Trimble, Ronald D
2008-Oct-10 16:53 UTC
[Samba] Server crash - Is it a Kernel or Samba problem?
Yesterday I had an unexpected server crash. Here is what appeared in the logs: Oct 9 20:16:21 USTR-LINUX-1 [powersaved][11654]: resmgr: server response code 200 Oct 9 20:16:53 USTR-LINUX-1 last message repeated 19 times Oct 9 20:17:26 USTR-LINUX-1 last message repeated 13 times Oct 9 20:17:26 USTR-LINUX-1 kernel: Unable to handle kernel paging request at virtual address 00100104 Oct 9 20:17:26 USTR-LINUX-1 kernel: printing eip: Oct 9 20:17:26 USTR-LINUX-1 kernel: c0134d50 Oct 9 20:17:26 USTR-LINUX-1 kernel: *pde = 09044001 Oct 9 20:17:26 USTR-LINUX-1 kernel: Oops: 0002 [#1] Oct 9 20:17:26 USTR-LINUX-1 kernel: SMP Oct 9 20:17:26 USTR-LINUX-1 kernel: CPU: 2 Oct 9 20:17:26 USTR-LINUX-1 kernel: EIP: 0060:[<c0134d50>] Tainted: G U Oct 9 20:17:26 USTR-LINUX-1 kernel: EFLAGS: 00010002 (2.6.5-7.286-bigsmp SLES9_SP3_BRANCH-20070531101258) Oct 9 20:17:26 USTR-LINUX-1 kernel: EIP is at free_uid+0x20/0x50 Oct 9 20:17:26 USTR-LINUX-1 kernel: eax: 00100100 ebx: ecd84500 ecx: ecd84514 edx: 00200200 Oct 9 20:17:26 USTR-LINUX-1 kernel: esi: c9460af8 edi: 00000009 ebp: 0000000a esp: cf66beb0 Oct 9 20:17:26 USTR-LINUX-1 kernel: ds: 007b es: 007b ss: 0068 Oct 9 20:17:26 USTR-LINUX-1 kernel: Process smbd (pid: 29272, threadinfo=cf66a000 task=ec3c4010) Oct 9 20:17:26 USTR-LINUX-1 kernel: Stack: c677d708 c0135f64 00000000 cf66bf28 00000000 cf66bf28 ec3c4010 ec3c4554 Oct 9 20:17:26 USTR-LINUX-1 kernel: c0137c22 cf66a000 083d7520 cf66bfc4 ffffe000 c0137ffa 2411f3bd cf66a000 Oct 9 20:17:26 USTR-LINUX-1 kernel: ec3c4554 cf66bfc4 cf66bf28 cf66a000 083d7520 cf66bfc4 ec3c4554 c010847a Oct 9 20:17:26 USTR-LINUX-1 kernel: Call Trace: Oct 9 20:17:26 USTR-LINUX-1 kernel: [<c0135f64>] __dequeue_signal+0x184/0x1a0 Oct 9 20:17:26 USTR-LINUX-1 kernel: [<c0137c22>] dequeue_signal+0x62/0xa0 Oct 9 20:17:26 USTR-LINUX-1 kernel: [<c0137ffa>] get_signal_to_deliver+0x7a/0x3d0 Oct 9 20:17:26 USTR-LINUX-1 kernel: [<c010847a>] do_signal+0x8a/0x640 Oct 9 20:17:26 USTR-LINUX-1 kernel: [<c0140874>] ckrm_invoke_event_cb_chain+0x24/0x30 Oct 9 20:17:26 USTR-LINUX-1 kernel: [<c013ab2c>] sys_setresuid+0x1dc/0x240 Oct 9 20:17:26 USTR-LINUX-1 kernel: [<c0108a67>] do_notify_resume+0x37/0x40 Oct 9 20:17:26 USTR-LINUX-1 kernel: [<c0109256>] work_notifysig+0x13/0x15 Oct 9 20:17:26 USTR-LINUX-1 kernel: Oct 9 20:17:26 USTR-LINUX-1 kernel: Code: 89 50 04 89 02 89 da c7 43 14 00 01 10 00 c7 41 04 00 02 20 Oct 10 00:24:53 USTR-LINUX-1 syslogd 1.4.1: restart. My question is is this a kernel or a samba problem? Has anyone experience this before? I do know that the server was under considerable SMB load (a build was being generated on another computer and written to this server) when the oops occurred. I am running SUSE SLES 9 SP4. Kernel is 2.6.5-7.286-bigsmp. Any help would be appreciated. Thanks!
Volker Lendecke
2008-Oct-10 19:32 UTC
[Samba] Server crash - Is it a Kernel or Samba problem?
On Fri, Oct 10, 2008 at 11:22:58AM -0500, Trimble, Ronald D wrote:> Oct 9 20:17:26 USTR-LINUX-1 kernel: Call Trace: > Oct 9 20:17:26 USTR-LINUX-1 kernel: [<c0135f64>] __dequeue_signal+0x184/0x1a0 > Oct 9 20:17:26 USTR-LINUX-1 kernel: [<c0137c22>] dequeue_signal+0x62/0xa0 > Oct 9 20:17:26 USTR-LINUX-1 kernel: [<c0137ffa>] get_signal_to_deliver+0x7a/0x3d0 > Oct 9 20:17:26 USTR-LINUX-1 kernel: [<c010847a>] do_signal+0x8a/0x640 > Oct 9 20:17:26 USTR-LINUX-1 kernel: [<c0140874>] ckrm_invoke_event_cb_chain+0x24/0x30 > Oct 9 20:17:26 USTR-LINUX-1 kernel: [<c013ab2c>] sys_setresuid+0x1dc/0x240 > Oct 9 20:17:26 USTR-LINUX-1 kernel: [<c0108a67>] do_notify_resume+0x37/0x40 > Oct 9 20:17:26 USTR-LINUX-1 kernel: [<c0109256>] work_notifysig+0x13/0x15 > Oct 9 20:17:26 USTR-LINUX-1 kernel: > Oct 9 20:17:26 USTR-LINUX-1 kernel: Code: 89 50 04 89 02 89 da c7 43 14 00 01 10 00 c7 41 04 00 02 20 > Oct 10 00:24:53 USTR-LINUX-1 syslogd 1.4.1: restart. > > > My question is is this a kernel or a samba problem? Has > anyone experience this before? I do know that the server > was under considerable SMB load (a build was being > generated on another computer and written to this server) > when the oops occurred. I am running SUSE SLES 9 SP4. > Kernel is 2.6.5-7.286-bigsmp.Kernel crashes are a kernel problem, or maybe flaky hardware. Samba might put a load on the kernel that only few other applications do, but it is a kernel problem. Volker -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 189 bytes Desc: not available Url : http://lists.samba.org/archive/samba/attachments/20081010/62face95/attachment.bin
Trimble, Ronald D
2008-Oct-10 19:39 UTC
[Samba] Server crash - Is it a Kernel or Samba problem?
Do you have any suggestions on how I may track this down. Obviously, the logs are sparse. Has anyone else reported a similar problem? -----Original Message----- From: Volker Lendecke [mailto:Volker.Lendecke@SerNet.DE] Sent: Friday, October 10, 2008 3:19 PM To: Trimble, Ronald D Cc: samba@lists.samba.org Subject: Re: [Samba] Server crash - Is it a Kernel or Samba problem? On Fri, Oct 10, 2008 at 11:22:58AM -0500, Trimble, Ronald D wrote:> Oct 9 20:17:26 USTR-LINUX-1 kernel: Call Trace: > Oct 9 20:17:26 USTR-LINUX-1 kernel: [<c0135f64>] > __dequeue_signal+0x184/0x1a0 Oct 9 20:17:26 USTR-LINUX-1 kernel: > [<c0137c22>] dequeue_signal+0x62/0xa0 Oct 9 20:17:26 USTR-LINUX-1 > kernel: [<c0137ffa>] get_signal_to_deliver+0x7a/0x3d0 Oct 9 20:17:26 > USTR-LINUX-1 kernel: [<c010847a>] do_signal+0x8a/0x640 Oct 9 > 20:17:26 USTR-LINUX-1 kernel: [<c0140874>] > ckrm_invoke_event_cb_chain+0x24/0x30 > Oct 9 20:17:26 USTR-LINUX-1 kernel: [<c013ab2c>] > sys_setresuid+0x1dc/0x240 Oct 9 20:17:26 USTR-LINUX-1 kernel: > [<c0108a67>] do_notify_resume+0x37/0x40 Oct 9 20:17:26 USTR-LINUX-1 > kernel: [<c0109256>] work_notifysig+0x13/0x15 Oct 9 20:17:26 USTR-LINUX-1 kernel: > Oct 9 20:17:26 USTR-LINUX-1 kernel: Code: 89 50 04 89 02 89 da c7 43 > 14 00 01 10 00 c7 41 04 00 02 20 Oct 10 00:24:53 USTR-LINUX-1 syslogd 1.4.1: restart. > > > My question is is this a kernel or a samba problem? Has anyone > experience this before? I do know that the server was under > considerable SMB load (a build was being generated on another computer > and written to this server) when the oops occurred. I am running SUSE > SLES 9 SP4. > Kernel is 2.6.5-7.286-bigsmp.Kernel crashes are a kernel problem, or maybe flaky hardware. Samba might put a load on the kernel that only few other applications do, but it is a kernel problem. Volker
Volker Lendecke
2008-Oct-10 20:15 UTC
[Samba] Server crash - Is it a Kernel or Samba problem?
On Fri, Oct 10, 2008 at 02:36:25PM -0500, Trimble, Ronald D wrote:> Do you have any suggestions on how I may track this down. > Obviously, the logs are sparse. Has anyone else reported > a similar problem?The only real suggestion I have is to contact Novell. SLES9 is a supported product. Volker -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 189 bytes Desc: not available Url : http://lists.samba.org/archive/samba/attachments/20081010/ccc73def/attachment.bin