From: | Ian Barwick <ian(dot)barwick(at)2ndquadrant(dot)com> |
---|---|
To: | ROS Didier <didier(dot)ros(at)edf(dot)fr>, "pgsql-general(at)lists(dot)postgresql(dot)org" <pgsql-general(at)lists(dot)postgresql(dot)org> |
Subject: | Re: repmgr and STONIT - SPLIT BAIN |
Date: | 2019-01-15 11:41:37 |
Message-ID: | d9df1c0d-b069-cf2c-272b-e8802ee4eae9@2ndquadrant.com |
Views: | Raw Message | Whole Thread | Download mbox | Resend email |
Thread: | |
Lists: | pgsql-general |
Hi
On 1/15/19 5:54 PM, ROS Didier wrote:
> Hi
>
> We are going to implement repmgr according to the following configuration :
>
> -Primary node on the site A
>
> -Standby node on the site B
>
> -Witness node on the same site A
>
> Do we need to use LOCATION repmgr parameter in this case to avoid Split Brain and permit STONIT ?
In this case, the "location" parameter should be the same for all nodes.
FYI the "location" parameter is intended for the use case where you have a large
number of nodes in multiple data centres and need to ensure that only a node in the
same data centre (location) as the primary can get promoted.
STONITH is something you'll need to implement yourself.
Regards
Ian Barwick
--
Ian Barwick http://www.2ndQuadrant.com/
PostgreSQL Development, 24x7 Support, Training & Services
From | Date | Subject | |
---|---|---|---|
Next Message | Pavel Stehule | 2019-01-15 11:44:41 | Re: Oracke BLOB to Postgres BYTEA using ora2pg |
Previous Message | Fabio Pardi | 2019-01-15 11:39:50 | Re: Read consistency when using synchronous_commit=off |