Re: Streaming replication failed to start scenarios - Mailing list pgsql-general

From Ben Chobot
Subject Re: Streaming replication failed to start scenarios
Date
Msg-id B017A84A-0438-4E2E-A6FA-5B91B7B70783@silentmedia.com
Whole thread Raw
In response to Re: Streaming replication failed to start scenarios  (chinnaobi <chinnaobi@gmail.com>)
List pgsql-general
On Oct 22, 2012, at 6:57 AM, chinnaobi wrote:

> Hi Laurenz Albe,
>
> I have tested using cygwin rsync in windows 2008 R2, just after restart the
> server.
>
> for 10 GB it took nearly 5 minutes to sync,
> for 50 GB it took nearly 30 minutes,  -- too long Though there were no big
> changes.
>
> My requirement is something less than 5 minutes. I am doing high
> availability. In my setup there is WAL archiving enabled as well.
>
> Is there any way I can speedup the configuring dead primary server to
> standby ??

If it's a planned switch, and not a crash, take a look at this thread:

http://archives.postgresql.org/pgsql-general/2012-08/msg00083.php

If it's a crash, then the only way to bring your crashed node back up as a slave to your new master is with another
fullcopy from that master node. 

pgsql-general by date:

Previous
From: Papiernik Anna-AAP053
Date:
Subject: oracle_fdw with oracle os authentication
Next
From: "Enke, Dr., Michael"
Date:
Subject: Q: ecpg and UPDATE/DELETE ... RETURNING ... for multiple returned rows