Hi Jeff.
Thanks for the clarification.
I'll adjust wal_keep_segments for the expected biggest table in the backup.
Best regards,
Mads
From: Jeff Janes <jeff.janes@gmail.com>
To: "Mads.Tandrup@schneider-electric.com"
<Mads.Tandrup@schneider-electric.com>,
Cc: Albe Laurenz <laurenz.albe@wien.gv.at>,
"pgsql-general@postgresql.org" <pgsql-general@postgresql.org>
Date: 06-06-2013 18:33
Subject: Re: [GENERAL] Streaming replication with sync slave, but
disconnects due to missing WAL segments
Sent by: pgsql-general-owner@postgresql.org
On Wed, Jun 5, 2013 at 11:26 PM, <Mads.Tandrup@schneider-electric.com>
wrote:
Hi
Thanks for your reply. Do you know of any options that I could give
pg_dump/psql to avoid creating one big transaction? I'm using the
plain text format for pg_dump.
For the plain text format, it is already not one big transaction, unless
you specify to -1 to the psql. However, the load of any individual table
will still be a single transaction, so for a very large table it will still
be a very long transaction.
Using pg_dump for --inserts could get around this, but it would probably be
better to fix the fundamental problem by increasing wal_keep_segments or
something of that nature.
Cheers,
Jeff
______________________________________________________________________
This email has been scanned by the Symantec Email Security.cloud service.
______________________________________________________________________
