From: | Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us> |
---|---|
To: | "Joshua D(dot) Drake" <jd(at)commandprompt(dot)com> |
Cc: | pgsql-hackers(at)postgresql(dot)org |
Subject: | Re: Simple thing to make pg_autovacuum more useful |
Date: | 2008-01-17 22:38:57 |
Message-ID: | 16252.1200609537@sss.pgh.pa.us |
Views: | Raw Message | Whole Thread | Download mbox | Resend email |
Thread: | |
Lists: | pgsql-hackers |
"Joshua D. Drake" <jd(at)commandprompt(dot)com> writes:
> Your objection is let's keep it as difficult as possible within the
> existing paradigm because nobody thought pg_autovacuum could be useful
> in the first place.
No, my point is that there's no value in putting band-aids on an object
that was never designed to be user-friendly. The extra ease of use from
putting defaults on that table's columns is insignificant compared to
what we'd get by fixing its *real* problems:
* superuser-only, no mechanism to let users admin their own tables
(nor any way to reconcile user-set values with a DBA's possible
wish to override them)
* no support for dumping and restoring settings
I don't think we should be encouraging direct manual insertions into
pg_autovacuum in any case.
So I'd rather see some effort spent on figuring out what the API really
*should* look like. I don't know, other than that it should hard-wire
as little as possible because we are likely to be changing the set of
available parameters in future. Maybe we need a concept like per-table
settings for GUC variables?
regards, tom lane
From | Date | Subject | |
---|---|---|---|
Next Message | Hope Ho | 2008-01-17 22:51:30 | modularity of PostgreSQL |
Previous Message | Joshua D. Drake | 2008-01-17 22:31:08 | Re: Simple thing to make pg_autovacuum more useful |