Bind 8 caching problems

Mark.Andrews at nominum.com Mark.Andrews at nominum.com
Wed Feb 9 00:25:18 UTC 2000


	If you turn on host-statistics the origin of each record
	is returned.

	As for the underlying "problem".  The DNS has never guarenteed
	to give out consistent information when it is "new" information.
	In this case it looks like someone forgot to pay registation bill
	and the domain was removed from the com zone.  The bill has now
	been paid and the domain is back, however not all of the com
	servers have transfered the current zone yet.  Hence some say the
	domain does not exist and some do.  Named caches both types of
	answers.

	Restarting named just changes which com server gets asked.

	Mark

> You need to find out who is giving you the bad information. The dumps used to
> clearly indicate where negative cache entries came from, but I just looked at
>  a
> dump from a bind-8.2.2-p5 named, and apparently this is not shown any more. Y
> ou
> may have to enable debugging output in order to track down the source of the 
> bad
> data.
> 
> 
> - Kevin
> 
> Gildas PERROT wrote:
> 
> > Hi,
> >
> > Regularly, I notice some problems with my Bind 8.2.2p5 servers (one on Sola
> ris
> > 2.6 and one on FreeBSD 3.3). For certain zones for which they are not
> > authoritative, they are unable to answer to requests due to their cache.
> >
> > Ex :
> >
> > ; <<>> DiG 8.2 <<>> @hyperbate www.livinghair.com
> > ; (1 server found)
> > ;; res options: init recurs defnam dnsrch
> > ;; got answer:
> > ;; ->>HEADER<<- opcode: QUERY, status: NXDOMAIN, id: 6
> > ;; flags: qr rd ra; QUERY: 1, ANSWER: 0, AUTHORITY: 1, ADDITIONAL: 0
> > ;; QUERY SECTION:
> > ;;      www.livinghair.com, type = A, class = IN
> >
> > ;; AUTHORITY SECTION:
> > com.                    2h17m14s IN SOA  A.ROOT-SERVERS.NET.
> > hostmaster.internic.NET. (
> >                                         2000020700      ; serial
> >                                         30M             ; refresh
> >                                         15M             ; retry
> >                                         1W              ; expiry
> >                                         1D )            ; minimum
> >
> > ;; Total query time: 1 msec
> > ;; FROM: vectra.francenet.fr to SERVER: hyperbate  193.149.97.80
> > ;; WHEN: Tue Feb  8 10:17:29 2000
> > ;; MSG SIZE  sent: 36  rcvd: 113
> >
> > The contents of the cache :
> >
> > ;livinghair     10496   IN      SOA     A.ROOT-SERVERS.NET.
> > hostmaster.internic.
> > NET. (
> > ;               2000020700 1800 900 604800 86400 );com.;NXDOMAIN        ;-$
> > ;Cr=auth
> >
> > $ORIGIN livinghair.com.
> > ;www    10464   IN      SOA     A.ROOT-SERVERS.NET. hostmaster.internic.NET
> . (
> > ;               2000020700 1800 900 604800 86400 );com.;NXDOMAIN        ;-$
> > ;Cr=auth
> >
> > If I query directly the NS of livinghair.com, no problem.
> > If I restart the server, the cache is emptied and the problem is solved.
> > It seems that the cache is not expired after the TTL of the record.
> >
> > Any idea about that problem ?
> >
> > How often is it suggested to restart named ?
> >
> > Thanks in advance for your help.                        Gildas.
> > --
> > Gildas PERROT, perrot at francenet.fr         __o
> > FranceNet, 28 rue Desaix, 75015 Paris ---_ \<,_
> > http://www.francenet.fr            ---- (_)/ (_)
> 
> 
> 
> 
> 
--
Mark Andrews, Nominum Inc. / Internet Software Consortium
1 Seymour St., Dundas Valley, NSW 2117, Australia
PHONE: +61 2 9871 4742                 INTERNET: Mark.Andrews at nominum.com



More information about the bind-users mailing list