From: Tom Lane <tgl@sss.pgh.pa.us>
To: "Zwettler Markus (OIZ)" <Markus.Zwettler@zuerich.ch>
cc: "pgsql-admin@lists.postgresql.org" <pgsql-admin@lists.postgresql.org>
Subject: Re: how can one infer the maximum possible restore time from a WAL
 file(name) in a backup?
In-reply-to: <173875f2fb734d46877eecf55ae0682e@zuerich.ch>
References: <173875f2fb734d46877eecf55ae0682e@zuerich.ch>
Comments: In-reply-to "Zwettler Markus (OIZ)" <Markus.Zwettler@zuerich.ch>
	message dated "Mon, 21 Nov 2022 12:45:10 +0000"
MIME-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
Content-ID: <3116563.1669041117.1@sss.pgh.pa.us>
Content-Transfer-Encoding: quoted-printable
Date: Mon, 21 Nov 2022 09:31:57 -0500
Message-ID: <3116564.1669041117@sss.pgh.pa.us>
Archived-At: 
 <https://www.postgresql.org/message-id/3116564.1669041117%40sss.pgh.pa.us>
Precedence: bulk

"Zwettler Markus (OIZ)" <Markus.Zwettler@zuerich.ch> writes:
> In Oracle there is a view v$archived_log in which for each logfile the t=
ime of the first and last transaction contained in it is indicated. By mea=
ns of this view one can infer from a log file in the backup the maximum po=
ssible recovery time.

> Is there something similar in Postgres? Or in other words, how can one i=
nfer the maximum possible restore time from a WAL file(name) in a backup?

There's no view, but you could use pg_waldump to examine the timestamps
of the first and last COMMIT/ROLLBACK records in a WAL file.

			regards, tom lane