Default servers are not available/Can't find server name for address - disaster after changing ip block

Peter Rose peter.rose1 at ntlworld.com
Thu Sep 14 17:24:00 UTC 2000


Hello -
Desperate plea for help here -
I have been running a Red Hat Linux 5.2 server for the last year with my

own dns (BIND 8) on the same box.
Everything was working fine.
My provider then decided to move the entire operation to another
datacenter and my box was relocated and I have just been assigned a new
block of ip addresses.
I amended my /etc/hosts to the new range of ip's and did the same in all

the zonefiles and amened named.conf also.
I also amended the entries in resolv.conf
However, although named starts, it is not getting the names of my local
nameservers.
When I do nslookup I get this:

[root at server cyberscreen]# nslookup
*** Can't find server name for address 63.141.195.52: Server failed
*** Default servers are not available

If I omit resolv.conf altogether and let it fall back on hostname, I get

this:

[root at server cyberscreen]# nslookup
Default Server:  server.cyberscreen.com
Address:  0.0.0.0

and it will then give correct lookups for my virtual hosts, like so:

> www.happyhands.ws
Server:  server.cyberscreen.com
Address:  0.0.0.0

Non-authoritative answer:
Name:    server.cyberscreen.com
Address:  63.141.195.100
Aliases:  www.happyhands.ws

So I figured that my resolv.conf is screwed - but I can't see where:

domain                          cyberscreen.com
nameserver                      63.141.195.52
nameserver                      63.141.195.52
-----------
To complicate matters, just prior to my box being moved, I was given one

new IP address to use for my secondary nameserver (ns2.cyberscreen.com)
and I made the changes with InterNic.

I have only just allocated a new IP to ns1.cyberscreen.com and this
change is still being processed by internic. As all my zonefiles have
ns1.cyberscreen.com as the SOA, I'm wondering if this could cause
problems - but I still don't see why nslookup can't get the names from
/etc/hosts

I am also getting database errors in the log for my local.reverse file,
ie

Sep 14 15:08:14 server named[304]: local.reverse:7: Database error ())
Sep 14 15:08:14 server named[304]: local.reverse:9: Database error ())
Sep 14 15:08:14 server named[304]: local.reverse:12: Database error
(ns2.cyberscreen.com.)
Sep 14 15:08:14 server named[304]: local.reverse:14: Database error
(ns2.cyberscreen.com.)

but I can't for the life of me see where the faults are because all I've

really done to a perfectly working set of files is change the ip's.

I'm now thinking that maybe it's something to do with the way my
provider has relocated the box - for example I notice this error
message:

Sep 14 16:11:55 server named[304]: Lame server on '104.195.141.63.in-
addr.arpa'
(in '141.63.in-addr.arpa'?): [207.98.128.65].53 'NS2.GOOD.NET'
Sep 14 16:11:55 server named[304]: Lame server on '104.195.141.63.in-
addr.arpa'
(in '141.63.in-addr.arpa'?): [207.98.128.33].53 'NS1.GOOD.NET'
Sep 14 16:11:55 server named[304]: Lame server on '104.195.141.63.in-
addr.arpa'
(in '141.63.in-addr.arpa'?): [207.98.128.97].53 'NS3.GOOD.NET'

As 104.195.141.63.in-addr.arpa is one of my new IP's, does this imply
that my nameservers are actually being delegated to NS1.GOOD.NET (which
is located remarkably close to my providers's previous datacenter) and
there is a problem there?

If anyone can shed any light on this or suggest whether it's my cockup
or something upstream, I would be grateful. I now have 32 virtual hosts
with their email aliasing down:-)

Peter Rose
cyberscreen
London UK

"I don't like your fashion business mister"
  - Leonard Cohen / First We Take Manhattan





More information about the bind-users mailing list