[Samba] SAMBA 3.0.22 unusual nmbd spawning

[Martin Vogt]

Hi,

we had upgraded our SAMBA PDC from 3.0.14 to 3.0.22 on a Gentoo system.
Now we have problems with peculiar spwaning of too many "nmbd" processes.
In smb.conf WINS support is activated, so I would expect two nmb
processes. But as you can see in the process list, during SAMBA is running
after some minutes another nmbd child is spawned. I tracked down the
problem to one example:

We have one primary domain ITCAMPUS, our PDC is BAIKONUR. PIONEER is an
ubuntu host running samba 3.0.22, WINS server disabled. As you can see
from the following snippet of nmbd.log nmbd on BAIKONUR tries to resolve
PIONEER, first per WINS then via DNS lookup and dies.
Afterwards I added PIONEER to our LDAP backend (serves to SAMBA+DNS+DHCP)
and spawning ends. We have several windows hosts in our network, not every
guest host is resolvable via DNS. So every time a foreign WINDOWS host
with its own WORKGROUP is joining our network I expect new nmbd spawning.
We never had such problems with samba 3.0.14, any ideas are appreciated?
thx (Hint: Process list and nmbd.log are taken from different days.)

greetings,
Martin


P.S.
another question: What causes this? (also in bigger nmbd.log snippet
below) BAIKONUR as WINS server don't recognises itself?
...
nmbd[15468]: 	ITCAMPUS(1) current master browser = UNKNOWN
nmbd[15468]: 		BAIKONUR 40099b2b (Samba Server 3.0.22)
...



# /etc/smb.conf
[global]
...
    domain logons = Yes
    os level = 254
    preferred master = Yes
    domain master = Yes
    wins support = yes
    name resolve order = lmhosts wins hosts bcast
...


# nmbd.log (LOGLEVEL 10)
nmbd[15468]: 	ITCAMPUS(1) current master browser = BAIKONUR
nmbd[15468]: 		BAIKONUR 400c9b2b (Samba Server 3.0.22)
nmbd[15468]: 		PROMETHEUS 40011003 ()
nmbd[15468]: 		TRANSPLUTO 40011003 (Jörg)
nmbd[15468]: 		LEDA 40011203 (R16)
nmbd[15468]: 		XPGSX 40011003 ()
nmbd[15468]: 		KALLISTO 40011403 ()
nmbd[15468]: 		XPVM 40011003 ()
nmbd[15468]: dump_workgroups()
nmbd[15468]: dump workgroup on subnet  UNICAST_SUBNET: netmask=172.16.x.xx:
nmbd[15468]: 	TEMP(5) current master browser = MESSE-PC-1
nmbd[15468]: 	MSHOME(4) current master browser = PIONEER
nmbd[15468]: 	ARBEITSGRUPPE(3) current master browser = BISONDEMO
nmbd[15468]: 	MSXNET(2) current master browser = MSX
nmbd[15468]: 	ITCAMPUS(1) current master browser = UNKNOWN
nmbd[15468]: 		BAIKONUR 40099b2b (Samba Server 3.0.22)
nmbd[15468]: find_workgroup_on_subnet: workgroup search for ITCAMPUS on
subnet UNICAST_SUBNET: found.
nmbd[15468]: find_workgroup_on_subnet: workgroup search for ITCAMPUS on
subnet UNICAST_SUBNET: found.
nmbd[15468]: Initiating DMB<->DMB sync with PIONEER(0.0.0.0)
nmbd[15650]: Initiating browse sync for MSHOME to PIONEER(0.0.0.0)
nmbd[15650]: no entry for PIONEER#20 found.
nmbd[15650]: resolve_lmhosts: Attempting lmhosts lookup for name
PIONEER<0x20>
nmbd[15650]: getlmhostsent: lmhost entry: 127.0.0.1 localhost
nmbd[15650]: getlmhostsent: lmhost entry: 192.168.xxx.x HOUSTON
nmbd[15650]: getlmhostsent: lmhost entry: 172.16.x.xx GSX
nmbd[15650]: resolve_wins: Attempting wins lookup for name PIONEER<0x20>
nmbd[15650]: wins_srv_is_dead: 127.0.0.1 is dead
nmbd[15650]: resolve_hosts: Attempting host lookup for name PIONEER<0x20>



# ps faux
root     32085  ... 00:59   0:00 /usr/sbin/smbd -D
root     32092  ... 00:59   0:00  \_ /usr/sbin/smbd -D
root       676  ... 06:37   0:01  \_ /usr/sbin/smbd -D
...
root     32087  ... 00:59   0:05 /usr/sbin/nmbd -D
root     32088  ... 00:59   0:00  \_ /usr/sbin/nmbd -D
root      1056  ... 08:05   0:00  \_ /usr/sbin/nmbd -D
root     30865  ... 08:36   0:00  \_ /usr/sbin/nmbd -D
root     31055  ... 09:21   0:00  \_ /usr/sbin/nmbd -D
root     31271  ... 10:11   0:00  \_ /usr/sbin/nmbd -D
root     32090  ... 00:59   0:00 /usr/sbin/winbindd
root     32091  ... 00:59   0:00  \_ /usr/sbin/winbindd
root     32551  ... 03:08   0:00  \_ /usr/sbin/winbindd