日志：
2024-01-22 17:02:23.522: main pid 3956: LOG: reaper handler
2024-01-22 17:02:23.523: main pid 3956: LOG: reaper handler: exiting normally
2024-01-22 17:02:35.423: [unknown] pid 11091: WARNING: write on backend 0 failed with error :"Broken pipe"
2024-01-22 17:02:35.423: [unknown] pid 11091: DETAIL: while trying to write data from offset: 0 wlen: 265642
2024-01-22 17:02:35.423: [unknown] pid 11091: ERROR: unable to write data to backend
2024-01-22 17:02:35.423: [unknown] pid 11091: DETAIL: pool_flush failed
2024-01-22 17:02:35.426: [unknown] pid 11091: WARNING: write on backend 0 failed with error :"Broken pipe"
2024-01-22 17:02:35.426: [unknown] pid 11091: DETAIL: while trying to write data from offset: 0 wlen: 18
2024-01-22 17:02:35.435: [unknown] pid 11091: LOG: received degenerate backend request for node_id: 0 from pid [11091]
2024-01-22 17:02:35.438: watchdog pid 3958: LOG: watchdog received the failover command from local pgpool-II on IPC interface
2024-01-22 17:02:35.438: watchdog pid 3958: LOG: watchdog is processing the failover command [DEGENERATE_BACKEND_REQUEST] received from local pgpool-II on IPC interface
2024-01-22 17:02:35.438: watchdog pid 3958: LOG: we have got the consensus to perform the failover
2024-01-22 17:02:35.438: watchdog pid 3958: DETAIL: 1 node(s) voted in the favor
2024-01-22 17:02:35.438: [unknown] pid 11091: LOG: signal_user1_to_parent_with_reason(0)
2024-01-22 17:02:35.438: [unknown] pid 11091: LOG: unable to flush data to backend
2024-01-22 17:02:35.438: [unknown] pid 11091: DETAIL: do not failover because I am the main process
2024-01-22 17:02:35.438: [unknown] pid 11091: WARNING: write on backend 0 failed with error :"Broken pipe"
2024-01-22 17:02:35.438: [unknown] pid 11091: DETAIL: while trying to write data from offset: 0 wlen: 5
2024-01-22 17:02:35.438: main pid 3956: LOG: Pgpool-II parent process received SIGUSR1
2024-01-22 17:02:35.439: main pid 3956: LOG: Pgpool-II parent process has received failover request
2024-01-22 17:02:35.439: watchdog pid 3958: LOG: received the failover indication from Pgpool-II on IPC interface
2024-01-22 17:02:35.439: watchdog pid 3958: LOG: watchdog is informed of failover start by the main process
问题：
（1）11091显示的是unknown，是不是这个进程本身出现了啥问题，导致pgpool和主库通信异常，所以触发pgpool对pg进行主备切换？11091的问题能查到嘛？
（2）DETAIL: do not failover because I am the main process 根据日志看，主进程不是3956嘛，还是说11091出问题了，导致他认为自己是主进程？
PS:整个过程中后段PG主备均正常。
根据提供的日志信息，可以得出以下推测：
（1）根据日志中的警告和错误信息，进程11091出现了问题，导致与主库的通信异常。具体地说，出现了"write on backend 0 failed with error: 'Broken pipe'"错误。这可能是由于网络连接中断或其他原因导致的通信问题。由于通信异常，pgpool可能触发了对主备切换的操作。要查找进程11091的问题，您可以检查更多的日志文件，特别是涉及进程11091的日志，以获取更多的错误信息和上下文。
（2）关于"DETAIL: do not failover because I am the main process"的日志信息，它表明进程11091认为自己是主进程，并阻止了主备切换。这可能是由于进程11091发生了故障或异常，导致其状态出现问题，使其错误地认为自己是主进程。这可能是由于与主库通信失败或其他原因导致的进程异常。进程11091的问题可能会导致整个切换过程中的异常行为。
需要进一步说明的是，上述推测仅基于您提供的日志信息，并且不能确定问题的确切原因。为了更好地理解问题，建议您进一步检查相关的日志文件、监控系统状态和进行必要的故障排除。如果问题持续存在，您可能需要考虑与技术支持团队或专家进行进一步的沟通和协助。
评论
有用 3
墨值悬赏