Postgres, High Availability, patching servers sequentially

From: David Gauthier <davegauthierpg(at)gmail(dot)com>
To: Postgres General <pgsql-general(at)postgresql(dot)org>
Subject: Postgres, High Availability, patching servers sequentially
Date: 2020-08-05 21:36:54
Message-ID: CAMBRECD268mGL_9SbEN39nEFcyZtCcWTPa5Z_5vQBQ9dG487Jg@mail.gmail.com
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-general

Hi:

version 9.6 on linux
Our IT dept configured our DB to be "High Availability" a couple months
back. I believe this means there's a backup server and disks that mirror
the main and can kick in should main go down.

They need to install a patch on the servers which will require server
downtime. Is there a way we can leverage the fact that there are two
servers, do them sequentially, without causing any effective downtime for
the users ? I'm thinking something like...

- Disconnect the backup DB from the main. No longer a HA at this point but
that's OK for the short term.

- Patch that backend server then bring it back up.

- Reconnect it to the main DB as the backup. I'm thinking it should sync
with the main and get the user changes/edits that happened while the thing
was being patched.

- Switch the roles of the main and the backup.

- Disconnect the second server, patch it, bring it back as the backup.
Back to HA.

Browse pgsql-general by date

  From Date Subject
Next Message Mohamed Wael Khobalatte 2020-08-06 01:18:10 Re: Unexplained disk usage in AWS Aurora Postgres
Previous Message Sreejith P 2020-08-05 19:01:54 Check replication lag