Re: PITR - warm standby switchover question

On Apr 14, 2009, at 3:47 PM, Dan Hayes wrote:

> I've followed the implementation instructions at 24.4.2:
> http://www.postgresql.org/docs/current/static/warm-standby.html
>
> And I've used the archive/restore commands from the example in F23.2:
> http://www.postgresql.org/docs/current/static/pgstandby.html
>
> This all works great. The primary backs up the WAL archives to a
> mutually accessible folder and the backup system is continuously
> reading in the logs. Monitoring the "standby.log" file shows that
> the system is working as it should.
>
> My question is, if the main system goes down, what do I do to the
> backup machine to make it come alive? Currently, when I attempt to
> login to the database, I get the error: "psql: FATAL: the database
> is starting up" I'm assuming this means it can't be accessed while
> its in continuous recovery mode. I attempted to turn off postgresql
> on the main server and try connecting again, same error (was
> obvious, but I do things one step at a time...) So then I removed
> the "recovery.conf" file on the backup server and attempted to
> restart postgresql. It stopped fine, but the restart failed.
>
> The pg_log/postgresql-XXXX-XX-XX.log file says:
> LOG: database system was interrupted while in recovery at log time
> 2009-04-14 17:36:14 CDT
> HINT: If this has occurred more than once some data might be
> corrupted and you might need to choose an earlier recovery target.
> LOG: could not open file "pg_xlog/0000000100000002000000DD" (log
> file 2, segment 221): No such file or directory
> LOG: invalid primary checkpoint record
> LOG: could not open file "pg_xlog/0000000100000002000000DC" (log
> file 2, segment 220): No such file or directory
> LOG: invalid secondary checkpoint record
> PANIC: could not locate a valid checkpoint record
> LOG: startup process (PID 3756) was terminated by signal 6: Aborted
> LOG: aborting startup due to startup process failure
>
> Any ideas what I'm doing wrong? How do I bring the standby server
> out of standby do act as the primary?

You should use pg_standby's -t flag to specify a stop file. Then all
you need to do is touch (create) that file and pg_standby will let the
server come out of recovery mode into normal operation mode. Be sure
to rm or mv the recovery.conf once that is complete.

Erik Jones, Database Administrator
Engine Yard
Support, Scalability, Reliability
866.518.9273 x 260
Location: US/Pacific
IRC: mage2k