dig +trace problems

Will Yardley &- at no.spam.veggiechinese.net
Tue Jul 16 00:38:46 UTC 2002


In article <agvosn$ftmf$1 at isrv4.isc.org>, Mark_Andrews at isc.org wrote:
>> In article <agvcfs$fm9u$1 at isrv4.isc.org>, Simon Waters wrote:
 
>>>> Why does dig +trace fail on this:
>>>> jazz% dig +trace ilove.lancaster.pa.us.
>
>>>> dig: couldn't get address for 'B.GTLD.BIZ': not found
  
>>> I can get BIND 9.2.1 to answer for b.gtld.biz from glue, and
>>> then if I query "dig gtld.biz ns" it caches the non-existence of
>>> "b". Your problem dig query still works with dig from 9.2.1,
>>> whichever way I do it.
 
>> d'oh - should have checked that.  Works fine with 9.2.1 DiG:

> Well it looks like your platform has a broken getaddrinfo()
> implementation.  BIND 9.3.0a0 (aka snapshot) sets AI_ADDRCONFIG
> whereas BIND 9.2.1 doesn't.

FWIW, it's running FreeBSD on i386 (4.6 stable):

jazz% uname -srmn
FreeBSD jazz.hq.newdream.net 4.6-STABLE i386

Don't know if this helps, but:

jazz% grep -n AI_ADDRCONFIG /usr/include/netdb.h
170:    (AI_PASSIVE | AI_CANONNAME | AI_NUMERICHOST | AI_ADDRCONFIG)
174:#define     AI_ADDRCONFIG   0x00000400 /* only if any address is
assigned */
177:#define     AI_DEFAULT      (AI_V4MAPPED_CFG | AI_ADDRCONFIG)

Let me know if there's information that might be helpful in figuring
this out.

What's odd is that I could reproduce the problem on another machine
before, but both machines now seem to work just fine now that
A.GTLD.BIZ. can be reached again.

-- 
No copies, please.
To reply privately, simply reply; don't remove anything.


More information about the bind-users mailing list