Hallo, I have a NT_STATUS_NO_LOGON_SERVERS problem, after i have updated my samba server from version 2.0.7 to samba 3.0.2a. Logon works fine for about 18 minutes, then suddenly the communication between the SAMBA process and the WINNT PDC does not work anymore. I have checked my configuration as good as i can. e.g. dns resolution, local master browser, smb signing settings,schannel settings, guest account, remote browse sync, socket options, packet filtering, wins registration I dont'find any problem. I found two bugs on bugzilla, which seemed to me, as if they have nearly the same problem: Bug 870 "smbd spuriosly reports NT_STATUS_NO_LOGON_SERVERS" and Bug 331 "User authentification fails NT_STATUS_NO_LOGON_SERVERS ... " I have an interesting Workarround: When i use the command "net rpc join -S alien -U'Administrator%********' -W EPIDAUROS" it works again and the problem is away for about 18 Minutes. The same workaround was reportet in Bug 870. So i do this in a cronjob every 5 Minutes, on my production server. I have a Test Server where the same problem exist. What i do: samba 3.0.2a on solaris 8 ( Servername DATENNETZ, 192.168.1.1 ) SunOS e450 5.8 Generic_117000-01 sun4u sparc SUNW,Ultra-4 Windows DOMAIN "EPIDAUROS" winnt is a PDC and WINS Server ( Servername ALIEN , 10.1.161.1 ) The servers are in different subnets. On the solaris machine i only start nmbd and smbd. ( as i understand, i do not need a winbind process, because i have a remote WINS Server ) I have only 30 Users. On the testserver, i am the single user. The user accounts exist on the WINNT PDC whith password "A". On the solaris machine the User has a NIS account, whith the same login name, but a password "B". This concept has been implemented before i was in the company, it has worked whith samba 2.0.7. Could this be a problem whith samba 3.x? My smb.conf file is: [global] workgroup = EPIDAUROS netbios name = DATENNETZ server string = TESTSERVER mit Samba 3.0.2a security = DOMAIN password server = ALIEN client plaintext auth = No log level = 10 name resolve order = wins host bcast socket options = SO_KEEPALIVE IPTOS_LOWDELAY TCP_NODELAY preferred master = Yes domain master = No wins server = 10.1.161.1 ldap ssl = no remote announce = 10.1.191.255 [playground] comment = TEST Server, do not use path = /daten/playground read only = No force directory mode = 0770 I have generated the Tracefile whith trace log 2 and 10: You see the following: 1) server start 2) "smbclient -L datennetz -UEPIDAUROS/juergen0" added interface ip=192.168.1.34 bcast=192.168.1.255 nmask=255.255.255.0 Password: session setup failed: NT_STATUS_NO_LOGON_SERVERS 3) "net rpc join -S alien -U'Administrator%********' -W EPIDAUROS" Joined domain EPIDAUROS. 4) "smbclient -L datennetz -UEPIDAUROS/juergen0" added interface ip=192.168.1.34 bcast=192.168.1.255 nmask=255.255.255.0 Password: Domain=[EPIDAUROS] OS=[Unix] Server=[Samba 3.0.2a] Sharename Type Comment --------- ---- ------- playground Disk TEST Server, do not use IPC$ IPC IPC Service (TESTSERVER mit Samba 3.0.2a) ADMIN$ IPC IPC Service (TESTSERVER mit Samba 3.0.2a) Server Comment --------- ------- DATENNETZ TESTSERVER mit Samba 3.0.2a Workgroup Master --------- ------- EPIDAUROS This work now, for about 18 Minutes 5) i do a monitoring of samba using netsaint, which tests every 2 minutes using smbclient 6) suddenly the error is reportet 7) "smbclient -L datennetz -UEPIDAUROS/juergen0" added interface ip=192.168.1.34 bcast=192.168.1.255 nmask=255.255.255.0 Password: session setup failed: NT_STATUS_NO_LOGON_SERVERS 8) stoping the server Regards Juergen Lepper
Created attachment 475 [details] trace log 2
Created attachment 476 [details] trace log 10 ( compressed tar ) of log.nmbd log.smbd log.winbind
What OS is ALIEN ? Any idea why it is not responding to a node status request on 137/udp ? internal_resolve_name: looking up ALIEN#20 name ALIEN#20 found. remove_duplicate_addrs2: looking for duplicate address/port pairs get_dc_list: returning 1 ip addresses in an ordered list get_dc_list: 10.1.161.1:0 name_status_find: looking up EPIDAUROS#1c at 10.1.161.1 Cache entry with key = NBT/EPIDAUROS#1C.20.10.1.161.1 couldn't be found namecache_status_fetch: no entry for NBT/EPIDAUROS#1C.20.10.1.161.1 found. Deleting cache entry (key = NBT/EPIDAUROS#1C.20.10.1.161.1) bind succeeded on port 0 Sending a packet of len 50 to (10.1.161.1) on port 137 Sending a packet of len 50 to (10.1.161.1) on port 137 name_status_find: name not found check_ntdomain_security: unable to locate a DC for domain EPIDAUROS
Thank's for your hint! The problem was caused by the PDC ( Servername ALIEN, WINNT 4.0 whith SP6 + hotfixes). As i checked netstat on the WINNT Server, i found different Problems on this server. netstat didn't work. The follwoing error where shown: 1) The interfaceinitialization had an error: 2 2) PopUp Windows whith snmpapi.dll Error I had to reinstall SP6 on this server to solve the above errors. Now netstat workes fine and shows 137/udp listening. Samba 3.x on solaris works fine too. I suppose, that the installation of several hotfixes from Microsoft, had caused the problem on the winnt box. A fellow has installed these hotfixes on the same date i was updating samba on my solaris server. Sorry, for taking your time. Regards Juergen
Windows NT 4.0 DC issue. Closing as 'not our bug'
database cleanup