| From: | Michael Banck <mbanck(at)gmx(dot)net> |
|---|---|
| To: | Klaus Darilion <klaus(dot)darilion(at)nic(dot)at> |
| Cc: | Mahesh Sathe <maheshsathe1(at)gmail(dot)com>, Raj <rajeshkumar(dot)dba09(at)gmail(dot)com>, Pgsql-admin <pgsql-admin(at)lists(dot)postgresql(dot)org> |
| Subject: | Re: Major upgrade in patroni cluster |
| Date: | 2026-07-06 10:52:10 |
| Message-ID: | 20260706105210.GA6368@p46.dedyn.io;lightning.p46.dedyn.io |
| Views: | Whole Thread | Raw Message | Download mbox | Resend email |
| Thread: | |
| Lists: | pgsql-admin |
On Mon, Jul 06, 2026 at 10:43:52AM +0000, Klaus Darilion wrote:
> Do I understand it right, that during the rolling upgrade you loose
> High Availability as for short time the Patroni nodes use a different
> Postgres version and hence the streaming replication will not work?
That is correct, and short of using logical replication for
zero-downtime upgrade into a separate, new, Patroni cluster, I don't see
how to circumvent this.
One can make the upgrade window smaller by upgrading the standbys with
rsync, but that is quite brittle and should only be done in controlled
environments after lots of testing. Also, it only works with the --link
pg_upgrade option, not --copy or --clone.
Michael
| From | Date | Subject | |
|---|---|---|---|
| Next Message | Michael Banck | 2026-07-06 10:56:19 | Re: Major upgrade in patroni cluster |
| Previous Message | Klaus Darilion | 2026-07-06 10:43:52 | AW: Major upgrade in patroni cluster |