Re: syslogging oddity - Mailing list pgsql-hackers

From Magnus Hagander
Subject Re: syslogging oddity
Date
Msg-id 20070723075307.GC29554@svr2.hagander.net
Whole thread Raw
In response to Re: syslogging oddity  (Tom Lane <tgl@sss.pgh.pa.us>)
Responses Re: syslogging oddity  (Tom Lane <tgl@sss.pgh.pa.us>)
List pgsql-hackers
On Sun, Jul 22, 2007 at 08:05:12PM -0400, Tom Lane wrote:
> Andrew Dunstan <andrew@dunslane.net> writes:
> > Somewhere along the way we seem to have made the syslogger's shutdown 
> > message go to stderr, even if we have redirected it:
> 
> I'm pretty sure it has done that all along; at least the design
> intention is that messages generated by syslogger itself should go to
> its stderr.  (Else, if the logger is having trouble, you might never get
> to find out why at all.)

Yeah, I think it's been that way all along.

> It might be reasonable to reduce "logger shutting down" to DEBUG1
> or so, now that the facility has been around for awhile.

+1.

For example, many windows system have *only* that message in the eventlog,
and nothing else... Which is kind of strange.

It could be interesting to have it write it *to the logfile* though, since
it'd then at least be in the same place as the others. As in special-casing
this one message, and just ignore logging it in case it fails. But think
we're fine just dropping the level.

//Magnus


pgsql-hackers by date:

Previous
From: Tom Lane
Date:
Subject: Re: syslogging oddity
Next
From: "Simon Riggs"
Date:
Subject: Re: 8.2 is 30% better in pgbench than 8.3