Re: pg_upgrade question

From: Adrian Klaver <adrian(dot)klaver(at)aklaver(dot)com>
To: Alan Stange <stange(at)rentec(dot)com>, pgsql-general(at)lists(dot)postgresql(dot)org
Subject: Re: pg_upgrade question
Date: 2021-12-02 16:10:07
Message-ID: b1780f09-fd51-3adf-b102-737d32404e3d@aklaver.com
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-general

On 12/2/21 08:05, Alan Stange wrote:
> Hello all,
>
> We're running a 13.x installation and looking to upgrade to 14.1.  This
> is all on Linux servers.   We have a main instance running with a number
> of hot standby replicas configured.   In the past, we have done a
> dump/restore on a slow evening and then rsynced all the bits around so
> that the main server and all replicas were identical at the start.   The
> database has gotten larger now (~1TB) so that is a less desirable option.
>
> So I am thinking to do a pg_upgrade --link on each of the main and the
> hot standby replicas, and then restarting all the processes.   I'm sure
> this will work fine on the main server, but I have not seen any mention
> of this working in the expected way on the hot standby replicas.   Can
> someone confirm that the steps I described here will work in the way I
> am optimistically expecting it will?

Have you looked at steps 9 & 11 here?:

https://www.postgresql.org/docs/current/pgupgrade.html

>
> I should mention that we are running on ZFS, and will take a snapshot
> prior to the upgrade, so that if something goes sideways with our use of
> --link  we can revert back instantly.
>
> Thank you,
>
> Alan
>
>

--
Adrian Klaver
adrian(dot)klaver(at)aklaver(dot)com

In response to

Responses

Browse pgsql-general by date

  From Date Subject
Next Message Alan Stange 2021-12-02 16:45:21 Re: pg_upgrade question
Previous Message Alan Stange 2021-12-02 16:05:50 pg_upgrade question