RE: Patch for migration of the pg_commit_ts directory

From: "Hayato Kuroda (Fujitsu)" <kuroda(dot)hayato(at)fujitsu(dot)com>
To: 'ls7777' <ls7777(at)yandex(dot)ru>
Cc: "pgsql-hackers(at)postgresql(dot)org" <pgsql-hackers(at)postgresql(dot)org>, "orlovmg(at)gmail(dot)com" <orlovmg(at)gmail(dot)com>, "amit(dot)kapila16(at)gmail(dot)com" <amit(dot)kapila16(at)gmail(dot)com>, Masahiko Sawada <sawada(dot)mshk(at)gmail(dot)com>
Subject: RE: Patch for migration of the pg_commit_ts directory
Date: 2026-07-27 05:54:28
Message-ID: OS9PR01MB121496DE9DDA05B06590DA734F5CC2@OS9PR01MB12149.jpnprd01.prod.outlook.com
Views: Whole Thread | Raw Message | Download mbox | Resend email
Thread:
Lists: pgsql-hackers

Dear Sergey,
(BTW, your mail looks broken on the web mail archive [1]. Please check settings)

> Are you talking about checking for files in the pg_commit_ts directory of the old cluster?
> The possibility of copying the catalog is checked here:

No. IIUC, the code check whether the old node has a commit timestamps, and copies them
if it has.

My scenario was that if the NEW node has already had a commit timestamps. In this case,
existing entries would be overwritten after the upgrade. Then my question was that whether
we need to ensure there are no commit timestamp entries in the new cluster.

[1]: https://www.postgresql.org/message-id/7105101784659908@mail.yandex.ru

Best regards,
Hayato Kuroda
FUJITSU LIMITED
 

In response to

Responses

Browse pgsql-hackers by date

  From Date Subject
Next Message Jacob Brazeal 2026-07-27 05:58:21 Write skew observed under serializable isolation
Previous Message Yugo Nagata 2026-07-27 05:36:02 Re: JIT works only partially with meson build?