From: | Evan Martin <postgresql(at)realityexists(dot)net> |
---|---|
To: | pgsql-general(at)postgresql(dot)org |
Subject: | Re: Please make it easy to drop a database that is in use |
Date: | 2012-06-21 14:50:07 |
Message-ID: | 4FE3349F.7040604@realityexists.net |
Views: | Raw Message | Whole Thread | Download mbox | Resend email |
Thread: | |
Lists: | pgsql-general |
On 22/06/2012 12:07 AM, Tom Lane wrote:
>> SELECT pg_terminate_backend(procpid)
>> FROM pg_stat_activity
>> WHERE datname = 'dropme';
>> ERROR: must be superuser to signal other server processes
> As far as that goes, there's a pending patch to reduce the privileges
> required to use pg_terminate_backend. I'm not in favor of having DROP
> DATABASE do it for you though --- that just seems like a very
> large-caliber foot gun.
>
Good to hear there's a patch coming for that. Just to be clear, I'm not
suggesting DROP DATABASE should do that by default. Basically, I'm
looking for the "--force" option here - a way to say "I know what I'm
doing, just drop this database if at all possible".
"rm -rf" is a foot gun, too, but if that -f wasn't there you would have
to write commands to set permissions on files when you couldn't care
less about the permissions and just want to delete them. Sure, a
competent Linux user could write such a command, but imagine how
annoying it would be to do that all the time! Not to mention the extra
room for errors in that command. To me, DROP DATABASE is a very similar
case.
Regards,
Evan
From | Date | Subject | |
---|---|---|---|
Next Message | David Pirotte | 2012-06-21 15:10:28 | Promoting sync slave to master without incrementing timeline counter? |
Previous Message | Marc Mamin | 2012-06-21 14:17:01 | (BUG ?) unprefixed oid -> ERROR: cache lookup failed for function |