Horton

Re: A recent message added to pg_upgade

Álvaro Herrera <alvherre@kurilemu.de>

View thread View on postgr.es
From: Álvaro Herrera <alvherre@kurilemu.de>
To: Dilip Kumar <dilipbalaut@gmail.com>
Cc: Amit Kapila <amit.kapila16@gmail.com>, Tom Lane <tgl@sss.pgh.pa.us>, Kyotaro Horiguchi <horikyota.ntt@gmail.com>, smithpb2250@gmail.com, bharath.rupireddyforpostgres@gmail.com, pgsql-hackers@lists.postgresql.org
Date: 2025-07-09T11:59:28Z
Lists: pgsql-hackers

Attachments

On 2025-Jul-09, Dilip Kumar wrote:

> On Wed, Jul 9, 2025 at 9:07 AM Dilip Kumar <dilipbalaut@gmail.com> wrote:

> > After further consideration, I believe your proposed method is
> > superior to forcing the max_slot_wal_keep_size to -1 via a check hook.
> > The ultimate goal is to prevent WAL removal during a binary upgrade,
> > and your approach directly addresses this issue rather than
> > controlling it by forcing the GUC value.  I am planning to send a
> > patch using this approach for both max_slot_wal_keep_size as well as
> > for idle_replication_slot_timeout.
> 
> PFA, with this approach.

This indeed makes the whole thing a lot simpler and more direct than the
original code, and solves this subthread's complaint.  It's a bit weird
that slot.c and xlog.c now have to worry about IsBinaryUpgrade, but I
don't feel any guilt about that.

I propose a few comment updates on top of your patch.

-- 
Álvaro Herrera               48°01'N 7°57'E  —  https://www.EnterpriseDB.com/
"Oh, great altar of passive entertainment, bestow upon me thy discordant images
at such speed as to render linear thought impossible" (Calvin a la TV)

Commits

  1. Fix the handling of two GUCs during upgrade.

  2. Rework how logirep launchers are stopped during pg_upgrade

  3. Prohibit max_slot_wal_keep_size to value other than -1 during upgrade.

  4. Prevent startup of logical replication launcher during pg_upgrade

  5. Migrate logical slots to the new node during an upgrade.