[pgpool-general: 2976] Re: PGPool and Watchdog

[Yugo Nagata]

Hi,

On Mon, 23 Jun 2014 16:33:35 +1000
James Sewell <james.sewell at lisasoft.com> wrote:

> Hello All,
> 
> I have a load balancer in front of a pair of pgpool nodes (3.3.1), in front
> of a pair of PostgreSQL   nodes (9.3) which are configured as a streaming
> pair. I am running on RHEL.
> 
> I am using another HA tool to manage the failover between the Postgres
> nodes, and the load is spread between the pgpool nodes with inactive nodes
> being taken out of the load balanced pool.
> 
> Do I need watchdog in this case? I have seen reports that pgpool still
> synchronises useful information using this connection, but can't find what
> this information is?
> 
> If I did use watchdog, which settings would I need in this limited case?

When watchdog enabled, backend status is shared between pgpools. 
For example, when backend goes down ** and when a pgpool detects
this and handles the failover **, the backend status change is sent
to another pgpool.

However, if you use another HA tools to protecte backends, you might
want to use DIALLOW_TO_FAILOVER flag in backend_flag. 
http://www.pgpool.net/docs/latest/pgpool-en.html#BACKEND_FLAG
If you use DIALLOW_TO_FAILOVER, the backend status change wouldn't
be sent to another pgpool, since pgpool doesn't handle backend failover.

And in your case, delegate_IP should be empty since virtual IP sn't necessary.

> 
> Also is there a pcp command which I can use to tell pgpool that a new node
> is master, but not actually promote it (this would have happened already)?
> Does pcp_promote_node do this when no promotion script is provided?

pcp_promote_node works as follow:

(1) This changes pgpool-II's internal status and set the specified
    node to primary. The internal status is used in online-recovery
    and loadbalancing etc..  
** Note that this doesn't control the backend nodes themselves.**

(2) All nodes other than the new primary are detached from pgpool-II.

(3) pgpool-II execute follow_master command for all the detached nodes.


As mentioned at (1), pcp_promote_node doesn't actually promote bacend node. 
However, please note that pcp_promote_node detaches all nodes other than the
new primary. You would have to execute pcp_attach_node to attach the nodes in
follow_master command.

When watchdog enabled, backend status changes by pcp_promote_node and
pcp_attach_node also are sent to another pgpool.

> 
> Thanks for the help.
> 
> Cheers,
> 
> James Sewell,
> Solutions Architect
> ______________________________________
> 
> 
>  Level 2, 50 Queen St, Melbourne VIC 3000
> 
> *P *(+61) 3 8370 8000  *W* www.lisasoft.com  *F *(+61) 3 8370 8099
> 
> -- 
> 
> 
> ------------------------------
> The contents of this email are confidential and may be subject to legal or 
> professional privilege and copyright. No representation is made that this 
> email is free of viruses or other defects. If you have received this 
> communication in error, you may not copy or distribute any part of it or 
> otherwise disclose its contents to anyone. Please advise the sender of your 
> incorrect receipt of this correspondence.


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