[Pgpool-general] How to obtain a synchronized pg_xlog dir?

[Tatsuo Ishii]

> Hi. I'm always dealing with pgpool-II-3.0.1 and postgres-9.0.1 in
> hot-standby/streaming replication...
> 
> I realized that when I try a failback of an ex primary as standby, it fails
> because I have different WALs inside the pg_xlog directory.
> For example in current primary, ex standby, I have:
> pg_xlog/00000008.history
> pg_xlog/00000009.history
> pg_xlog/000000070000000000000025
> pg_xlog/000000070000000000000026
> pg_xlog/000000080000000000000026
> pg_xlog/000000090000000000000026
> pg_xlog/000000090000000000000027
> pg_xlog/000000090000000000000028
> pg_xlog/archive_status/00000008.history.ready
> pg_xlog/archive_status/00000009.history.ready
> 
> In ex primary, now supposed to be the standby I have only:
> pg_xlog/000000090000000000000027
> pg_xlog/000000090000000000000028
> pg_xlog/000000090000000000000029
> pg_xlog/00000009000000000000002A

I would say this is normal. I see same phenomena on my streaming rep
system. I think we make it cleaner which is wrong. Possible causes are:

1) Rsync bug. Rsync does not copy properly.
2) Pgpool-II bug. I can't think of any possibilty right now though.
3) PostgreSQL bug.
4) other than above

For 1) you can try, for example, plain old tar.
For 3) You could try 9.0-stable branch HEAD.
--
Tatsuo Ishii
SRA OSS, Inc. Japan
English: http://www.sraoss.co.jp/index_en.php
Japanese: http://www.sraoss.co.jp

> With this situation tha failback fails and postgresql log reports:
> LOG:  entering standby mode
> LOG:  streaming replication successfully connected to primary
> LOG:  redo starts at 0/27000020
> LOG:  consistent recovery state reached at 0/28000000
> FATAL:  the database system is starting up
> FATAL:  the database system is starting up
> FATAL:  the database system is starting up
> ...
> 
> But if I shut down the servers and perform another manual sync from the current
> primary, ex standby:
> rsync -av /home/database/9.0.1/data/pg_xlog/
> ${currentStandbyIP}:/home/database/9.0.1/data/pg_xlog
> and I restart the servers, the switch between them works immediately...
> 
> But how can automatize this? It seems that during basebackup.sh execution, other
> WALs are producted and so my rsync inside it does not copy everything...
> 
> How can I solve this?
> 
> Thanks in advance!
> 
> _______________________________________________
> Pgpool-general mailing list
> Pgpool-general at pgfoundry.org
> http://pgfoundry.org/mailman/listinfo/pgpool-general