[pgpool-general: 4893] Re: Database in read only transaction

[Daniel Huhardeaux]

Le 10/08/2016 à 16:37, Lucas Luengas a écrit :
> Hello.
>
> You can check Postgresql log files. When standby node is converted in
> master node, pgpool run shell script indicated in failover_command
> (pgpool.conf). This shell script must create a trigger file in standby
> server. After trigger file is created, standby node is in write-mode,
> and Posgresql log files show this.

Hello,

problem I have is that failover and failback command from pgpool.conf 
are never executed ! :( (i'm in max debug mode on one server)

Question I have: pgpool.conf is replication=on load_balance=on 
use-watchdog=on fail_over_on_backend_error=on  and master_slave=off

 From the documentation I read, especially 
http://www.pgpool.net/pgpool-web/contrib_docs/simple_sr_setting2_3.3/ 
and sample from pgpool-source
it should be possible. I now have some doubt.

Main problem is that when a postgresql DB was done, it reconnect to the 
pool but there is no recovery from master DB to this new standby.

Should I better use master/slave ?

>
> Regards.
>
>
> On Tue, Aug 9, 2016 at 9:39 PM, Daniel Huhardeaux <tech at tootai.net
> <mailto:tech at tootai.net>> wrote:
>
>     Sorry, another problem. After having shutted down pgpool on master
>     (previous post), the standby node became MASTER but unusable. I
>     connect to it and want to delete a table. I receive
>
>     verification_cluster=# drop table utilisateur;
>     WARNING:  packet kind of backend 1 ['E'] does not match with
>     master/majority nodes packet kind ['C']
>     FATAL:  failed to read kind from backend
>     DETAIL:  kind mismatch among backends. Possible last query was:
>     "drop table utilisateur;" kind details are: 0[C] 1[E: cannot execute
>     DROP TABLE in a read-only transaction]
>     HINT:  check data consistency among db nodes
>     server closed the connection unexpectedly
>             This probably means the server terminated abnormally
>             before or while processing the request.
>     The connection to the server was lost. Attempting reset: Succeeded.
>
>     Previous master being down, I tried various solutions including
>     pg_ctl reload or server reboot. It stay in this state.
>
>     >From the test I made, everything is working as it should except the
>     recovery from a standby node when it come back to life. What am I
>     missing ? How can I debug this ?
>
>     Thanks for any hint, sorry to take your time.
>
>     Regards
>
>     Daniel
>
>     --
>     TOOTAi Networks
>     _______________________________________________
>     pgpool-general mailing list
>     pgpool-general at pgpool.net <mailto:pgpool-general at pgpool.net>
>     http://www.pgpool.net/mailman/listinfo/pgpool-general
>     <http://www.pgpool.net/mailman/listinfo/pgpool-general>
>
>


-- 
TOOTAi Networks