bind9 logging

[Joerg Schumacher]

On Thu, Jul 10, 2003 at 08:37:07AM +1000, Mark.Andrews at isc.org wrote:

> > I've upgraded one of our servers from bind 8 to bind 9 and now I'm 
> > reworking the logfile analyser.  Some "popular" errors made by our
> > customers we've been able to catch in the bind 8 logs are:
> > 
> >    serial number decrements:
> >    	xfer-in: notice: Zone... SOA serial# (xxx) ... is < ours (xxx)
> 
> 	It's currently a debug message.
> 
> 	zone_debuglog(zone, me, 1, "ahead");

Thanks for the hint.  Any plans to raise the level from debug to error 
and include both serials?   A slave being "ahead" indicates almost 
always an error.  The operator of the slave should be able to detect 
this without debugging.

In the meantime I've found a message in the bind9-workers archive.
Looks like others do share my concerns:

   From: Danny Thomas <d.thomas () its ! uq ! edu ! au>
   Subject: patches for improved logging in 9.2.1
   Date:    2002-07-20 10:24:19

Any comments on the proposed changes in this article?

> >    listing our server as slave, but forgot to tell us about:
> > 	default: info: ns_forw: query(...) contains our address ...
> 
> 	With views sending to yourself is sometimes normal.

I'm not keen on the shortcut of bind8, I'm keen on the log message.
I think it's a horrible idea to log the "lame server" only on the
client side which might not be interested enough to get the error
fixed.  I do care about the quality of the dns, so I'd love to see 
in the logs for which zones our server is considered lame.  

Regards,
Joerg 

-- 
 Gaertner Datensysteme                         38114 Braunschweig 
 Joerg Schumacher                              Hamburger Str. 273a
 Tel: 0531-2335555                             Fax: 0531-2335556