Re: Streaming-only cascading replica won't come up without writes on the master

From: Heikki Linnakangas <hlinnakangas(at)vmware(dot)com>
To: Josh Berkus <josh(at)agliodbs(dot)com>
Cc: PostgreSQL-development <pgsql-hackers(at)postgreSQL(dot)org>
Subject: Re: Streaming-only cascading replica won't come up without writes on the master
Date: 2013-02-01 17:04:45
Message-ID: 510BF5AD.4000903@vmware.com
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-hackers

On 01.02.2013 02:04, Josh Berkus wrote:
> I thought this was only a 9.3 issue, but it turns out to be
> reproduceable on 9.2.2. Basically, I did:
>
> 1. master is queicent ... no writes occuring.
> 2. createded cascading replica (reprep1) from replica (repmaster)
> 3. reprep1 remains "in recovery mode" until a write occurs on master
>
> I've been able to reproduce this several times on my laptop using
> postmasters running on different ports.

Hmm, that sounds like the issue discussed here:
http://www.postgresql.org/message-id/50C7612E.5060008@vmware.com, and
fixed by this commit:

http://git.postgresql.org/gitweb/?p=postgresql.git;a=commit;h=fb565f8c9616ec8ab1b5176d16f310725e581e6e

Can you check if you still see that behavior with REL9_2_STABLE?

- Heikki

In response to

Responses

Browse pgsql-hackers by date

  From Date Subject
Next Message Peter Eisentraut 2013-02-01 17:11:13 Re: PL/PgSQL STRICT
Previous Message Andrew Dunstan 2013-02-01 16:35:39 Re: obsolete code