[Pgpool-general] pgpool + warm standby presentation

Tatsuo Ishii ishii at sraoss.co.jp
Mon Sep 27 12:50:53 UTC 2010


You should use "raw mode", which means both replication_mode and
master_slave_mode are turned off.
--
Tatsuo Ishii
SRA OSS, Inc. Japan
English: http://www.sraoss.co.jp/index_en.php
Japanese: http://www.sraoss.co.jp

> I'm trying to implement pgpool2 3.0 + 8.4.4 with pg_standby.
> Both servers are working correctly (with standby running).
> 
> 
> 
> It's weird for me:
> 
> 2010-09-27 11:49:31 DEBUG: pid 12161: health_check: 0 th DB node status: 2
> 2010-09-27 11:49:31 DEBUG: pid 12161: health_check: 1 th DB node status: 2
> 2010-09-27 11:49:55 DEBUG: pid 12210: I am 12210 accept fd 6
> 2010-09-27 11:49:55 DEBUG: pid 12210: Protocol Major: 1234 Minor: 5679
> database:  user:
> 2010-09-27 11:49:55 DEBUG: pid 12210: SSLRequest from client
> 2010-09-27 11:49:55 DEBUG: pid 12210: Protocol Major: 3 Minor: 0
> database: postgres user: postgres
> 2010-09-27 11:49:55 DEBUG: pid 12210: new_connection: connecting 0 backend
> 2010-09-27 11:49:55 DEBUG: pid 12210: new_connection: connecting 1 backend
> 2010-09-27 11:49:55 ERROR: pid 12210: pool_read_kind: kind does not
> match between master(52) slot[1] (45)
> 
> 
> 
> pcp_node_info:
> 
>  pgpool2/bin/pcp_node_info 5 bbdd2 9898 postgres xxxxx 2
> BackendError
>  pgpool2/bin/pcp_node_info 5 bbdd2 9898 postgres xxxxxx 1
> bbdd2 5432 2 0.000000
> 
> 
> How I can implement warm standby with pgpool? Obviusly, node 2 cannot connect
> because is recovering.
> 
> I must deattach the node and attach it in the failover?
> 
> Regards
> 
> 
> -- 
> ,A  ,A  ,A  ,A  ,A  ,A  ,A  Emanuel Calvo Franco
> ,A  ,A  ,A  ,A  ,A  ,A  ,A  ,A  Independient DBA
> ,A  ,A  ,A  ,A  www.emanuelcalvofranco.com.ar


More information about the Pgpool-general mailing list