On Wed, Jun 07, 2006 at 05:21:47PM +0200, Marco Fretz wrote: > hello there > > we have a a problem with Bind 9.3.0 and caching for a domain xxx.net. > the domain xxx.net is delegated to some other dns server. we try to > resolve relay.xxx.net and no problems. a few hours later we're unable to > get a answer for IN A relay.xxx.net from our dns server. > > after a "rndc flush" it's working again for some hours. may theres a > problem with "recursive querys" or "updates"? >
Probably the best place to ask is a BIND mailing list, which is for the app, rather than at at OS list, but I'll have a go: When there is a problem with the domain in question, try the following: $ dig @your.cache A relay.xxx.net +short $ dig @ns.xxx.net A relay.xxx.net +short (to get the authoritive domain name servers, try this: $ dig xxx.net NS +short) If there is a difference in what your cache delivers, check the zone's serial number: $ dig @your.cache SOA relay.xxx.net +short $ dig @ns.xxx.net SOA relay.xxx.net +short If there is a different serial number, the authoritive servers have an updated zone, and you cache does not. The hostmaster for the zone may have set the zone expire record to be of a longer time period than he expected changes to occur, therefore any caching servers will hold a record until it expires. There could be other reasons, but as I'm pissed, that's all I can think of at the moment. Hic. Like that Stella Artois...... And Caledonain 80/-. -- Craig Skinner | http://www.kepax.co.uk | [EMAIL PROTECTED]
