[pgpool-general: 1985] Re: pgpool 3.3.0 RC1 master/master - reset when on db is down

[Yugo Nagata]

Hi,

I'll investigate this. However, I'm afraid the reply would be late for 
I'll take vacation to next Tuesday..

On Thu, 1 Aug 2013 20:02:02 +0200 (CEST)
Jakub Krzemiński <krzeminski at arise.pl> wrote:

> Hello,
> i'm trying set up two pgpool in a master/master ( active/active) configuration, etch pgpool is in replication mode with two postgresql db 9.0.10.
> pgpool1 in replication mode with postgres_A nad postgres_B
> pgpool2 in replication mode with postgres_A nad postgres_B
> ( pgpool1 is on one server with postgres_A and  on other server i have pgpool2 with postgres_B both are virtual server Ubuntu 12.04 64bit LTS) 
> 
> 
> my case is when one node or db or both is down - second one is serving sql  
> 
> now my problem is that: when i kill one of db or pgpool node the second one restarting all sql session 
> ( witch is very bad for me because external services connected to pgpool cant reconnect ) 
> 
> In log file have( i hope it is right part) :
> 
> 
> Aug  1 19:12:20 waw1-pgpool1test pgpool[1901]: detect_error: kind: E
> Aug  1 19:12:20 waw1-pgpool1test pgpool[1901]: detect_postmaster_down_error: receive admin shutdown error from a node.
> Aug  1 19:12:20 waw1-pgpool1test pgpool[1900]: detect_error: kind: E
> Aug  1 19:12:20 waw1-pgpool1test pgpool[1900]: detect_error: kind: E
> Aug  1 19:12:20 waw1-pgpool1test pgpool[1900]: detect_postmaster_down_error: receive admin shutdown error from a node.
> Aug  1 19:12:20 waw1-pgpool1test pgpool[1900]: postmaster on DB node 1 was shutdown by administrative command
> Aug  1 19:12:20 waw1-pgpool1test pgpool[1900]: degenerate_backend_set: 1 fail over request from pid 1900
> Aug  1 19:12:20 waw1-pgpool1test pgpool[1901]: postmaster on DB node 1 was shutdown by administrative command
> Aug  1 19:12:20 waw1-pgpool1test pgpool[1900]: wd_recv_packet: recv failed
> Aug  1 19:12:20 waw1-pgpool1test pgpool[1900]: degenerate_backend_set: failover request from pid 1900 is canceled by other pgpool
> Aug  1 19:12:20 waw1-pgpool1test pgpool[1901]: degenerate_backend_set: 1 fail over request from pid 1901
> Aug  1 19:12:20 waw1-pgpool1test pgpool[1857]: wd_unlock: send unlock request: 1
> Aug  1 19:12:20 waw1-pgpool1test pgpool[1857]: failover_handler called
> Aug  1 19:12:20 waw1-pgpool1test pgpool[1857]: failover_handler: starting to select new master node
> Aug  1 19:12:20 waw1-pgpool1test pgpool[1857]: wd_start_interlock: start interlocking
> Aug  1 19:12:20 waw1-pgpool1test pgpool[1857]: wd_assume_lock_holder: become a new lock holder
> Aug  1 19:12:20 waw1-pgpool1test pgpool[1857]: wd_unlock: send unlock request: 0
> Aug  1 19:12:20 waw1-pgpool1test pgpool[1857]: starting degeneration. shutdown host 192.168.89.53(5433)
> Aug  1 19:12:20 waw1-pgpool1test pgpool[1857]: Restart all children
> Aug  1 19:12:20 waw1-pgpool1test pgpool[1857]: failover_handler: kill 1870
> Aug  1 19:12:20 waw1-pgpool1test pgpool[1870]: child received shutdown request signal 3
> Aug  1 19:12:20 waw1-pgpool1test pgpool[1857]: failover_handler: kill 1871
> Aug  1 19:12:20 waw1-pgpool1test pgpool[1871]: child received shutdown request signal 3
> Aug  1 19:12:20 waw1-pgpool1test pgpool[1857]: failover_handler: kill 1872
> Aug  1 19:12:20 waw1-pgpool1test pgpool[1872]: child received shutdown request signal 3
> Aug  1 19:12:20 waw1-pgpool1test pgpool[1857]: failover_handler: kill 1873
> Aug  1 19:12:20 waw1-pgpool1test pgpool[1873]: child received shutdown request signal 3
> Aug  1 19:12:20 waw1-pgpool1test pgpool[1857]: failover_handler: kill 1874
> Aug  1 19:12:20 waw1-pgpool1test pgpool[1874]: child received shutdown request signal 3
> 
> 
> 
> 
> _______________________________________________
> pgpool-general mailing list
> pgpool-general at pgpool.net
> http://www.pgpool.net/mailman/listinfo/pgpool-general


-- 
Yugo Nagata <nagata at sraoss.co.jp>