From: | Scott Ribe <scott_ribe(at)elevated-dev(dot)com> |
---|---|
To: | "AL-Temimi, Muthana" <muthana(dot)al-temimi(at)tu-harburg(dot)hamburg(dot)de> |
Cc: | "pgsql-admin(at)postgresql(dot)org" <pgsql-admin(at)postgresql(dot)org> |
Subject: | Re: Latest checkpoint's NextOID |
Date: | 2014-09-25 13:53:00 |
Message-ID: | 8D4BB505-C111-446E-81F9-91EA382E9371@elevated-dev.com |
Views: | Raw Message | Whole Thread | Download mbox | Resend email |
Thread: | |
Lists: | pgsql-admin |
On 25 September 2014 11:12, AL-Temimi, Muthana
<muthana(dot)al-temimi(at)tu-harburg(dot)hamburg(dot)de> wrote:
> i’m using 9.1 PostgreSQL and looking for a function to write WAL to table in
> order to save all the data into database and to make a recovery on the 2nd
> DB Server.
>
>
>
> The idea is to help recovery on other server and get the same OID on the 2nd
> Server and that’s by using the Latest checkpoint's NextOID from the first
> Server.
Why not use WAL-shipping or streaming replication the normal way?
You do realize that everything that goes into tables goes into the WAL, so you write WAL into a table, the table gets logged to WAL, and then that WAL... That's just not going to work, unless you use an unlogged table, in which case you have a replication/recovery mechanism which is inherently unreliable, which is pretty worthless. So the whole idea seems to be unrealistic.
Now, if you want to log changes to tables into a table, that's doable via triggers, usually for audit purposes, but it can be used to keep replicas in synch. It's just that it's hard to imagine what you'd do that Slony-I doesn't already do, assuming you even need its features as opposed to streaming replication.
--
Scott Ribe
scott_ribe(at)elevated-dev(dot)com
http://www.elevated-dev.com/
(303) 722-0567 voice
From | Date | Subject | |
---|---|---|---|
Next Message | Kevin Grittner | 2014-09-25 15:06:34 | Re: Out of shared memory while creating a backup with pg_dump |
Previous Message | Simon Riggs | 2014-09-25 13:36:53 | Re: Latest checkpoint's NextOID |