resolver change?

[Harti Brandt]

On Sun, 27 Jan 2008, Doug Barton wrote:

DB>Hartmut Brandt wrote:
DB>> Alexander Leidinger wrote:
DB>>> Quoting Dag-Erling SmЪЪЪЪrgrav <des at des.no> (from Sun, 27 Jan 2008
DB>>> 19:45:46 +0100):
DB>>>
DB>>>> Hajimu UMEMOTO <ume at mahoroba.org> writes:
DB>>>>> Dag-Erling SmЪЪrgrav <des at des.no> writes:
DB>>>>> > OK, so the resolver now uses a process-internal cache?  Is there
DB>>>>> any way
DB>>>>> > to turn it off?
DB>>>>> No, our resolver doesn't have a process-internal cache at all.
DB>>>>
DB>>>> So what's going on?
DB>>>>
DB>>>> Looking back through my logs, it was working correctly as late as
DB>>>> January 13, so something broke between then and January 21.
DB>>>>
DB>>>> My name server does *not* forward queries, it goes straight to the
DB>>>> source.  Everything looks fine if I run host(1) multiple times, it only
DB>>>> seems to fail when successive lookups are made from the same process.
DB>>>
DB>>> Are you seeing this with or without activated nscd/cached?
DB>> 
DB>> I see the same effect as DES. Tried to disable hosts caching in nscd.
DB>> Didn't change anything.
DB>
DB>What happens if you disable nscd completely?

Could it be that this was a problem with the DNS server? Today it doesn't
matter whether I disable or enable the cache, I get three different IP
addresses for 5 servers in ntp.conf. Maybe it was rotating the addresses to
slow or what?

harti