From: | "Andrus" <kobruleht2(at)hot(dot)ee> |
---|---|
To: | Paul Förster <paul(dot)foerster(at)gmail(dot)com> |
Cc: | "pgsql-general" <pgsql-general(at)postgresql(dot)org> |
Subject: | Re: How to start slave after pg_basebackup. Why min_wal_size and wal_keep_segments are duplicated |
Date: | 2020-05-31 20:56:05 |
Message-ID: | 69FF4F48E6FE4A63A2EBD820C5B2B3A3@dell2 |
Views: | Whole Thread | Raw Message | Download mbox | Resend email |
Thread: | |
Lists: | pgsql-general |
Hi!
> On 31. May, 2020, at 21:47, Andrus <kobruleht2(at)hot(dot)ee> wrote:
> replikaator(at)[unknown] ERROR: requested WAL segment 00000001000002CF000000E9 has already been removed
>the message says it all. You need to copy the WAL file 00000001000002CF000000E9 and newer to the replica's pg_wal directory because
>it has been removed >already on the master site.
>Obviously, you can only do that if the files have been archived.
wal files are not archieved.
>Otherwise, you'd have to fully reinitiate the replica.
I have tried to re-initiate replica serveral times in low-use time but this error occurs again.
>Replication will start again as soon as the requested WALs are copied over to the replica.
>Use a replication slot to avoid this situation. If you use a replication slot, the master will only remove WAL files which are not
>needed by any one replica.
If named replication slot is used commands like
vacuumdb --all --full
will cause main server crash due to disk space limit. pg_wal directory will occupy free disk space. After that main server stops.
I tried using wal_keep_segments =180
Will setting wal_keep_segments to higher value allw replication start after pg_basebackup ?
Andrus.
From | Date | Subject | |
---|---|---|---|
Next Message | Adrian Klaver | 2020-05-31 21:02:08 | Re: pg_dump of database with numerous objects |
Previous Message | Christophe Pettus | 2020-05-31 20:38:11 | Re: pg_dump of database with numerous objects |