That process merely sets up a new server, it doesn't start streaming,
unless the server has been configured correctly. You state that the
slave crashed after two hours. How did you make this determination?
All you seem to be doing is setting up the slave from scratch
repeatedly, and assuming that it will magically just work, rather than
understanding & debugging why its not working. Where is the log
output from your servers that shows what is really transpiring?
On Sat, Mar 9, 2013 at 6:51 AM, akp geek <akpgeek@gmail.com> wrote:
> Hi all -
>
> I am in desperate need of your help. The replication/streaming
> stopped working on March5th. I followed the following procedure to restart
> the streaming. After running it for couple of hours , the database is
> crashing on the slave. This is on our production server. Thanks for your
> help.
>
> Steps that I followed
>
> 1. SELECT pg_start_backup('label', true);
> 2. I created tar of data.
> 3. un tar the data on the slave.
> 4. deleted the pid file on the slave
> 5. created recovery.conf
> 5. SELECT pg_stop_backup(); on the primary
> 6. started the postgres on slave
> 7. After 2 hours, the slave server crashes
>
> Please help.
>
> Regards
>