Results 1 to 3 of 3
  1. #1
    Join Date
    Nov 2012
    Posts
    3

    Streaming Replication error - WAL contains references to invalid pages

    Hi guys,
    First post here so hi!
    Wondering if anyone has any idea what's going on here.
    Cheers
    Gaz

    PostgreSQL 9.2 on Windows Server 2008 R2 64-bit

    Streaming Replication initialised & working on a 500GB database.
    Following day on slave, following errors in the log:

    2012-11-27 09:57:53 GMT WARNING: page 911726 of relation pg_tblspc/16570/PG_9.2_201204301/16571/16595 is uninitialized
    2012-11-27 09:57:53 GMT CONTEXT: xlog redo vacuum: rel 16570/16571/16595; blk 911727, lastBlockVacuumed 911725
    2012-11-27 09:57:53 GMT PANIC: WAL contains references to invalid pages
    2012-11-27 09:57:53 GMT CONTEXT: xlog redo vacuum: rel 16570/16571/16595; blk 911727, lastBlockVacuumed 911725
    2012-11-27 09:57:54 GMT LOG: startup process (PID 1392) exited with exit code 3
    2012-11-27 09:57:54 GMT LOG: terminating any other active server processes

    Restarting the PostgreSQL service on the slave, the log shows the same errors:

    2012-11-27 10:26:39 GMT LOG: database system was interrupted while in recovery at log time 2012-11-27 09:37:29 GMT
    2012-11-27 10:26:39 GMT HINT: If this has occurred more than once some data might be corrupted and you might need to choose an earlier recovery target.
    2012-11-27 10:26:39 GMT LOG: entering standby mode
    2012-11-27 10:26:39 GMT LOG: redo starts at 3A8/11A5A9A8
    2012-11-27 10:26:40 GMT FATAL: the database system is starting up
    2012-11-27 10:26:41 GMT FATAL: the database system is starting up
    2012-11-27 10:26:42 GMT FATAL: the database system is starting up
    2012-11-27 10:26:43 GMT FATAL: the database system is starting up
    2012-11-27 10:26:44 GMT FATAL: the database system is starting up
    2012-11-27 10:26:45 GMT FATAL: the database system is starting up
    2012-11-27 10:26:46 GMT FATAL: the database system is starting up
    2012-11-27 10:26:47 GMT FATAL: the database system is starting up
    2012-11-27 10:26:48 GMT FATAL: the database system is starting up
    2012-11-27 10:26:49 GMT FATAL: the database system is starting up
    2012-11-27 10:26:50 GMT LOG: consistent recovery state reached at 3A8/387BFF48
    2012-11-27 10:26:50 GMT LOG: database system is ready to accept read only connections
    2012-11-27 10:26:50 GMT WARNING: page 911726 of relation pg_tblspc/16570/PG_9.2_201204301/16571/16595 is uninitialized
    2012-11-27 10:26:50 GMT CONTEXT: xlog redo vacuum: rel 16570/16571/16595; blk 911727, lastBlockVacuumed 911725
    2012-11-27 10:26:50 GMT PANIC: WAL contains references to invalid pages
    2012-11-27 10:26:50 GMT CONTEXT: xlog redo vacuum: rel 16570/16571/16595; blk 911727, lastBlockVacuumed 911725
    2012-11-27 10:26:50 GMT LOG: startup process (PID 2640) exited with exit code 3
    2012-11-27 10:26:50 GMT LOG: terminating any other active server processes

    I noticed a autovacuum process started on the master at 09:04am and was still running when the slave tripped over could this be a factor?
    Or could this be a consequence of the WAL corruption issue fixed in 9.2.1/9.1.6?

    Full page writes is switched on.

  2. #2
    Join Date
    Nov 2012
    Posts
    3
    Anyone have any ideas on this, or pointers where to look? Really tearing my hair out on this
    Thanks

  3. #3
    Join Date
    Nov 2012
    Posts
    3
    If anyone else encounters this - it was an error with the copy mechanism moving data from the master to the slave. Used a better utility (robocopy) and all's been well since.
    Cheers

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •