From: | Magnus Hagander <magnus(at)hagander(dot)net> |
---|---|
To: | FAST PostgreSQL <fastpgs(at)fast(dot)fujitsu(dot)com(dot)au> |
Cc: | pgsql-patches(at)postgresql(dot)org |
Subject: | Re: COPY-able sql log outputs |
Date: | 2007-04-02 08:44:43 |
Message-ID: | 20070402084443.GA31512@svr2.hagander.net |
Views: | Raw Message | Whole Thread | Download mbox | Resend email |
Thread: | |
Lists: | pgsql-patches |
On Mon, Apr 02, 2007 at 04:27:33PM +0000, FAST PostgreSQL wrote:
> On Sat, 31 Mar 2007 14:09, you wrote:
> > FAST PostgreSQL wrote:
> > >> Am Dienstag, 3. April 2007 20:33 schrieb FAST PostgreSQL:
> > >>> Attached is the completed patch for the COPY-able sql log outputs.
> > >>
> > >> Could you please remove random whitespace changes from this patch?
> > >
> > > Done and attached.
> >
> > A couple of comments.
> >
> > First one that's not at all about the code - but your emails appear to
> > be dated several days into the future. This one, for example, is
> > datestamped 2007-04-04. AFAIK timezones only go 13 (or is it 14?) hours
> >
> > :) You'll want to look into that. (otherwise, it'll look like you missed
> >
> > feature freeze..)
>
> Sorry.. Fixed.
Not quite. Time looks correct, but it specifies a timezone that's ten hours
off:
Received: from there (unknown 137.172.75.153)
by fast.fujitsu.com.au (Scalix SMTP Relay 10.0.1.3)
via ESMTP; Mon, 02 Apr 2007 16:28:34 +1000 (EST)
Date: Mon, 2 Apr 2007 16:27:33 +0000
> > Looking at the code, I was first going to ask what happens when you
> > enable both redirect_stderr and this one - would the data be put in
> > random order in the same file. But looking at the code, it seems you're
> > appending ".sql" to the filename for this log. Two comments about that:
> >
> > 1) This is not mentioned anywhere in the docs or comments. It needs to be.
>
> Agreed. As I mentioned in the previous mail, preferably alongwith the
> recommended table structure to COPY the logs into. But where? Is adding a new
> paragraph under the log_destination section of 'Where to Log' chapter ok ?
I think that makes sense. The comment in postgresql.conf might also need
some updating to reflect it.
> > 2) Not sure if .sql is such a great appendix to use, given that it's not
> > actual SQL. (I will also echo the other comment that was posted about
> > the whole name being sqllog, when it's not SQL. I think a better name is
> > needed since the data is csv. Goes for all parameters and function names)
>
> 'csvlog' is another option. I can change that. Anyone have other preferences ?
I assume csvlog for parameter names, and .csv as the file extension? Or did
you mean csvlog for extension as well? In that case, I'd suggest .csv ;-)
//Magnus
From | Date | Subject | |
---|---|---|---|
Next Message | Andrew Dunstan | 2007-04-02 12:11:11 | Re: Current enums patch |
Previous Message | Heikki Linnakangas | 2007-04-02 08:27:18 | Re: Current enums patch |