#4106

Committed 19 May 2025 07:15AM UTC coverage: 62.857% (-0.2%) from 63.042%

Build # #4106

Build Type

push

travis-ci

Committed by GitHub

Commit Message

Merge pull request #31115 from taosdata/merge/mainto3.0

Run Details

156749 of 318088 branches covered (49.28%)

Branch coverage included in aggregate %.

242535 of 317143 relevant lines covered (76.47%)

18746393.97 hits per line

Source File
Press 'n' to go to next uncovered line, 'b' for previous

65.79

/source/libs/sync/src/syncTimeout.c

/*
 * Copyright (c) 2019 TAOS Data, Inc. <jhtao@taosdata.com>
 *
 * 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 <http://www.gnu.org/licenses/>.
 */

#define _DEFAULT_SOURCE
#include "syncTimeout.h"
#include "syncElection.h"
#include "syncRaftCfg.h"
#include "syncRaftLog.h"
#include "syncReplication.h"
#include "syncRespMgr.h"
#include "syncSnapshot.h"
#include "syncUtil.h"

static void syncNodeCleanConfigIndex(SSyncNode* ths) {
#if 0
  int32_t   newArrIndex = 0;
  SyncIndex newConfigIndexArr[MAX_CONFIG_INDEX_COUNT] = {0};
  SSnapshot snapshot = {0};

  ths->pFsm->FpGetSnapshotInfo(ths->pFsm, &snapshot);
  if (snapshot.lastApplyIndex != SYNC_INDEX_INVALID) {
    for (int32_t i = 0; i < ths->raftCfg.configIndexCount; ++i) {
      if (ths->raftCfg.configIndexArr[i] < snapshot.lastConfigIndex) {
        // pass
      } else {
        // save
        newConfigIndexArr[newArrIndex] = ths->raftCfg.configIndexArr[i];
        ++newArrIndex;
      }
    }

    int32_t oldCnt = ths->raftCfg.configIndexCount;
    ths->raftCfg.configIndexCount = newArrIndex;
    memcpy(ths->raftCfg.configIndexArr, newConfigIndexArr, sizeof(newConfigIndexArr));

    int32_t code = syncWriteCfgFile(ths);
    if (code != 0) {
      sNFatal(ths, "failed to persist cfg");
    } else {
      sNTrace(ths, "clean config index arr, old-cnt:%d, new-cnt:%d", oldCnt, ths->raftCfg.configIndexCount);
    }
  }
#endif
}

static int32_t syncNodeTimerRoutine(SSyncNode* ths) {
  ths->tmrRoutineNum++;

  sDebug("vgId:%d, timer routine, status report", ths->vgId);
  if (ths->tmrRoutineNum % (tsRoutineReportInterval / (ths->pingTimerMS / 1000)) == 0) {
    sNInfo(ths, "timer routines");
  } else {
    sNTrace(ths, "timer routines");
  }

  // timer replicate
  sDebug("vgId:%d, timer routine, node replicate", ths->vgId);
  TAOS_CHECK_RETURN(syncNodeReplicate(ths));

  // clean mnode index
  sDebug("vgId:%d, timer routine, clean config index", ths->vgId);
  if (syncNodeIsMnode(ths)) {
    syncNodeCleanConfigIndex(ths);
  }

  sDebug("vgId:%d, timer routine, snapshot resend", ths->vgId);
  int64_t timeNow = taosGetTimestampMs();

  for (int i = 0; i < ths->peersNum; ++i) {
    SSyncSnapshotSender* pSender = syncNodeGetSnapshotSender(ths, &(ths->peersId[i]));
    if (pSender != NULL) {
      if (ths->isStart && (ths->state == TAOS_SYNC_STATE_LEADER || ths->state == TAOS_SYNC_STATE_ASSIGNED_LEADER) &&
          pSender->start) {
        int64_t elapsedMs = timeNow - pSender->lastSendTime;
        if (elapsedMs < SYNC_SNAP_RESEND_MS) {
          continue;
        }

        if (elapsedMs > SYNC_SNAP_TIMEOUT_MS) {
          sSError(pSender, "snap replication timeout, terminate.");
          snapshotSenderStop(pSender, false);
        } else {
          sSWarn(pSender, "snap replication resend.");
          TAOS_CHECK_RETURN(snapshotReSend(pSender));
        }
      }
    }
  }

  sDebug("vgId:%d, timer routine, resp clean", ths->vgId);
  if (!syncNodeIsMnode(ths)) {
    syncRespClean(ths->pSyncRespMgr);
  }

  return 0;
}

int32_t syncNodeOnTimeout(SSyncNode* ths, const SRpcMsg* pRpc) {
  int32_t      ret = 0;
  SyncTimeout* pMsg = pRpc->pCont;

  syncLogRecvTimer(ths, pMsg, &pRpc->info.traceId);

  if (pMsg->timeoutType == SYNC_TIMEOUT_PING) {
    if (atomic_load_64(&ths->pingTimerLogicClockUser) <= pMsg->logicClock) {
      ++(ths->pingTimerCounter);

      TAOS_CHECK_RETURN(syncNodeTimerRoutine(ths));
    }

  } else if (pMsg->timeoutType == SYNC_TIMEOUT_ELECTION) {
    if (atomic_load_64(&ths->electTimerLogicClock) <= pMsg->logicClock) {
      ++(ths->electTimerCounter);

      TAOS_CHECK_RETURN(syncNodeElect(ths));
    }

  } else if (pMsg->timeoutType == SYNC_TIMEOUT_HEARTBEAT) {
    if (atomic_load_64(&ths->heartbeatTimerLogicClockUser) <= pMsg->logicClock) {
      ++(ths->heartbeatTimerCounter);
      sTrace("vgId:%d, sync timer, type:replicate count:%" PRIu64 ", lc-user:%" PRIu64, ths->vgId,
             ths->heartbeatTimerCounter, ths->heartbeatTimerLogicClockUser);
    }

  } else {
    sError("vgId:%d, recv unknown timer-type:%d", ths->vgId, pMsg->timeoutType);
  }

  return ret;
}

1	/*
2	* Copyright (c) 2019 TAOS Data, Inc. <jhtao@taosdata.com>
3	*
4	* This program is free software: you can use, redistribute, and/or modify
5	* it under the terms of the GNU Affero General Public License, version 3
6	* or later ("AGPL"), as published by the Free Software Foundation.
7	*
8	* This program is distributed in the hope that it will be useful, but WITHOUT
9	* ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
10	* FITNESS FOR A PARTICULAR PURPOSE.
11	*
12	* You should have received a copy of the GNU Affero General Public License
13	* along with this program. If not, see <http://www.gnu.org/licenses/>.
14	*/
15
16	#define _DEFAULT_SOURCE
17	#include "syncTimeout.h"
18	#include "syncElection.h"
19	#include "syncRaftCfg.h"
20	#include "syncRaftLog.h"
21	#include "syncReplication.h"
22	#include "syncRespMgr.h"
23	#include "syncSnapshot.h"
24	#include "syncUtil.h"
25
26	static void syncNodeCleanConfigIndex(SSyncNode* ths) {	24,850✔
27	#if 0
28	int32_t newArrIndex = 0;
29	SyncIndex newConfigIndexArr[MAX_CONFIG_INDEX_COUNT] = {0};
30	SSnapshot snapshot = {0};
31
32	ths->pFsm->FpGetSnapshotInfo(ths->pFsm, &snapshot);
33	if (snapshot.lastApplyIndex != SYNC_INDEX_INVALID) {
34	for (int32_t i = 0; i < ths->raftCfg.configIndexCount; ++i) {
35	if (ths->raftCfg.configIndexArr[i] < snapshot.lastConfigIndex) {
36	// pass
37	} else {
38	// save
39	newConfigIndexArr[newArrIndex] = ths->raftCfg.configIndexArr[i];
40	++newArrIndex;
41	}
42	}
43
44	int32_t oldCnt = ths->raftCfg.configIndexCount;
45	ths->raftCfg.configIndexCount = newArrIndex;
46	memcpy(ths->raftCfg.configIndexArr, newConfigIndexArr, sizeof(newConfigIndexArr));
47
48	int32_t code = syncWriteCfgFile(ths);
49	if (code != 0) {
50	sNFatal(ths, "failed to persist cfg");
51	} else {
52	sNTrace(ths, "clean config index arr, old-cnt:%d, new-cnt:%d", oldCnt, ths->raftCfg.configIndexCount);
53	}
54	}
55	#endif
56	}	24,850✔
57
58	static int32_t syncNodeTimerRoutine(SSyncNode* ths) {	260,035✔
59	ths->tmrRoutineNum++;	260,035✔
60
61	sDebug("vgId:%d, timer routine, status report", ths->vgId);	260,035✔
62	if (ths->tmrRoutineNum % (tsRoutineReportInterval / (ths->pingTimerMS / 1000)) == 0) {	260,158✔
63	sNInfo(ths, "timer routines");	3,146!
64	} else {
65	sNTrace(ths, "timer routines");	257,012✔
66	}
67
68	// timer replicate
69	sDebug("vgId:%d, timer routine, node replicate", ths->vgId);	260,158✔
70	TAOS_CHECK_RETURN(syncNodeReplicate(ths));	260,158✔
71
72	// clean mnode index
73	sDebug("vgId:%d, timer routine, clean config index", ths->vgId);	260,214✔
74	if (syncNodeIsMnode(ths)) {	260,214✔
75	syncNodeCleanConfigIndex(ths);	24,850✔
76	}
77
78	sDebug("vgId:%d, timer routine, snapshot resend", ths->vgId);	259,517✔
79	int64_t timeNow = taosGetTimestampMs();	259,427✔
80
81	for (int i = 0; i < ths->peersNum; ++i) {	290,087✔
82	SSyncSnapshotSender* pSender = syncNodeGetSnapshotSender(ths, &(ths->peersId[i]));	30,656✔
83	if (pSender != NULL) {	30,661!
84	if (ths->isStart && (ths->state == TAOS_SYNC_STATE_LEADER \|\| ths->state == TAOS_SYNC_STATE_ASSIGNED_LEADER) &&	30,662!
85	pSender->start) {	9,472!
86	int64_t elapsedMs = timeNow - pSender->lastSendTime;	×
87	if (elapsedMs < SYNC_SNAP_RESEND_MS) {	×
88	continue;	×
89	}
90
91	if (elapsedMs > SYNC_SNAP_TIMEOUT_MS) {	×
92	sSError(pSender, "snap replication timeout, terminate.");	×
93	snapshotSenderStop(pSender, false);	×
94	} else {
95	sSWarn(pSender, "snap replication resend.");	×
96	TAOS_CHECK_RETURN(snapshotReSend(pSender));	×
97	}
98	}
99	}
100	}
101
102	sDebug("vgId:%d, timer routine, resp clean", ths->vgId);	259,431✔
103	if (!syncNodeIsMnode(ths)) {	259,431✔
104	syncRespClean(ths->pSyncRespMgr);	235,086✔
105	}
106
107	return 0;	259,841✔
108	}
109
110	int32_t syncNodeOnTimeout(SSyncNode* ths, const SRpcMsg* pRpc) {	260,846✔
111	int32_t ret = 0;	260,846✔
112	SyncTimeout* pMsg = pRpc->pCont;	260,846✔
113
114	syncLogRecvTimer(ths, pMsg, &pRpc->info.traceId);	260,846✔
115
116	if (pMsg->timeoutType == SYNC_TIMEOUT_PING) {	261,016✔
117	if (atomic_load_64(&ths->pingTimerLogicClockUser) <= pMsg->logicClock) {	259,468!
118	++(ths->pingTimerCounter);	260,229✔
119
120	TAOS_CHECK_RETURN(syncNodeTimerRoutine(ths));	260,229✔
121	}
122
123	} else if (pMsg->timeoutType == SYNC_TIMEOUT_ELECTION) {	1,548!
124	if (atomic_load_64(&ths->electTimerLogicClock) <= pMsg->logicClock) {	1,548✔
125	++(ths->electTimerCounter);	1,547✔
126
127	TAOS_CHECK_RETURN(syncNodeElect(ths));	1,547!
128	}
129
130	} else if (pMsg->timeoutType == SYNC_TIMEOUT_HEARTBEAT) {	×
131	if (atomic_load_64(&ths->heartbeatTimerLogicClockUser) <= pMsg->logicClock) {	×
132	++(ths->heartbeatTimerCounter);	×
133	sTrace("vgId:%d, sync timer, type:replicate count:%" PRIu64 ", lc-user:%" PRIu64, ths->vgId,	×
134	ths->heartbeatTimerCounter, ths->heartbeatTimerLogicClockUser);
135	}
136
137	} else {
138	sError("vgId:%d, recv unknown timer-type:%d", ths->vgId, pMsg->timeoutType);	×
139	}
140
141	return ret;	261,119✔
142	}

taosdata / TDengine / #4106

Source File Press 'n' to go to next uncovered line, 'b' for previous

Source File
Press 'n' to go to next uncovered line, 'b' for previous