From: | John Scalia <jayknowsunix(at)gmail(dot)com> |
---|---|
To: | "pgsql-admin(at)postgresql(dot)org" <pgsql-admin(at)postgresql(dot)org> |
Subject: | recovery.conf restore_command |
Date: | 2014-07-08 16:19:59 |
Message-ID: | 53BC1A2F.4010501@gmail.com |
Views: | Raw Message | Whole Thread | Download mbox | Resend email |
Thread: | |
Lists: | pgsql-admin |
Hi all,
I was experimenting with the restore_command on one of my standby systems. Originally this command just had:
restore_command='cp /mnt/wallogs/archive/%f %p'
And it was, and is working just fine, but since that doesn't delete the WAL files from the archive directory, I wanted to try:
restore_command='pg_standby /mnt/wallogs/archive %f %p %r'
But, the database then refused to start. The log showed:
LOG: entering standby mode
FATAL: could not restore file "00000003.history" from archive: return code 32512
LOG: startup process (PID 22847) existed with exit code 1
Now, that history file was already in the standby's pg_xlog directory. So, I tried copying it back to the archive directory and trying again, but no joy. I double-checked the
permissions on the file and the directory and those were good. So, I just changed the restore_command back to its original form, and the standby then started successfully. Did I
use pg_standby incorrectly? FYI, these servers are streaming replication using Version 9.3.3.
--
Jay
From | Date | Subject | |
---|---|---|---|
Next Message | Gilberto Castillo | 2014-07-08 17:29:56 | Re: recovery.conf restore_command |
Previous Message | Kevin Grittner | 2014-07-08 14:22:39 | Re: [ADMIN] WARNING: database must be vacuumed within 8439472 transactions |