Hi Chris,
Assuming the error condition is resolved, the process should restart automatically. However, one issue we have found is that there appears to be an upper limit to how many files `test` can actually check for existence. We found that it was returning 1 in error and hence wasn't archiving the files. Replacing test with the perl equivalent seemed to sidestep this issue.
Your archive command would look something like:
perl -e 'exit 1 if -e "/data/pgsql/backups/wal_arch/%f.gz"' && cp %p /var/lib/pgsql/backups/wal_arch/%f
Don't forget to properly escape the quotes.
You should be able to update the archive command by reloading the configuration: '/sbin/service postgresql reload' or equivalent for your system.
After the configuration is loaded (and you aren't hitting any other system limitations) it should begin archiving the files again.
Cheers,
Andrew
Hi, hoping someone can tell me how to get this running again.
I have pitr running and noticed that the slave is far out of date with wal logs.
Upon investigation I see that on the master that /data/pgsql/backups/wal_logs has over 6000 logs that haven't been moved
to the /var/lib/pgsql/backups/wal_arch/ folder.
The root partition that has /var had been full and logs may have been removed, and it may have run some time in this condition until someone corrected it.
postgresql.conf file.
test ! -f /data/pgsql/backups/wal_arch/%f.gz && cp %p /var/lib/pgsql/backups/wal_arch/%f
Is there a way to restart a service without taking the database down/up?
Any help would be appreciated.
Cheers,
Chris Barnes
Windows Live: Make it easier for your friends to see
what youre up to on Facebook.
