[Samba] remove wins entries - samba 3

[Eric Hines]

Geoff,

Sorry about the hour; I didn't realize you were still up--I went to bed....

A number of questions, and some updates.  I can find no evidence of 
active named logging, although I did find one log with named entries.  
In particular, what is the relevant log(s)?  There is no syslog or 
system log.  Running a FIND on *log didn't turn up anything even 
remotely close.  I've obviously not got logging turned on properly....

I also notice that, where John's example has several instantiations of 
named running, I have only one, and it's very difficult to terminate 
that one--I have to kill <pid> to do it.  Service <daemon> restart works 
fine for all the others, and service named start works fine, too.  Just 
service stop/restart do not work--the latter hangs on the stop part. 

In the files below, why all the changes to mail from lserver1?  I 
thought from John's examples these were supposed to be the server name?

Geoff Scott wrote:

>Eric Hines wrote:
>  
>
>>Geoff Scott wrote:
>>    
>>
>>>What do your logs say for bind starting up?  Can you restart bind and
>>>watch your logs?  Do you have any errors for it?
>>>
>>f you mean winbind, a tail -f on log.winbindd just showed it
>>    
>>
>No Berkely Internet Name Daemon - BIND
>The daemon is actually named "named"
>
>Grep for the entries for that daemon (named)in the relevant log,
>/var/log/....... Syslog?
>  
>
In log /var/log/messages, named starts successfully, loads all the zone 
files OK, and it outputs the log entry "lame server resolving 
'lserver1.test.biz' (in 'test.biz'?): 206.16.250.17#53, also ... .18#53 
several times.  These are owned by a company in Barcelona, Spain.  There 
also are cases (fewer) of resolving localhost.lserver1.test.biz to the 
same IP addresses/ports.  tail -f messages and pinging lserver1 produced 
no immediate result. I could find no other log that had named entires in 
it.  According to log.nmbd, Samba server LSERVER1 and samba name server 
LSERVER1 repeatedly became domain master browser and local master 
browser, respectively, on 192.168.1.103.  tail -f log.nmbd also did not 
respond to an unsuccessful ping of lserver1.

You asked whether I could tell my router/firewall not to send dhcp stuff 
to lserver1 only.  That would take a specific MAC address exclusion 
capability, and this router/firewall does not have that.  Can I, 
instead, tell lserver1 not to look to the router/firewall, but only to 
look to itself (/e.g./, via the dhcpd.conf or via lserver1's System 
Settings|Network GUI, using the DNS and/or hosts tab)?  Or would that 
lock lserver1 into itself, never to get access to the Internet?

I've done some other poking around in response to the DNS doc for which 
you sent me the URL last night, and noticed these things:
    /etc/sysconfig/networking/devices/ifcfg-eth0 is set as follows 
(emphasis added)
       DEVICE=eth0
       BOOTPROTO=dhcp
       ONBOOT=yes
       TYPE=Ethernet
       DHCP_HOSTNAME=*lserver1*
I have the same thing for eth1 (there are two NIC chips on the 
motherboard), except it's turned off.

dhcpd.leases has pserver1 (my print server) at 192.168.1.96, even though 
it's hardwired via its own setup functionality to a static address of 
198.162.1.10, and it responds to pings at the .10 address.

Finally, I made the zone file changes, and I still cannot ping lserver1 
or lserver1.test.biz--unknown host in both cases.

><snip>
>
>Regards Geoff Scott
>
-- 
He can compress the most words into the smallest ideas of any man
I ever met.
  - Abraham Lincoln