Hi,
Problem: When our network has any glitches causing a network time-out
for more than a few
minutes, our samba servers lose their connectivity with our WINS server
even after the network is
restored. This causes the WINS server to tombstone the name
registration for the samba server(s),
thereby causing clients to not connect through the browser. SMB
connections are not affected so the
workaround is to either connect via ip address or stop/restart samba.
This is prevalent across all of
our samba servers regardless of operating system.
The following is an excerpt from the log.nmb of a Sun Solaris system,
running Solaris 2.6, Samba 2.0.6.
This was run at debug level 5 (nmbd -D -d 5). This entry shows timeouts
to our WINS server
(134.253.181.26) after network connectivity was lost for 5 minutes and
then restored. Any insight
would be appreciated.
Regards,
Ken Hatfield
[2000/02/24 10:02:57, 4] nmbd/nmbd_packets.c:(1632)
retransmit_or_expire_response_records: timeout for packet id 20094 to
IP 134.2
53.181.26 on subnet UNICAST_SUBNET
[2000/02/24 10:02:57, 4] nmbd/nmbd_packets.c:(1632)
retransmit_or_expire_response_records: timeout for packet id 20095 to
IP 134.2
53.181.26 on subnet UNICAST_SUBNET
[2000/02/24 10:02:57, 4] nmbd/nmbd_packets.c:(1632)
retransmit_or_expire_response_records: timeout for packet id 20096 to
IP 134.2
53.181.26 on subnet UNICAST_SUBNET
[2000/02/24 10:02:57, 4] nmbd/nmbd_packets.c:(1632)
retransmit_or_expire_response_records: timeout for packet id 20097 to
IP 134.2
53.181.26 on subnet UNICAST_SUBNET
[2000/02/24 10:02:57, 4] nmbd/nmbd_packets.c:(1632)
retransmit_or_expire_response_records: timeout for packet id 20098 to
IP 134.2
53.181.26 on subnet UNICAST_SUBNET
