Re: pg_start/stop_backup non-exclusive scripts to snapshot

Melvin,

* Melvin Davidson (melvin6925(at)gmail(dot)com) wrote:
> On Wed, Jul 5, 2017 at 10:14 AM, Stephen Frost <sfrost(at)snowman(dot)net> wrote:
> >Part of my concern is that such a script is unlikely to show any problems
> until it comes time to do a restore
> As previously stated, the script was used to set up a slave and has done so
> successfully many times. There are subsequent scripts
> that check results.

Ah, the impression was that this was being suggested as a way of
performing regular backups. For simply creating a replica,
pg_basebackup works pretty well for small systems. For larger
databases, being able to perform parallel backup/restore is very useful,
even if it's just for building a replica.

> >What might be worse would be to pull the plug while the backup is running
> and then try to bring the primary back online.
> Uh, whom would be so stupid as to do that?

Pulling the plug in the middle of various operations is a good way to
simulate what happens if the system crashes, for whatever reason, and to
make sure that processes and procedures are in place to address such a
failure scenario.

> >Right, there's little sense in trying to perfect a shell script when
> proper solutions exist.
> >>It's better to create something that others criticise than to create
> nothing and criticise others. Go create, have fun!!

Indeed, I'd certainly welcome criticism of pgBackRest.

Thanks!

Stephen