Re: Use pg_rewind when target timeline was switched

From: Michael Paquier <michael(dot)paquier(at)gmail(dot)com>
To: Alexander Korotkov <a(dot)korotkov(at)postgrespro(dot)ru>
Cc: pgsql-hackers <pgsql-hackers(at)postgresql(dot)org>
Subject: Re: Use pg_rewind when target timeline was switched
Date: 2015-09-18 23:26:57
Message-ID: CAB7nPqTuVnAfi_xi6oSdeqfX77af8VhMz1FXRoEA9MaRw-nTyg@mail.gmail.com
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-hackers

On Fri, Sep 18, 2015 at 12:34 PM, Michael Paquier
<michael(dot)paquier(at)gmail(dot)com> wrote:
> On Fri, Sep 18, 2015 at 9:00 AM, Alexander Korotkov wrote:
>> BTW, it would be an option to generate system_identifier to each new
>> timeline, by analogy of initdb do for the whole WAL.
>> Having such system_identifiers we can distinguish different timeline which
>> have assigned same ids.
>> Any thoughts?
>
> If you mean a new field incorporated in XLogLongPageHeader and
> ControlFile to ensure that a new timeline generated is unique across
> the same installation identified with system_identifier, then I'm not
> against it for something up to 2 bytes (even 1 byte would be fine),

Er, 2 bytes may be a bad idea as well, 1/16k% chance of collision
looks dangerous when rewinding a node... It could cause silent data
corruption on the standby being rewounded.
--
Michael

In response to

Browse pgsql-hackers by date

  From Date Subject
Next Message Michael Paquier 2015-09-18 23:27:44 Re: Use pg_rewind when target timeline was switched
Previous Message Michael Paquier 2015-09-18 23:25:12 Re: Use pg_rewind when target timeline was switched