refresh_callback: zone , after bind9 upgrade

[Mark.Andrews at nominum.com]

> In article <9kppvt$qf at pub3.rc.vix.com>,  <Mark.Andrews at nominum.com> wrote:
> >
> >	A refresh_callback is the code called when a refresh query
> >	succeeds or fails.  A refresh query is the query made to
> >	see if the zone is up to dat.  The server will make three
> >	of these before it gives up and reschedules roughly retry
> >	seconds later.  The failure is that the nameserver didn't
> >	get a response in the expected time (timed out).
> 
> Is this the log message that replaced:
> 
> named[<pid>: Err/TO getting serial# for "<domain>"
> named-xfer[<pid>]: connect(<addr>) for zone <domain> failed: Connection timed
>  out

	Basically.

> 
> BIND's log messages have never really been the epitome of clarity.

	Syslog messages are not supposed to be large.  We try and
	get the pertinent information into the message (BIND 9.1
	dropped the ball somewhat there).  In this case there is
	not much more we could report other that "we timed out".

	The ARM does need a appendix which covers the error messages
	and what they mean.

>  But
> from what I've seen in the newsgroup (we're not yet running BIND 9 here),
> it seems like BIND 9 has made things even worse.  Shouldn't a complete
> rewrite have given you the opportunity to improve this?

	The re-write had to do more with code managability and
	correctness of operation, new protocol features, IPv6
	support.  Clearer error messages was further down the list.
	I think we have actually achieved the primary goals of the
	BIND 9 re-write.

	Mark

> 
> -- 
> Barry Margolin, barmar at genuity.net
> Genuity, Woburn, MA
> *** DON'T SEND TECHNICAL QUESTIONS DIRECTLY TO ME, post them to newsgroups.
> Please DON'T copy followups to me -- I'll assume it wasn't posted to the grou
> p.
> 
--
Mark Andrews, Nominum Inc.
1 Seymour St., Dundas Valley, NSW 2117, Australia
PHONE: +61 2 9871 4742                 INTERNET: Mark.Andrews at nominum.com