pg.ddx.io  pgsql-admin@postgresql.org mailing list archive  
help / color / mirror / Atom feed
Re: Change "shutdown time" was: Change "database system
3+ messages / 2 participants
[nested] [flat]

* Re: Change "shutdown time" was: Change "database system
@ 2006-12-15 15:23  Joshua Colson <joshua.colson@ination.com>
  0 siblings, 1 reply; 3+ messages in thread

From: Joshua Colson @ 2006-12-15 15:23 UTC (permalink / raw)
  To: pgsql-admin

On Thu, 2006-12-14 at 19:10 -0500, Tom Lane wrote: 
> You'd have to modify pg_resetxlog to force that value into the sysid
> rather than whatever is there.  Hand-editing wouldn't work very well
> because of (a) the CRC on pg_control and (b) the need to copy the
> value into extant WAL segments.

Well, I ended up just using the pg_control file from the running
installation. The seemed to work in combination with pg_resetxlog but
now the restore is not working because the postmaster has a shutdown
timestamp of today's date and the recovery is thinking that it has
already recovered past the proper point in time. :(

Does anyone know how I can convince the postmaster to believe it was
shutdown at approximately the time the pg_dumpall happened?

Thank you all again for any assistance!

-- 
Joshua Colson <joshua.colson@ination.com>
iNation, LLC




^ permalink  raw  reply  [nested|flat] 3+ messages in thread

* Re: Change "shutdown time" was: Change "database system
@ 2006-12-15 17:57  Tom Lane <tgl@sss.pgh.pa.us>
  parent: Joshua Colson <joshua.colson@ination.com>
  0 siblings, 1 reply; 3+ messages in thread

From: Tom Lane @ 2006-12-15 17:57 UTC (permalink / raw)
  To: Joshua Colson <joshua.colson@ination.com>; +Cc: pgsql-admin

Joshua Colson <joshua.colson@ination.com> writes:
> Well, I ended up just using the pg_control file from the running
> installation.

The problem with that is that it'll have a fairly current
last-checkpoint pointer.  You really need a last-checkpoint pointer
pointing at where the rollforward needs to start.

			regards, tom lane



^ permalink  raw  reply  [nested|flat] 3+ messages in thread

* Re: Change "shutdown time" was: Change "database system
@ 2006-12-15 18:07  Joshua Colson <joshua.colson@ination.com>
  parent: Tom Lane <tgl@sss.pgh.pa.us>
  0 siblings, 0 replies; 3+ messages in thread

From: Joshua Colson @ 2006-12-15 18:07 UTC (permalink / raw)
  To: pgsql-admin

On Fri, 2006-12-15 at 12:57 -0500, Tom Lane wrote:
> The problem with that is that it'll have a fairly current
> last-checkpoint pointer.  You really need a last-checkpoint pointer
> pointing at where the rollforward needs to start.

Can't I just reset the last-checkpoint pointer using pg_resetxlog, or
will it not do that?

Thanks.

-- 
Joshua Colson <joshua.colson@ination.com>
iNation, LLC




^ permalink  raw  reply  [nested|flat] 3+ messages in thread


end of thread, other threads:[~2006-12-15 18:07 UTC | newest]

Thread overview: 3+ messages (download: mbox mbox.gz follow: Atom feed)
-- links below jump to the message on this page --
2006-12-15 15:23 Re: Change "shutdown time" was: Change "database system Joshua Colson <joshua.colson@ination.com>
2006-12-15 17:57 ` Tom Lane <tgl@sss.pgh.pa.us>
2006-12-15 18:07   ` Joshua Colson <joshua.colson@ination.com>

This inbox is served by DDX for PostgreSQL; see mirroring instructions
for how to clone and mirror all data and code used for this inbox