PostgreSQL實戰(zhàn)之啟動恢復(fù)讀取checkpoint記錄失敗的條件詳解

1、首先讀取ControlFile->checkPoint指向的checkpoint

2、如果讀取失敗，slave直接abort退出，master再次讀取ControlFile->prevCheckPoint指向的checkpoint

StartupXLOG->
 |--checkPointLoc = ControlFile->checkPoint;
 |--record = ReadCheckpointRecord(xlogreader, checkPointLoc, 1, true):
 |-- if (record != NULL){
   ...
  }else if (StandbyMode){
   ereport(PANIC,(errmsg("could not locate a valid checkpoint record")));
  }else{
   checkPointLoc = ControlFile->prevCheckPoint;
   record = ReadCheckpointRecord(xlogreader, checkPointLoc, 2, true);
   if (record != NULL){
    InRecovery = true;//標(biāo)記下面進入recovery
   }else{
    ereport(PANIC,(errmsg("could not locate a valid checkpoint record")));
   }
  }

一、那么什么條件下讀取的checkpoint記錄record==NULL?

1、ControlFile->checkPoint % XLOG_BLCKSZ SizeOfXLogShortPHD
2、ReadRecord(xlogreader, ControlFile->checkPoint, LOG, true)返回NULL
3、ReadRecord讀到的record!=NULL record->xl_rmid != RM_XLOG_ID
4、ReadRecord讀到的record!=NULL info != XLOG_CHECKPOINT_SHUTDOWN info != XLOG_CHECKPOINT_ONLINE
5、ReadRecord讀到的record!=NULL record->xl_tot_len != SizeOfXLogRecord + SizeOfXLogRecordDataHeaderShort + sizeof(CheckPoint)

二、ReadRecord函數(shù)返回NULL的條件

ReadRecord(xlogreader, ControlFile->checkPoint, LOG, true)
 |--record = XLogReadRecord(xlogreader, ControlFile->checkPoint, errormsg);
 |-- 2.1 record==NULL  ！StandbyMode
 |-- 2.2 record!=NULL  !tliInHistory(xlogreader->latestPageTLI, expectedTLEs)
 /*-----
 note:只要讀取了一頁xlog,就會賦值為該頁第一個記錄的時間線
 XLogReaderValidatePageHeader
  -->xlogreader->latestPageTLI=hdr->xlp_tli;
 ------*/

三、XlogReadRecord讀取checkpoint返回NULL的條件？

XLogReadRecord(xlogreader, ControlFile->checkPoint, errormsg)
    targetPagePtr = ControlFile->checkPoint - (ControlFile->checkPoint % XLOG_BLCKSZ);
    targetRecOff = ControlFile->checkPoint % XLOG_BLCKSZ;
    readOff = ReadPageInternal(state,targetPagePtr, Min(targetRecOff + SizeOfXLogRecord, XLOG_BLCKSZ));
    pageHeaderSize = XLogPageHeaderSize((XLogPageHeader) state->readBuf);
    record = (XLogRecord *) (state->readBuf + RecPtr % XLOG_BLCKSZ);
    total_len = record->xl_tot_len;
    -------------
    1、readOff 0
    2、0 targetRecOff pageHeaderSize
    3、(((XLogPageHeader) state->readBuf)->xlp_info XLP_FIRST_IS_CONTRECORD) targetRecOff == pageHeaderSize
       page頭有跨頁的record并且checkpoint定位的偏移正好在頁頭尾部
    4、targetRecOff = XLOG_BLCKSZ - SizeOfXLogRecord
       !ValidXLogRecordHeader(state, ControlFile->checkPoint, state->ReadRecPtr, record,randAccess)
       ---（record->xl_tot_len SizeOfXLogRecord || record->xl_rmid > RM_MAX_ID || record->xl_prev != state->ReadRecPtr）
    5、targetRecOff > XLOG_BLCKSZ - SizeOfXLogRecord total_len SizeOfXLogRecord
    6、total_len > state->readRecordBufSize !allocate_recordbuf(state, total_len)
       一旦該記錄損壞，total_len的長度非常大的話，就需要allocate_recordbuf擴展state->readbuf，可能因此分配失敗abort
       記錄的checksum需要等待全部讀取完整記錄后才校驗
    -------------

三、ReadPageInternal返回的readOff返回小于0的條件

ReadPageInternal(state,targetPagePtr, Min(targetRecOff + SizeOfXLogRecord, XLOG_BLCKSZ))

    1、第一次read wal文件，readLen = state->read_page：讀取第一頁。readLen 0

    2、readLen>0 !XLogReaderValidatePageHeader(state, targetSegmentPtr, state->readBuf)
    --

    3、讀取checkpoint所在頁readLen = state->read_page: readLen 0

    4、readLen > 0 readLen = SizeOfXLogShortPHD

    5、!XLogReaderValidatePageHeader(state, pageptr, (char *) hdr)

四、XLogPageRead何時返回值0 ?

/*
 1、WaitForWALToBecomeAvailable open失敗
 2、lseek 失敗  ！StandbyMode
 3、read失敗  ！StandbyMode
 4、校驗page頭失敗  ！StandbyMode
 如果是StandbyMode，則會重新retry->WaitForWALToBecomeAvailable,切換日志源進行open
 */
 !WaitForWALToBecomeAvailable(targetPagePtr + reqLen,private->randAccess,1,targetRecPtr)//open
 |-- return -1
 readOff = targetPageOff;
 if (lseek(readFile, (off_t) readOff, SEEK_SET)  0){
  ！StandbyMode:: return -1
 }
 if (read(readFile, readBuf, XLOG_BLCKSZ) != XLOG_BLCKSZ){
  ！StandbyMode:: return -1
 }
 XLogReaderValidatePageHeader(xlogreader, targetPagePtr, readBuf)
 ！StandbyMode:: return -1

五、WaitForWALToBecomeAvailable何時返回false？

--XLOG_FROM_ARCHIVE | XLOG_FROM_PG_WAL
    1、先XLogFileReadAnyTLI open日志：
        1、遍歷時間線列表里的每一個時間線，從最新的開始
        2、當(dāng)讀取checkpoint的時候，source是XLOG_FROM_ANY
        3、先找歸檔的日志進行open；如果open失敗再找WAL日志進行open
        4、如果都沒有open成功，則向前找時間線，open前一個時間線segno和文件號相同的文件進行open
        5、open成功后expectedTLEs被賦值為當(dāng)前時間線列表的所有值
    2、如果open失敗，則切換日志源：XLOG_FROM_ARCHIVE | XLOG_FROM_PG_WAL -> XLOG_FROM_STREAM
    3、切換日志源后，XLOG_FROM_ARCHIVE | XLOG_FROM_PG_WAL 則：
       slave promote ：return false
       !StandbyMode：return false
    --XLOG_FROM_STREAM
    1、!WalRcvStreaming()即receiver進程掛了，切換日志源
    2、CheckForStandbyTrigger()切換日志源
    3、XLOG_FROM_STREAM->XLOG_FROM_ARCHIVE

總結(jié)

以上就是這篇文章的全部內(nèi)容了，希望本文的內(nèi)容對大家的學(xué)習(xí)或者工作具有一定的參考學(xué)習(xí)價值，如果有疑問大家可以留言交流，謝謝大家對腳本之家的支持。