Re: Streaming replication breakdown

On Mon, Jun 6, 2022 at 1:15 PM Arun Maddukuri <arun(dot)maddukuri07(at)gmail(dot)com>
wrote:

> Hi Postgresql team,
>
> My streaming replication has broken down. A request to provide a
> solution to fix it.
>
> Please find below details:
>
> This is Ubuntu (18.04), Postgresql 13.7 version.
>
> Log details:
>
> 2022-06-05 09:19:00.006 UTC,,,14121,,629c7503.3729,1,,2022-06-05 09:18:59
> UTC,,0,LOG,00000,"started streaming WAL from primary at 0/A6000000 on
> timeline 1",,,,,,,,,"","walreceiver"
> 2022-06-05 09:19:00.006 UTC,,,14121,,629c7503.3729,2,,2022-06-05 09:18:59
> UTC,,0,FATAL,XX000,"could not receive data from WAL stream: ERROR:
> requested starting point 0/A6000000 is ahead of the WAL flush position of
> this server 0/A50000D8",,,,,,,,,"","walreceiver"
>

It looks like the master server got removed and replaced with an earlier
version of itself, such as from a backup. Talk to however might have been
in a position to make that happen.

You will probably need to recreate the replica to get it streaming again.
But first save a copy of the current replica someplace safe, as it may have
the only copy which exists of some of the data lost from the master durings
it mysterious adventure.

Cheers,

Jeff