weird name resolution problem w/ bind

[Klaus A. Kreil]

Hello BIND experts out there,
I do have a strange behaviour of bind (version 9.2.3 running on SuSE
9.1,
kernel 2.6.5-7.75-smp) as follows:

ping hp8550 (hp8550 is a local networked printer) from a ssh window
(putty) times out without resolving the name yielding an error message
as follows:

ping: unknown host hp8550

This is also the case if started with the FQDN (ping
hp8550.xxxxx.local) to
resolve for the IP address:
ping: unknown host hp8550.xxxxx.local

nslookup hp8550 from the same ssh console returns the correct FQDN
together
with the IP-number as follows:

Note: nslookup is deprecated and may be removed from future releases.
Consider using the `dig' or `host' programs instead. Run nslookup with
the `-sil[ent]' option to prevent this message from appearing.
Server: 192.168.xx.1
Address: 192.168.xx.1#53

Name: hp8550.xxxxx.local
Address: 192.168.xx.64

Same ping from a XP cmd window of a local PC connected to the local
domain
works without any problems: ping hp8550 resolves to the correct IP and
pings.



All external pings (i.e. hosts not in the local domain) with domain
names
resolve for the IP address and work as expected regardless whether
from XP
or ssh window. You can substitute bind with telnet, ftp, etc - all do
not
resolve correctly if executed from the ssh console.
BTW the same is true for the linux console

Setup is as follows:
authoritative bind for the domain xxxxx.local with "forwarders ..." to
the
ISP's name servers with "forward first". Restricted to allow queries
from
the local subnet + loopback and listening on dedicated local subnet
address
(192.168.xx.1) plus loopback.

At the moment I am clueless and would appreciate any suggestion.

Many thanks and best regards,

Klaus