| From: | "Matthew T(dot) O'Connor" <matthew(at)zeut(dot)net> |
|---|---|
| To: | Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us> |
| Cc: | Magnus Hagander <mha(at)sollentuna(dot)net>, PostgreSQL Patches <pgsql-patches(at)postgresql(dot)org> |
| Subject: | Re: pg_autovacuum integration patch |
| Date: | 2004-06-16 14:08:29 |
| Message-ID: | 1087394909.20076.30.camel@zedora2 |
| Views: | Whole Thread | Raw Message | Download mbox | Resend email |
| Thread: | |
| Lists: | pgsql-patches |
On Wed, 2004-06-16 at 10:01, Tom Lane wrote:
> "Magnus Hagander" <mha(at)sollentuna(dot)net> writes:
> > I also noticed:
> > ! elog(ERROR, "pg_autovacuum: GUC variable stats_row_level
> > must be enabled.");
> > ! elog(ERROR, " Please fix the problems and try
> > again.");
>
> > If you use the ereport() call instead of elog, you can set the second
> > one as a HINT. Since it's really the same error, I don't think you want
> > multiple errors logged...
>
> Even more to the point, control will never *reach* the second elog().
> Matthew clearly needs to spend more time studying the backend error
> message reporting mechanism. There is some documentation here:
> http://developer.postgresql.org/docs/postgres/error-message-reporting.html
> and the backend code is certainly chock-full of examples.
No doubt, I'll try to take a look at this soon. Since the deadline is
coming up, I wanted to find out my shortcoming sooner rather than
later.
Put another way, I thought it was time to stop operating in a "VACUUM"
;-)
Matthew
| From | Date | Subject | |
|---|---|---|---|
| Next Message | Andrew Dunstan | 2004-06-16 14:21:27 | Re: pg_autovacuum integration patch |
| Previous Message | Matthew T. O'Connor | 2004-06-16 14:04:37 | Re: pg_autovacuum integration patch |