVnode->dropped) {
dDebug("vgId:%d, vnode is destroyed for dropped:%d", pVnode->vgId, pVnode->dropped);
dInfo("vgId:%d, vnode is destroyed, dropped:%d", pVnode->vgId, pVnode->dropped);
snprintf(path, TSDB_FILENAME_LEN, "vnode%svnode%d", TD_DIRSEP, pVnode->vgId);
vnodeDestroy(path, pMgmt->pTfs);
}
......
......@@ -53,9 +53,9 @@ static void vmProcessMgmtQueue(SQueueInfo *pInfo, SRpcMsg *pMsg) {
}
if (IsReq(pMsg)) {
if (code != 0 && terrno != 0) {
if (code != 0) {
if (terrno != 0) code = terrno;
dError("msg:%p failed to process since %s", pMsg, terrstr());
code = terrno;
}
vmSendRsp(pMsg, code);
}
......@@ -97,110 +97,6 @@ static void vmProcessFetchQueue(SQueueInfo *pInfo, SRpcMsg *pMsg) {
taosFreeQitem(pMsg);
}
static void vmProcessWriteQueue(SQueueInfo *pInfo, STaosQall *qall, int32_t numOfMsgs) {
int32_t code = 0;
SRpcMsg *pMsg = NULL;
SVnodeObj *pVnode = pInfo->ahandle;
int64_t sync = vnodeGetSyncHandle(pVnode->pImpl);
SArray *pArray = taosArrayInit(numOfMsgs, sizeof(SRpcMsg **));
for (int32_t m = 0; m < numOfMsgs; m++) {
if (taosGetQitem(qall, (void **)&pMsg) == 0) continue;
dTrace("vgId:%d, msg:%p get from vnode-write queue", pVnode->vgId, pMsg);
if (taosArrayPush(pArray, &pMsg) == NULL) {
dError("vgId:%d, failed to push msg:%p to vnode-write array", pVnode->vgId, pMsg);
vmSendRsp(pMsg, TSDB_CODE_OUT_OF_MEMORY);
}
}
for (int32_t m = 0; m < taosArrayGetSize(pArray); m++) {
pMsg = *(SRpcMsg **)taosArrayGet(pArray, m);
code = vnodePreprocessReq(pVnode->pImpl, pMsg);
if (code == TSDB_CODE_ACTION_IN_PROGRESS) {
dTrace("vgId:%d, msg:%p in progress and no rsp", pVnode->vgId, pMsg);
continue;
}
if (pMsg->msgType != TDMT_VND_ALTER_REPLICA) {
code = syncPropose(sync, pMsg, false);
}
if (code == TAOS_SYNC_PROPOSE_SUCCESS) {
dTrace("vgId:%d, msg:%p is proposed and no rsp", pVnode->vgId, pMsg);
continue;
} else if (code == TAOS_SYNC_PROPOSE_NOT_LEADER) {
SEpSet newEpSet = {0};
syncGetEpSet(sync, &newEpSet);
SEp *pEp = &newEpSet.eps[newEpSet.inUse];
if (pEp->port == tsServerPort && strcmp(pEp->fqdn, tsLocalFqdn) == 0) {
newEpSet.inUse = (newEpSet.inUse + 1) % newEpSet.numOfEps;
}
dTrace("vgId:%d, msg:%p is redirect since not leader, numOfEps:%d inUse:%d", pVnode->vgId, pMsg,
newEpSet.numOfEps, newEpSet.inUse);
for (int32_t i = 0; i < newEpSet.numOfEps; ++i) {
dTrace("vgId:%d, msg:%p ep:%s:%u", pVnode->vgId, pMsg, newEpSet.eps[i].fqdn, newEpSet.eps[i].port);
}
SRpcMsg rsp = {.code = TSDB_CODE_RPC_REDIRECT, .info = pMsg->info};
tmsgSendRedirectRsp(&rsp, &newEpSet);
} else {
dError("vgId:%d, msg:%p failed to propose write since %s, code:0x%x", pVnode->vgId, pMsg, tstrerror(code), code);
vmSendRsp(pMsg, code);
}
}
for (int32_t i = 0; i < numOfMsgs; i++) {
pMsg = *(SRpcMsg **)taosArrayGet(pArray, i);
dTrace("vgId:%d, msg:%p is freed", pVnode->vgId, pMsg);
rpcFreeCont(pMsg->pCont);
taosFreeQitem(pMsg);
}
taosArrayDestroy(pArray);
}
static void vmProcessApplyQueue(SQueueInfo *pInfo, STaosQall *qall, int32_t numOfMsgs) {
SVnodeObj *pVnode = pInfo->ahandle;
SRpcMsg *pMsg = NULL;
for (int32_t i = 0; i < numOfMsgs; ++i) {
if (taosGetQitem(qall, (void **)&pMsg) == 0) continue;
dTrace("vgId:%d, msg:%p get from vnode-apply queue", pVnode->vgId, pMsg);
// init response rpc msg
SRpcMsg rsp = {0};
// get original rpc msg
assert(pMsg->msgType == TDMT_SYNC_APPLY_MSG);
SyncApplyMsg *pSyncApplyMsg = syncApplyMsgFromRpcMsg2(pMsg);
syncApplyMsgLog2("==vmProcessApplyQueue==", pSyncApplyMsg);
SRpcMsg originalRpcMsg;
syncApplyMsg2OriginalRpcMsg(pSyncApplyMsg, &originalRpcMsg);
// apply data into tsdb
if (vnodeProcessWriteReq(pVnode->pImpl, &originalRpcMsg, pSyncApplyMsg->fsmMeta.index, &rsp) < 0) {
rsp.code = terrno;
dError("vgId:%d, msg:%p failed to apply since %s", pVnode->vgId, pMsg, terrstr());
}
syncApplyMsgDestroy(pSyncApplyMsg);
rpcFreeCont(originalRpcMsg.pCont);
// if leader, send response
if (pMsg->info.handle != NULL) {
rsp.info = pMsg->info;
tmsgSendRsp(&rsp);
}
dTrace("vgId:%d, msg:%p is freed, code:0x%x", pVnode->vgId, pMsg, rsp.code);
rpcFreeCont(pMsg->pCont);
taosFreeQitem(pMsg);
}
}
static void vmProcessSyncQueue(SQueueInfo *pInfo, STaosQall *qall, int32_t numOfMsgs) {
SVnodeObj *pVnode = pInfo->ahandle;
SRpcMsg *pMsg = NULL;
......@@ -322,7 +218,7 @@ int32_t vmPutRpcMsgToQueue(SVnodeMgmt *pMgmt, EQueueType qtype, SRpcMsg *pRpc) {
if (pMsg == NULL) return -1;
SMsgHead *pHead = pRpc->pCont;
dTrace("vgId:%d, msg:%p is created, type:%s", pHead->vgId, pMsg, TMSG_INFO(pMsg->msgType));
dTrace("vgId:%d, msg:%p is created, type:%s", pHead->vgId, pMsg, TMSG_INFO(pRpc->msgType));
pHead->contLen = htonl(pHead->contLen);
pHead->vgId = htonl(pHead->vgId);
......@@ -362,9 +258,9 @@ int32_t vmGetQueueSize(SVnodeMgmt *pMgmt, int32_t vgId, EQueueType qtype) {
}
int32_t vmAllocQueue(SVnodeMgmt *pMgmt, SVnodeObj *pVnode) {
pVnode->pWriteQ = tWWorkerAllocQueue(&pMgmt->writePool, pVnode, (FItems)vmProcessWriteQueue);
pVnode->pWriteQ = tWWorkerAllocQueue(&pMgmt->writePool, pVnode->pImpl, (FItems)vnodeProposeMsg);
pVnode->pSyncQ = tWWorkerAllocQueue(&pMgmt->syncPool, pVnode, (FItems)vmProcessSyncQueue);
pVnode->pApplyQ = tWWorkerAllocQueue(&pMgmt->writePool, pVnode, (FItems)vmProcessApplyQueue);
pVnode->pApplyQ = tWWorkerAllocQueue(&pMgmt->writePool, pVnode->pImpl, (FItems)vnodeApplyMsg);
pVnode->pQueryQ = tQWorkerAllocQueue(&pMgmt->queryPool, pVnode, (FItem)vmProcessQueryQueue);
pVnode->pFetchQ = tQWorkerAllocQueue(&pMgmt->fetchPool, pVnode, (FItem)vmProcessFetchQueue);
pVnode->pMergeQ = tWWorkerAllocQueue(&pMgmt->mergePool, pVnode, (FItems)vmProcessMergeQueue);
......@@ -381,8 +277,8 @@ int32_t vmAllocQueue(SVnodeMgmt *pMgmt, SVnodeObj *pVnode) {
void vmFreeQueue(SVnodeMgmt *pMgmt, SVnodeObj *pVnode) {
tWWorkerFreeQueue(&pMgmt->writePool, pVnode->pWriteQ);
tWWorkerFreeQueue(&pMgmt->syncPool, pVnode->pSyncQ);
tWWorkerFreeQueue(&pMgmt->writePool, pVnode->pApplyQ);
tWWorkerFreeQueue(&pMgmt->syncPool, pVnode->pSyncQ);
tQWorkerFreeQueue(&pMgmt->queryPool, pVnode->pQueryQ);
tQWorkerFreeQueue(&pMgmt->fetchPool, pVnode->pFetchQ);
tWWorkerFreeQueue(&pMgmt->mergePool, pVnode->pMergeQ);
......
......@@ -71,6 +71,9 @@ int32_t vnodeSnapshotReaderClose(SVSnapshotReader *pReader);
int32_t vnodeSnapshotRead(SVSnapshotReader *pReader, const void **ppData, uint32_t *nData);
int32_t vnodeProcessCreateTSma(SVnode *pVnode, void *pCont, uint32_t contLen);
void vnodeProposeMsg(SQueueInfo *pInfo, STaosQall *qall, int32_t numOfMsgs);
void vnodeApplyMsg(SQueueInfo *pInfo, STaosQall *qall, int32_t numOfMsgs);
// meta
typedef struct SMeta SMeta; // todo: remove
typedef struct SMetaReader SMetaReader;
......
......@@ -84,7 +84,6 @@ int32_t vnodeAsyncCommit(SVnode* pVnode);
int32_t vnodeSyncOpen(SVnode* pVnode, char* path);
void vnodeSyncStart(SVnode* pVnode);
void vnodeSyncClose(SVnode* pVnode);
int32_t vnodeSyncAlter(SVnode* pVnode, SRpcMsg* pMsg);
#ifdef __cplusplus
}
......
......@@ -232,8 +232,10 @@ struct SVnode {
SWal* pWal;
STQ* pTq;
SSink* pSink;
int64_t sync;
tsem_t canCommit;
int64_t sync;
int32_t syncCount;
sem_t syncSem;
SQHandle* pQuery;
};
......
......@@ -81,7 +81,9 @@ SVnode *vnodeOpen(const char *path, STfs *pTfs, SMsgCb msgCb) {
pVnode->state.applied = info.state.committed;
pVnode->pTfs = pTfs;
pVnode->msgCb = msgCb;
pVnode->syncCount = 0;
tsem_init(&pVnode->syncSem, 0, 0);
tsem_init(&(pVnode->canCommit), 0, 1);
// open buffer pool
......@@ -175,6 +177,7 @@ void vnodeClose(SVnode *pVnode) {
vnodeCloseBufPool(pVnode);
// destroy handle
tsem_destroy(&(pVnode->canCommit));
tsem_destroy(&pVnode->syncSem);
taosMemoryFree(pVnode);
}
}
......
......@@ -91,9 +91,6 @@ int32_t vnodePreprocessReq(SVnode *pVnode, SRpcMsg *pMsg) {
}
} break;
case TDMT_VND_ALTER_REPLICA: {
code = vnodeSyncAlter(pVnode, pMsg);
} break;
default:
break;
}
......@@ -107,7 +104,7 @@ int32_t vnodeProcessWriteReq(SVnode *pVnode, SRpcMsg *pMsg, int64_t version, SRp
int32_t len;
int32_t ret;
vTrace("vgId:%d, start to process write request %s, version %" PRId64, TD_VID(pVnode), TMSG_INFO(pMsg->msgType),
vTrace("vgId:%d, start to process write request %s, index:%" PRId64, TD_VID(pVnode), TMSG_INFO(pMsg->msgType),
version);
pVnode->state.applied = version;
......@@ -173,7 +170,7 @@ int32_t vnodeProcessWriteReq(SVnode *pVnode, SRpcMsg *pMsg, int64_t version, SRp
break;
}
vTrace("vgId:%d, process %s request success, version: %" PRId64, TD_VID(pVnode), TMSG_INFO(pMsg->msgType), version);
vTrace("vgId:%d, process %s request success, index:%" PRId64, TD_VID(pVnode), TMSG_INFO(pMsg->msgType), version);
if (tqPushMsg(pVnode->pTq, pMsg->pCont, pMsg->contLen, pMsg->msgType, version) < 0) {
vError("vgId:%d, failed to push msg to TQ since %s", TD_VID(pVnode), tstrerror(terrno));
......
......@@ -16,41 +16,42 @@
#define _DEFAULT_SOURCE
#include "vnd.h"
static int32_t vnodeSyncEqMsg(const SMsgCb *msgcb, SRpcMsg *pMsg);
static int32_t vnodeSyncSendMsg(const SEpSet *pEpSet, SRpcMsg *pMsg);
static SSyncFSM *vnodeSyncMakeFsm(SVnode *pVnode);
static void vnodeSyncCommitMsg(SSyncFSM *pFsm, const SRpcMsg *pMsg, SFsmCbMeta cbMeta);
static void vnodeSyncPreCommitMsg(SSyncFSM *pFsm, const SRpcMsg *pMsg, SFsmCbMeta cbMeta);
static void vnodeSyncRollBackMsg(SSyncFSM *pFsm, const SRpcMsg *pMsg, SFsmCbMeta cbMeta);
static int32_t vnodeSyncGetSnapshot(SSyncFSM *pFsm, SSnapshot *pSnapshot);
static inline bool vnodeIsMsgBlock(tmsg_t type) {
return (type == TDMT_VND_CREATE_TABLE) || (type == TDMT_VND_ALTER_CONFIRM) || (type == TDMT_VND_ALTER_REPLICA);
}
int32_t vnodeSyncOpen(SVnode *pVnode, char *path) {
SSyncInfo syncInfo = {
.vgId = pVnode->config.vgId,
.isStandBy = pVnode->config.standby,
.syncCfg = pVnode->config.syncCfg,
.pWal = pVnode->pWal,
.msgcb = NULL,
.FpSendMsg = vnodeSyncSendMsg,
.FpEqMsg = vnodeSyncEqMsg,
};
static inline bool vnodeIsMsgWeak(tmsg_t type) { return false; }
snprintf(syncInfo.path, sizeof(syncInfo.path), "%s%ssync", path, TD_DIRSEP);
syncInfo.pFsm = vnodeSyncMakeFsm(pVnode);
static inline void vnodeAccumBlockMsg(SVnode *pVnode, tmsg_t type) {
if (!vnodeIsMsgBlock(type)) return;
pVnode->sync = syncOpen(&syncInfo);
if (pVnode->sync <= 0) {
vError("vgId:%d, failed to open sync since %s", pVnode->config.vgId, terrstr());
return -1;
}
int32_t count = atomic_add_fetch_32(&pVnode->syncCount, 1);
vTrace("vgId:%d, accum block, count:%d type:%s", pVnode->config.vgId, count, TMSG_INFO(type));
}
setPingTimerMS(pVnode->sync, 3000);
setElectTimerMS(pVnode->sync, 500);
setHeartbeatTimerMS(pVnode->sync, 100);
return 0;
static inline void vnodeWaitBlockMsg(SVnode *pVnode) {
int32_t count = atomic_load_32(&pVnode->syncCount);
if (count <= 0) return;
vTrace("vgId:%d, wait block finish, count:%d", pVnode->config.vgId, count);
tsem_wait(&pVnode->syncSem);
vTrace("vgId:%d, ===> block finish, count:%d", pVnode->config.vgId, count);
}
static inline void vnodePostBlockMsg(SVnode *pVnode, tmsg_t type) {
if (!vnodeIsMsgBlock(type)) return;
int32_t count = atomic_load_32(&pVnode->syncCount);
if (count <= 0) return;
count = atomic_sub_fetch_32(&pVnode->syncCount, 1);
vTrace("vgId:%d, post block, count:%d type:%s", pVnode->config.vgId, count, TMSG_INFO(type));
if (count <= 0) {
tsem_post(&pVnode->syncSem);
}
}
int32_t vnodeSyncAlter(SVnode *pVnode, SRpcMsg *pMsg) {
static int32_t vnodeProcessSyncReconfigReq(SVnode *pVnode, SRpcMsg *pMsg) {
SAlterVnodeReq req = {0};
if (tDeserializeSAlterVnodeReq((char *)pMsg->pCont + sizeof(SMsgHead), pMsg->contLen - sizeof(SMsgHead), &req) != 0) {
terrno = TSDB_CODE_INVALID_MSG;
......@@ -66,29 +67,107 @@ int32_t vnodeSyncAlter(SVnode *pVnode, SRpcMsg *pMsg) {
vInfo("vgId:%d, replica:%d %s:%u", TD_VID(pVnode), r, pNode->nodeFqdn, pNode->nodePort);
}
int32_t code = syncReconfig(pVnode->sync, &cfg);
if (code == TAOS_SYNC_PROPOSE_SUCCESS) {
// todo refactor
SRpcMsg rsp = {.info = pMsg->info, .code = 0};
tmsgSendRsp(&rsp);
return TSDB_CODE_ACTION_IN_PROGRESS;
return syncReconfig(pVnode->sync, &cfg);
}
void vnodeProposeMsg(SQueueInfo *pInfo, STaosQall *qall, int32_t numOfMsgs) {
SVnode *pVnode = pInfo->ahandle;
int32_t vgId = pVnode->config.vgId;
int32_t code = 0;
SRpcMsg *pMsg = NULL;
for (int32_t m = 0; m < numOfMsgs; m++) {
if (taosGetQitem(qall, (void **)&pMsg) == 0) continue;
vTrace("vgId:%d, msg:%p get from vnode-write queue handle:%p", vgId, pMsg, pMsg->info.handle);
if (pMsg->msgType == TDMT_VND_ALTER_REPLICA) {
code = vnodeProcessSyncReconfigReq(pVnode, pMsg);
} else {
code = vnodePreprocessReq(pVnode, pMsg);
if (code != 0) {
vError("vgId:%d, failed to pre-process msg:%p since %s", vgId, pMsg, terrstr());
} else {
code = syncPropose(pVnode->sync, pMsg, vnodeIsMsgWeak(pMsg->msgType));
}
}
return code;
}
if (code == 0) {
vnodeAccumBlockMsg(pVnode, pMsg->msgType);
} else if (code == TAOS_SYNC_PROPOSE_NOT_LEADER) {
SEpSet newEpSet = {0};
syncGetEpSet(pVnode->sync, &newEpSet);
SEp *pEp = &newEpSet.eps[newEpSet.inUse];
if (pEp->port == tsServerPort && strcmp(pEp->fqdn, tsLocalFqdn) == 0) {
newEpSet.inUse = (newEpSet.inUse + 1) % newEpSet.numOfEps;
}
void vnodeSyncStart(SVnode *pVnode) {
syncSetMsgCb(pVnode->sync, &pVnode->msgCb);
if (pVnode->config.standby) {
syncStartStandBy(pVnode->sync);
vTrace("vgId:%d, msg:%p is redirect since not leader, numOfEps:%d inUse:%d", vgId, pMsg, newEpSet.numOfEps,
newEpSet.inUse);
for (int32_t i = 0; i < newEpSet.numOfEps; ++i) {
vTrace("vgId:%d, msg:%p redirect:%d ep:%s:%u", vgId, pMsg, i, newEpSet.eps[i].fqdn, newEpSet.eps[i].port);
}
SRpcMsg rsp = {.code = TSDB_CODE_RPC_REDIRECT, .info = pMsg->info};
tmsgSendRedirectRsp(&rsp, &newEpSet);
} else {
syncStart(pVnode->sync);
if (terrno != 0) code = terrno;
vError("vgId:%d, msg:%p failed to propose since %s, code:0x%x", vgId, pMsg, tstrerror(code), code);
SRpcMsg rsp = {.code = code, .info = pMsg->info};
tmsgSendRsp(&rsp);
}
vTrace("vgId:%d, msg:%p is freed, code:0x%x", vgId, pMsg, code);
rpcFreeCont(pMsg->pCont);
taosFreeQitem(pMsg);
}
vnodeWaitBlockMsg(pVnode);
}
void vnodeSyncClose(SVnode *pVnode) { syncStop(pVnode->sync); }
void vnodeApplyMsg(SQueueInfo *pInfo, STaosQall *qall, int32_t numOfMsgs) {
SVnode *pVnode = pInfo->ahandle;
int32_t vgId = pVnode->config.vgId;
int32_t code = 0;
SRpcMsg *pMsg = NULL;
int32_t vnodeSyncEqMsg(const SMsgCb *msgcb, SRpcMsg *pMsg) {
for (int32_t i = 0; i < numOfMsgs; ++i) {
if (taosGetQitem(qall, (void **)&pMsg) == 0) continue;
vTrace("vgId:%d, msg:%p get from vnode-apply queue", vgId, pMsg);
// init response rpc msg
SRpcMsg rsp = {0};
// get original rpc msg
assert(pMsg->msgType == TDMT_SYNC_APPLY_MSG);
SyncApplyMsg *pSyncApplyMsg = syncApplyMsgFromRpcMsg2(pMsg);
syncApplyMsgLog2("==vmProcessApplyQueue==", pSyncApplyMsg);
SRpcMsg originalRpcMsg;
syncApplyMsg2OriginalRpcMsg(pSyncApplyMsg, &originalRpcMsg);
// apply data into tsdb
if (vnodeProcessWriteReq(pVnode, &originalRpcMsg, pSyncApplyMsg->fsmMeta.index, &rsp) < 0) {
rsp.code = terrno;
vError("vgId:%d, msg:%p failed to apply since %s", vgId, pMsg, terrstr());
}
syncApplyMsgDestroy(pSyncApplyMsg);
rpcFreeCont(originalRpcMsg.pCont);
vnodePostBlockMsg(pVnode, originalRpcMsg.msgType);
// if leader, send response
if (pMsg->info.handle != NULL) {
rsp.info = pMsg->info;
tmsgSendRsp(&rsp);
}
vTrace("vgId:%d, msg:%p is freed, code:0x%x handle:%p", vgId, pMsg, rsp.code, pMsg->info.handle);
rpcFreeCont(pMsg->pCont);
taosFreeQitem(pMsg);
}
}
static int32_t vnodeSyncEqMsg(const SMsgCb *msgcb, SRpcMsg *pMsg) {
int32_t code = tmsgPutToQueue(msgcb, SYNC_QUEUE, pMsg);
if (code != 0) {
rpcFreeCont(pMsg->pCont);
......@@ -97,7 +176,7 @@ int32_t vnodeSyncEqMsg(const SMsgCb *msgcb, SRpcMsg *pMsg) {
return code;
}
int32_t vnodeSyncSendMsg(const SEpSet *pEpSet, SRpcMsg *pMsg) {
static int32_t vnodeSyncSendMsg(const SEpSet *pEpSet, SRpcMsg *pMsg) {
int32_t code = tmsgSendReq(pEpSet, pMsg);
if (code != 0) {
rpcFreeCont(pMsg->pCont);
......@@ -106,19 +185,19 @@ int32_t vnodeSyncSendMsg(const SEpSet *pEpSet, SRpcMsg *pMsg) {
return code;
}
int32_t vnodeSyncGetSnapshot(SSyncFSM *pFsm, SSnapshot *pSnapshot) {
static int32_t vnodeSyncGetSnapshot(SSyncFSM *pFsm, SSnapshot *pSnapshot) {
vnodeGetSnapshot(pFsm->data, pSnapshot);
return 0;
}
void vnodeSyncReconfig(struct SSyncFSM *pFsm, SSyncCfg newCfg, SReConfigCbMeta cbMeta) {
static void vnodeSyncReconfig(struct SSyncFSM *pFsm, SSyncCfg newCfg, SReConfigCbMeta cbMeta) {
SVnode *pVnode = pFsm->data;
vInfo("vgId:%d, sync reconfig is confirmed", TD_VID(pVnode));
// todo rpc response here
}
void vnodeSyncCommitMsg(SSyncFSM *pFsm, const SRpcMsg *pMsg, SFsmCbMeta cbMeta) {
static void vnodeSyncCommitMsg(SSyncFSM *pFsm, const SRpcMsg *pMsg, SFsmCbMeta cbMeta) {
SyncIndex beginIndex = SYNC_INDEX_INVALID;
if (pFsm->FpGetSnapshot != NULL) {
SSnapshot snapshot = {0};
......@@ -164,7 +243,7 @@ void vnodeSyncCommitMsg(SSyncFSM *pFsm, const SRpcMsg *pMsg, SFsmCbMeta cbMeta)
}
}
void vnodeSyncPreCommitMsg(SSyncFSM *pFsm, const SRpcMsg *pMsg, SFsmCbMeta cbMeta) {
static void vnodeSyncPreCommitMsg(SSyncFSM *pFsm, const SRpcMsg *pMsg, SFsmCbMeta cbMeta) {
char logBuf[256] = {0};
snprintf(logBuf, sizeof(logBuf),
"==callback== ==PreCommitCb== pFsm:%p, index:%ld, isWeak:%d, code:%d, state:%d %s \n", pFsm, cbMeta.index,
......@@ -172,14 +251,14 @@ void vnodeSyncPreCommitMsg(SSyncFSM *pFsm, const SRpcMsg *pMsg, SFsmCbMeta cbMet
syncRpcMsgLog2(logBuf, (SRpcMsg *)pMsg);
}
void vnodeSyncRollBackMsg(SSyncFSM *pFsm, const SRpcMsg *pMsg, SFsmCbMeta cbMeta) {
static void vnodeSyncRollBackMsg(SSyncFSM *pFsm, const SRpcMsg *pMsg, SFsmCbMeta cbMeta) {
char logBuf[256] = {0};
snprintf(logBuf, sizeof(logBuf), "==callback== ==RollBackCb== pFsm:%p, index:%ld, isWeak:%d, code:%d, state:%d %s \n",
pFsm, cbMeta.index, cbMeta.isWeak, cbMeta.code, cbMeta.state, syncUtilState2String(cbMeta.state));
syncRpcMsgLog2(logBuf, (SRpcMsg *)pMsg);
}
SSyncFSM *vnodeSyncMakeFsm(SVnode *pVnode) {
static SSyncFSM *vnodeSyncMakeFsm(SVnode *pVnode) {
SSyncFSM *pFsm = taosMemoryCalloc(1, sizeof(SSyncFSM));
pFsm->data = pVnode;
pFsm->FpCommitCb = vnodeSyncCommitMsg;
......@@ -188,6 +267,42 @@ SSyncFSM *vnodeSyncMakeFsm(SVnode *pVnode) {
pFsm->FpGetSnapshot = vnodeSyncGetSnapshot;
pFsm->FpRestoreFinishCb = NULL;
pFsm->FpReConfigCb = vnodeSyncReconfig;
return pFsm;
}
int32_t vnodeSyncOpen(SVnode *pVnode, char *path) {
SSyncInfo syncInfo = {
.vgId = pVnode->config.vgId,
.isStandBy = pVnode->config.standby,
.syncCfg = pVnode->config.syncCfg,
.pWal = pVnode->pWal,
.msgcb = NULL,
.FpSendMsg = vnodeSyncSendMsg,
.FpEqMsg = vnodeSyncEqMsg,
};
snprintf(syncInfo.path, sizeof(syncInfo.path), "%s%ssync", path, TD_DIRSEP);
syncInfo.pFsm = vnodeSyncMakeFsm(pVnode);
pVnode->sync = syncOpen(&syncInfo);
if (pVnode->sync <= 0) {
vError("vgId:%d, failed to open sync since %s", pVnode->config.vgId, terrstr());
return -1;
}
setPingTimerMS(pVnode->sync, 3000);
setElectTimerMS(pVnode->sync, 500);
setHeartbeatTimerMS(pVnode->sync, 100);
return 0;
}
void vnodeSyncStart(SVnode *pVnode) {
syncSetMsgCb(pVnode->sync, &pVnode->msgCb);
if (pVnode->config.standby) {
syncStartStandBy(pVnode->sync);
} else {
syncStart(pVnode->sync);
}
}
void vnodeSyncClose(SVnode *pVnode) { syncStop(pVnode->sync); }
Markdown is supported
0% .
You are about to add 0 people to the discussion. Proceed with caution.
先完成此消息的编辑!
想要评论请 注册