Authoritative server recursing?

Daniele Orlandi daniele at orlandi.com
Wed Jun 1 14:21:34 UTC 2005


Hello,

I'm seeing this strange problem, maybe it's me... but it's the first time I
see such behaviour.

I have an authoritative master name server for the 10.212.62.in-addr.arpa
The zone contains a 16.10.212.62.in-addr.arpa IN NS record with
mix.b-soft.it

Another autoritative slave server has older, incorrect data.

The master server can recurse though it's not used as a resolver.

Here is the strangeness:

stef:~ # dig @vega -x 62.212.10.16 NS

; <<>> DiG 9.2.2 <<>> @vega -x 62.212.10.16 NS
;; global options:  printcmd
;; Got answer:
;; ->>HEADER<<- opcode: QUERY, status: NOERROR, id: 33042
;; flags: qr rd ra; QUERY: 1, ANSWER: 2, AUTHORITY: 0, ADDITIONAL: 0

;; QUESTION SECTION:
;16.10.212.62.in-addr.arpa.     IN      NS

;; ANSWER SECTION:
16.10.212.62.in-addr.arpa. 85140 IN     NS     
mix.b-soft.com.16.10.212.62.in-addr.arpa.
16.10.212.62.in-addr.arpa. 85140 IN     NS     
zagor.b-soft.com.16.10.212.62.in-addr.arpa.

;; Query time: 2 msec
;; SERVER: 62.212.0.2#53(vega)
;; WHEN: Wed Jun  1 16:17:14 2005
;; MSG SIZE  rcvd: 92

------------------------------------------------------

stef:~ # dig +norecurse @vega -x 62.212.10.16 NS

; <<>> DiG 9.2.2 <<>> +norecurse @vega -x 62.212.10.16 NS
;; global options:  printcmd
;; Got answer:
;; ->>HEADER<<- opcode: QUERY, status: NOERROR, id: 35930
;; flags: qr ra; QUERY: 1, ANSWER: 0, AUTHORITY: 1, ADDITIONAL: 1

;; QUESTION SECTION:
;16.10.212.62.in-addr.arpa.     IN      NS

;; AUTHORITY SECTION:
16.10.212.62.in-addr.arpa. 86400 IN     NS      mix.b-soft.com.

;; ADDITIONAL SECTION:
mix.b-soft.com.         171528  IN      A       62.212.3.230

;; Query time: 1 msec
;; SERVER: 62.212.0.2#53(vega)
;; WHEN: Wed Jun  1 16:17:26 2005
;; MSG SIZE  rcvd: 87

In other words, if I ask for recursion, the master server asks to the slave
and returns data from it. If I don't ask for recursion it responds
correctly.

The same happens if I disable recursion on the master.

Any idea?

thanks!

-- 
Daniele Orlandi



More information about the bind-users mailing list