[Pgpool-general] Two primary servers

[Yan Chunlu]

I remember in a ppt it says not using automatic failover, it is easily bring
up another primary when the network fluctuate.

If pgpool-II detects anything wrong on a node, it executes

failover.sh (or any script you configured)

A non-reachable node is not a down node sometimes

AVOID automatic failover. Prefer sending messages in

your failover.sh script to admins, Nagios or whatever

Anyway, the node won’t be used by pgpool-II... Yes, thats

problematic if it is the MASTER node that is down


I changed failover.sh to just send email notification instead of do the real
job.

2011/8/10 RickDong <rick_segno at hotmail.com>

>  Hi pgpool experts, I am just experiencing with the 3.1alpha and having
> this case where I am not sure if it is the expected behavior.
>
> I set up pgpool and a postgres 9.0.2 on one box and another postgres 9.0.2
> on another box. I have also configured the pgpooladmin to work with them. On
> initial start, pgpool chooses the one with smaller id as primary and the
> other as standby. If I disconnect the primary via the PHP page, failover
> kicks in and makes the second box primary. But in the status, for the old
> primary, it still says "Down Running as primary server". Now try to bring
> the the connection back by clicking the 'Return' button, I get two primary
> servers. Please note that the old primary wasn't actually shutdown
> physically. Is this an expected c ase?
>
> Thanks,
>
> Rick
>
> _______________________________________________
> Pgpool-general mailing list
> Pgpool-general at pgfoundry.org
> http://pgfoundry.org/mailman/listinfo/pgpool-general
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://pgfoundry.org/pipermail/pgpool-general/attachments/20110812/8d3e1fe1/attachment-0001.html>