Frederik Vogelsang
2012-Nov-02 17:35 UTC
[Samba] Samba 4: Issues with Linux clients (CIFS timeouts)
Hi, I have set up Samba 4 to act as a standalone AD domain controller and this setup is working so far. Windows clients work like a charm, even group policy and advanced stuff, which is pretty exciting. What is still causing a lot of pain are the Linux clients. Authentication works great (pam_krb5 & nss_ldap) and the users can even mount shares with their Kerberos credentials. Now the issue is that when the users access their home folders *sometimes* directory/file lookups fail or stall, which is causing apps to crash. After two minutes files and folders on CIFS shares can be accessed again, but only for a random period of time. After a while these timeouts are coming back, this is what I get in dmesg on the Linux client: CIFS VFS: Server horst has not responded in 120 seconds. Reconnecting... CIFS VFS: Unexpected lookup error -112 It looks like these timeouts happen when there is a lot of I/O load on the CIFS shares. This is the output of log.smbd when a freeze happens: [2012/11/02 18:19:12.454597, 2] ../source3/smbd/open.c:899(open_file) TESTDOMAIN\testuser opened file files/sample.txt read=Yes write=No (numopen=1) [2012/11/02 18:19:12.455161, 3] ../source3/smbd/process.c:1789(process_smb) Transaction 173 of length 63 (0 toread) [2012/11/02 18:19:12.455206, 3] ../source3/smbd/process.c:1392(switch_message) switch message SMBreadX (pid 11302) conn 0x1598260 [2012/11/02 18:19:12.455255, 3] ../source3/smbd/reply.c:3841(send_file_readX) send_file_readX fnum 13004 max=65536 nread=65536 [2012/11/02 18:19:12.459192, 3] ../source3/smbd/process.c:1789(process_smb) Transaction 174 of length 63 (0 toread) [2012/11/02 18:19:12.459259, 3] ../source3/smbd/process.c:1392(switch_message) switch message SMBreadX (pid 11302) conn 0x1598260 [2012/11/02 18:19:12.464500, 3] ../source3/smbd/process.c:1789(process_smb) Transaction 175 of length 63 (0 toread) [2012/11/02 18:19:12.464549, 3] ../source3/smbd/process.c:1392(switch_message) switch message SMBreadX (pid 11302) conn 0x1598260 [2012/11/02 18:19:12.476760, 3] ../source3/smbd/process.c:1789(process_smb) Transaction 176 of length 63 (0 toread) [2012/11/02 18:19:12.476829, 3] ../source3/smbd/process.c:1392(switch_message) switch message SMBreadX (pid 11302) conn 0x1598260 [2012/11/02 18:19:12.501988, 3] ../source3/smbd/process.c:1789(process_smb) Transaction 177 of length 63 (0 toread) [2012/11/02 18:19:12.502039, 3] ../source3/smbd/process.c:1392(switch_message) switch message SMBreadX (pid 11302) conn 0x1598260 [2012/11/02 18:19:12.546623, 3] ../source3/smbd/process.c:1789(process_smb) Transaction 178 of length 63 (0 toread) [2012/11/02 18:19:12.546692, 3] ../source3/smbd/process.c:1392(switch_message) switch message SMBreadX (pid 11302) conn 0x1598260 [2012/11/02 18:19:12.619302, 3] ../source3/smbd/process.c:1789(process_smb) Transaction 179 of length 63 (0 toread) [2012/11/02 18:19:12.619371, 3] ../source3/smbd/process.c:1392(switch_message) switch message SMBreadX (pid 11302) conn 0x1598260 [2012/11/02 18:19:12.677597, 3] ../source3/smbd/process.c:1789(process_smb) Transaction 180 of length 63 (0 toread) [2012/11/02 18:19:12.677664, 3] ../source3/smbd/process.c:1392(switch_message) switch message SMBreadX (pid 11302) conn 0x1598260 [2012/11/02 18:19:12.719097, 3] ../source3/smbd/process.c:1789(process_smb) Transaction 181 of length 63 (0 toread) [2012/11/02 18:19:12.719166, 3] ../source3/smbd/process.c:1392(switch_message) switch message SMBreadX (pid 11302) conn 0x1598260 [2012/11/02 18:19:12.766312, 3] ../source3/smbd/process.c:1789(process_smb) Transaction 182 of length 63 (0 toread) [2012/11/02 18:19:12.766380, 3] ../source3/smbd/process.c:1392(switch_message) switch message SMBreadX (pid 11302) conn 0x1598260 [2012/11/02 18:19:12.810994, 3] ../source3/smbd/process.c:1789(process_smb) Transaction 183 of length 63 (0 toread) [2012/11/02 18:19:12.811063, 3] ../source3/smbd/process.c:1392(switch_message) switch message SMBreadX (pid 11302) conn 0x1598260 [2012/11/02 18:19:12.848722, 3] ../source3/smbd/process.c:1789(process_smb) Transaction 184 of length 63 (0 toread) [2012/11/02 18:19:12.848791, 3] ../source3/smbd/process.c:1392(switch_message) switch message SMBreadX (pid 11302) conn 0x1598260 [2012/11/02 18:19:12.886421, 3] ../source3/smbd/process.c:1789(process_smb) Transaction 185 of length 63 (0 toread) [2012/11/02 18:19:12.886488, 3] ../source3/smbd/process.c:1392(switch_message) switch message SMBreadX (pid 11302) conn 0x1598260 [2012/11/02 18:19:12.925811, 3] ../source3/smbd/process.c:1789(process_smb) Transaction 186 of length 63 (0 toread) [2012/11/02 18:19:12.925879, 3] ../source3/smbd/process.c:1392(switch_message) switch message SMBreadX (pid 11302) conn 0x1598260 [2012/11/02 18:19:12.973718, 3] ../source3/smbd/process.c:1789(process_smb) Transaction 187 of length 63 (0 toread) [2012/11/02 18:19:12.973787, 3] ../source3/smbd/process.c:1392(switch_message) switch message SMBreadX (pid 11302) conn 0x1598260 [2012/11/02 18:19:13.040761, 3] ../source3/smbd/process.c:1789(process_smb) Transaction 188 of length 63 (0 toread) [2012/11/02 18:19:13.040830, 3] ../source3/smbd/process.c:1392(switch_message) switch message SMBreadX (pid 11302) conn 0x1598260 [2012/11/02 18:19:13.041172, 3] ../source3/smbd/reply.c:3841(send_file_readX) send_file_readX fnum 13004 max=475136 nread=472741 [2012/11/02 18:19:28.283505, 2] ../source3/smbd/server.c:436(remove_child_pid) Could not find child 11310 -- ignoring [2012/11/02 18:19:28.473748, 3] ../source3/smbd/process.c:1789(process_smb) Transaction 189 of length 78 (0 toread) [2012/11/02 18:19:28.473821, 3] ../source3/smbd/process.c:1392(switch_message) switch message SMBtrans2 (pid 11302) conn 0x1598260 [2012/11/02 18:19:28.473860, 3] ../source3/smbd/trans2.c:5217(call_trans2qfilepathinfo) call_trans2qfilepathinfo: TRANSACT2_QPATHINFO: level = 512 [2012/11/02 18:19:28.473902, 3] ../source3/smbd/vfs.c:1118(check_reduced_name) check_reduced_name [.] [/mnt/test] [2012/11/02 18:19:28.473938, 3] ../source3/smbd/vfs.c:1248(check_reduced_name) check_reduced_name: . reduced to /mnt/test [2012/11/02 18:19:28.473981, 3] ../source3/locking/share_mode_lock.c:408(fetch_share_mode_unlocked) Could not fetch share entry [2012/11/02 18:19:28.474011, 3] ../source3/smbd/trans2.c:5362(call_trans2qfilepathinfo) call_trans2qfilepathinfo . (fnum [fsp is NULL]) level=512 call=5 total_data=0 [2012/11/02 18:19:37.799385, 3] ../source3/smbd/process.c:1789(process_smb) Transaction 36 of length 42 (0 toread) [2012/11/02 18:19:37.799466, 3] ../source3/smbd/process.c:1392(switch_message) switch message SMBecho (pid 29523) conn 0x0 [2012/11/02 18:19:37.799527, 3] ../source3/smbd/reply.c:5383(reply_echo) echo 1 times I have tried many different kernels (currently on 3.7-rc3) and the problem persists. I believe that this is Samba 4 issue, because with Samba 3 I did not encounter any of these timeouts. Are there any thoughts what could be going wrong here? Are there any options I should set somewhere? Regards, Frederik