[Samba] sometimes users fails to login

[Rowland Penny]

On Tue, 12 Mar 2019 11:32:46 +0100
Andrea Cucciarre' via samba <samba at lists.samba.org> wrote:

> Hello,
> 
> I have Samba 4.6 as AD domain member and sometime the users fails to 
> login, the issue disappear after some minutes.
> I have enabled log leve 10 and I can see the following errors:
> 
> 2019/03/12 09:20:32.280799,  5, pid=15466, effective(0, 0), real(0,
> 0)] ../source3/lib/username.c:181(Get_Pwnam_alloc)
>    Finding user BITINTRA\U002489
> [2019/03/12 09:20:32.281111,  5, pid=15466, effective(0, 0), real(0,
> 0)] ../source3/lib/username.c:128(Get_Pwnam_internals)
>    Trying _Get_Pwnam(), username as given is BITINTRA\U002489
> [2019/03/12 09:20:32.281222,  5, pid=15466, effective(0, 0), real(0,
> 0)] ../source3/lib/username.c:153(Get_Pwnam_internals)
>    Get_Pwnam_internals didn't find user [BITINTRA\U002489]!
> [2019/03/12 09:20:32.282015,  3, pid=15466, effective(0, 0), real(0,
> 0),
> class=auth] ../source3/auth/user_krb5.c:164(get_user_from_kerberos_info)
> get_user_from_kerberos_info: Username BITINTRA\U002489 is invalid on
> this system [2019/03/12 09:20:32.282043,  3, pid=15466, effective(0,
> 0), real(0,
> 0)] ../source3/auth/auth_generic.c:145(auth3_generate_session_info_pac)
> auth3_generate_session_info_pac: Failed to map kerberos principal to
> system user (NT_STATUS_LOGON_FAILURE) [2019/03/12 09:20:32.282196,
> 3, pid=15466, effective(0, 0), real(0,
> 0)] ../source3/smbd/smb2_server.c:3097(smbd_smb2_request_error_ex)
> smbd_smb2_request_error_ex: smbd_smb2_request_error_ex: idx[1]
> status[NT_STATUS_ACCESS_DENIED] ||
> at ../source3/smbd/smb2_sesssetup.c:134
> 
> my understanding of the code is that getpwnam fails, which is
> supposed to query winbindd.
> In the log file log.wb-BITINTRA I can see the following error:
> 
> [2019/03/12 09:20:24.540456, 10, pid=15439, effective(0, 0), real(0,
> 0),
> class=winbind] ../source3/winbindd/winbindd_cm.c:1014(cm_prepare_connection)
> cm_prepare_connection: connecting to DC WG101SC0002.BITIntra.de for
> domain BITINTRA [2019/03/12 09:21:04.540067,  5, pid=15439,
> effective(0, 0), real(0,
> 0)] ../lib/tdb_wrap/tdb_wrap.c:64(tdb_wrap_log)
> tdb(/opt/samba/var/lock/mutex.tdb): tdb_brlock failed (fd=22) at
> offset 592 rw_type=2 flags=1 len=1 [2019/03/12 09:21:04.540189,  1,
> pid=15439, effective(0, 0), real(0,
> 0)] ../lib/tdb_wrap/tdb_wrap.c:64(tdb_wrap_log)
> tdb(/opt/samba/var/lock/mutex.tdb): tdb_lock failed on list 106
> ltype=2 (Interrupted system call) [2019/03/12 09:21:04.540219,  0,
> pid=15439, effective(0, 0), real(0,
> 0)] ../source3/lib/util_tdb.c:497(tdb_chainlock_with_timeout_internal)
> tdb_chainlock_with_timeout_internal: alarm (40) timed out for key
> WG101SC0002.BITIntra.de in tdb /opt/samba/var/lock/mutex.tdb
> [2019/03/12 09:21:04.540384,  1, pid=15439, effective(0, 0), real(0,
> 0)] ../source3/lib/server_mutex.c:97(grab_named_mutex) Could not get
> the lock for WG101SC0002.BITIntra.de [2019/03/12 09:21:04.540508,  0,
> pid=15439, effective(0, 0), real(0, 0),
> class=winbind] ../source3/winbindd/winbindd_cm.c:1023(cm_prepare_connection)
> cm_prepare_connection: mutex grab failed for WG101SC0002.BITIntra.de
> [2019/03/12 09:21:04.540667,  1, pid=15439, effective(0, 0), real(0,
> 0),
> class=winbind] ../source3/winbindd/winbindd_cm.c:1320(cm_prepare_connection)
> Failed to prepare SMB connection to WG101SC0002.BITIntra.de:
> NT_STATUS_POSSIBLE_DEADLOCK
> 
> my understanding is that it was hanging locking an offset in the file 
> /opt/samba/var/lock/mutex.tdb, so when the timeout elapsed the
> process was interrupted (I guess the offset was that of the mutex for 
> WG101SC0002.BITIntra.de)
> Could it be a corrupted mutex.tdb file? A slow responding DC?
> Any other suggestion?

Can you please post your smb.conf.
What OS ?
What is your AD DC ?

Rowland