Hello all,

I'm new here but I've being using bind from time to time since year 2002.
However I'm still quite noob.

I'm trying to configure bind 9.3.4/Debian Etch with a master zone and a
subdomain in it (I'm following this howto [1]). The master zone is
example.com and its subdomain is tx.example.com (I obviously used a real
domain of mine instead of example.com, but I prefer not reporting the real
name because I'm not sure about the security of nstx).

Based on what's written in the howto I've compiled the following
example.com.db file:

; BIND data file for example.com
$TTL 604800
@ IN SOA example.com. info.example.com. (
2008072502 ; Serial
7200 ; Refresh
120 ; Retry
2419200 ; Expire
604800) ; Default TTL
@ IN NS ns1.example.com.
;@ IN NS ns2.example.com.
example.com. IN MX 10 mail2.example.com.
example.com. IN A x.y.z.47
ns1 IN A x.y.z.47
www IN CNAME example.com.
mail2 IN CNAME www.example.com.
fioredec IN CNAME www.example.com.
tx IN NS ns.tx.example.com.
$ORIGIN tx.example.com.
@ IN NS ns.tx.example.com.
ns IN A x.y.z.48

Bind9 is listening on the interface x.y.z.47 and nstxd is listening (udp
only) on the interface x.y.z.48.

If I query nstxd directly I get what seems to be a reasonable reply (for
that daemon at least):

# dig @x.y.z.48 whatever.tx.example.com
[...] NOERROR [...] IN TXT [...]

So I assume nstxd it's working ok. But if I ask Bind9 who's the nameserver
for tx.example.com I get a SERVFAIL:

# dig @x.y.z.47 -t ns tx.example.com
[...] SERVFAIL [...]

The same if I dig whatever.tx.example.com using Bind9.

What am I doing wrong?

Thanks in advance,

[1] http://thomer.com/howtos/nstx.html