On Thu, Jul 24, 2014 at 12:53 PM, Fabio Milano <fmilano@crdsgroup.com> wrote:
> Any assistance in interpreting the logs is much appreciated.
> Replication server crashes. Below is snippet from log.
> 2014-07-22 23:36:23 EDT LOG: started streaming WAL from pr
> imary at 12/B0000000 on timeline 1
> 2014-07-22 23:43:12 EDT FATAL: could not receive data from
> WAL stream: server closed the connection unexpectedly
> This probably means the server terminated a
> bnormally
> before or while processing the request.
The replication stream has finished. Did you root server crash in some way?
> cp: cannot stat '/var/lib/postgresql/archive/00000001000000
> 12000000D0': No such file or directory
> 2014-07-22 23:43:12 EDT LOG: unexpected pageaddr 12/8D0000
> 00 in log segment 0000000100000012000000D0, offset 0
This means that server expected new WAL, but found instead an old
record, something that happens when a WAL file is recycled and is not
yet overwritten with new data. Did your archives get corrupted in some
way by a crash of master?
> 2014-07-22 23:43:28 EDT LOG: restored log file "0000000100
> 000012000000D0" from archive
> cp: cannot stat '/var/lib/postgresql/archive/00000001000000
> 12000000D1': No such file or directory
> 2014-07-22 23:43:28 EDT LOG: unexpected pageaddr 12/A50000
> 00 in log segment 0000000100000012000000D1, offset 0
Same here.
> Upon trying a /etc/init.d/postgresql restart
> 2014-07-23 14:47:55 EDT LOG: restored log file "0000000 00001B0000001D"
> from archive
> 2014-07-23 14:47:55 EDT LOG: consistent recovery state ched at
> 1B/1DFC64C0
> 2014-07-23 14:47:55 EDT LOG: database system is ready t ccept read only
> connections
> 2014-07-23 14:47:55 EDT LOG: restored log file "0000000 00001B0000001E"
> from archive
> 2014-07-23 14:47:56 EDT LOG: startup process (PID 730) terminated by
> signal 7: Bus error
This may find its root to a hardware problem...
What is the version of 9.3 you are using btw?
--
Michael