Postgres Pro
Downloads
Home   >   mailing lists

Thread: WAL shipping and ever expanding pg_xlog

WAL shipping and ever expanding pg_xlog

From
"Michael Dengler"
Date:
Hi,

I'm doing WAL shipping to do a warm standby system (8.2.4).

The problem is that the pg_xlog dir on the master just gets bigger and bigger (never seems to truncate) and the corresponding archive directory on the slave also gets bigger and bigger. Is there a way to moderate this?

Thanks

Mike

Re: WAL shipping and ever expanding pg_xlog

From
"Michael Dengler"
Date:
OK....it looks like the slave machine has run out of space and that caused the xlog files to pile up on the master.

Still...how do I prevent such all of the shipped WAL segments from remaining on the slave machine? Do I need to retain every single one? Can they be safely removed after the slave machine has restored the particular segment?

Thanks

Mike


On 6/13/07, Michael Dengler <michael.dengler@gmail.com> wrote:
Hi,

I'm doing WAL shipping to do a warm standby system (8.2.4).

The problem is that the pg_xlog dir on the master just gets bigger and bigger (never seems to truncate) and the corresponding archive directory on the slave also gets bigger and bigger. Is there a way to moderate this?

Thanks

Mike


Re: WAL shipping and ever expanding pg_xlog

From
"Merlin Moncure"
Date:
On 6/13/07, Michael Dengler <michael.dengler@gmail.com> wrote:
> OK....it looks like the slave machine has run out of space and that caused
> the xlog files to pile up on the master.
>
> Still...how do I prevent such all of the shipped WAL segments from remaining
> on the slave machine? Do I need to retain every single one? Can they be
> safely removed after the slave machine has restored the particular segment?

Are you using the pg_standy utility?  It has options to control this...

merlin

Re: WAL shipping and ever expanding pg_xlog

From
"Michael Dengler"
Date:
Doug Knight just informed me about the pg_standby module.

Works like a charm!

Thanks

Mike


On 6/13/07, Merlin Moncure < mmoncure@gmail.com> wrote:
On 6/13/07, Michael Dengler < michael.dengler@gmail.com> wrote:
> OK....it looks like the slave machine has run out of space and that caused
> the xlog files to pile up on the master.
>
> Still...how do I prevent such all of the shipped WAL segments from remaining
> on the slave machine? Do I need to retain every single one? Can they be
> safely removed after the slave machine has restored the particular segment?

Are you using the pg_standy utility?  It has options to control this...

merlin