Parity check hangs "kernel BUG"

[fatsindey]

I initiated a parity check before I was planning to add another drive to the unRAID server, but it never finished the parity check and noticed a strange error in the syslog file.  After this the media is not available but I can access the flash drive.  I have search the forum for this but I have only found one post with it but that post was never answered.  Everything was running fine before.  I restarted the machine and got to 64% done with the parity check with only 2 hours to go, then it spiked to 15 days to finished, then I noticed the kernel bug again. 

 

I am on unRAID v4.5.6

 

I am attaching the complete syslog and below is the specific issue I am seeing:

 

Feb 13 20:36:20 unRAID kernel: ------------[ cut here ]------------

Feb 13 20:36:20 unRAID kernel: kernel BUG at drivers/md/unraid.c:780!

Feb 13 20:36:20 unRAID kernel: invalid opcode: 0000 [#1] SMP

Feb 13 20:36:20 unRAID kernel: last sysfs file: /sys/devices/pci0000:00/0000:00:1f.2/host5/target5:0:0/5:0:0:0/block/sdc/stat

Feb 13 20:36:20 unRAID kernel: Modules linked in: md_mod xor i2c_i801 i2c_core ahci pata_marvell e1000e

Feb 13 20:36:20 unRAID kernel:

Feb 13 20:36:20 unRAID kernel: Pid: 2007, comm: unraidd Not tainted (2.6.32.9-unRAID #5)       

Feb 13 20:36:20 unRAID kernel: EIP: 0060:[<f832aaa3>] EFLAGS: 00010246 CPU: 0

Feb 13 20:36:20 unRAID kernel: EIP is at handle_stripe+0xdb/0xbc9 [md_mod]

Feb 13 20:36:20 unRAID kernel: EAX: 00000009 EBX: c3f6f0b0 ECX: 00000001 EDX: c3f6f0c8

Feb 13 20:36:20 unRAID kernel: ESI: c3f6f080 EDI: c3f6f080 EBP: c32f5f9c ESP: c32f5f24

Feb 13 20:36:20 unRAID kernel:  DS: 007b ES: 007b FS: 00d8 GS: 0000 SS: 0068

Feb 13 20:36:20 unRAID kernel: Process unraidd (pid: 2007, ti=c32f4000 task=f75ea940 task.ti=c32f4000)

Feb 13 20:36:20 unRAID kernel: Stack:

Feb 13 20:36:20 unRAID kernel:  f756fe20 f75ea940 c13ce774 c1442240 f75ea940 c32f5f9c f756fde0 00000002

Feb 13 20:36:20 unRAID kernel: <0> 00000000 00000001 00000000 00000000 00000000 00000000 00000000 00000000

Feb 13 20:36:20 unRAID kernel: <0> 00000000 00000003 c3f6e720 00000001 f75ea940 c32f5f8c f832a4e1 00000000

Feb 13 20:36:20 unRAID kernel: Call Trace:

Feb 13 20:36:20 unRAID kernel:  [<f832a4e1>] ? _release_stripe+0xb8/0xdc [md_mod]

Feb 13 20:36:20 unRAID kernel:  [<f832ba72>] ? unraidd+0x8f/0xb0 [md_mod]

Feb 13 20:36:20 unRAID kernel:  [<f832b9e3>] ? unraidd+0x0/0xb0 [md_mod]

Feb 13 20:36:20 unRAID kernel:  [<c1033869>] ? kthread+0x61/0x68

Feb 13 20:36:20 unRAID kernel:  [<c1033808>] ? kthread+0x0/0x68

Feb 13 20:36:20 unRAID kernel:  [<c100339f>] ? kernel_thread_helper+0x7/0x1a

Feb 13 20:36:20 unRAID kernel: Code: 75 a8 68 c5 d0 32 f8 e8 27 a5 cf c8 83 c4 18 f0 0f ba 73 04 0d 19 c0 85 c0 0f 84 a9 00 00 00 f6 43 05 02 8d 53 18 8b 43 28 75 04 <0f> 0b eb fe 89 c6 83 e6 01 83 7a 14 00 75 47 8b 45 88 e8 78 5a

Feb 13 20:36:20 unRAID kernel: EIP: [<f832aaa3>] handle_stripe+0xdb/0xbc9 [md_mod] SS:ESP 0068:c32f5f24

Feb 13 20:36:20 unRAID kernel: ---[ end trace d36e40d9b28aada0 ]---

syslog-unRAID.txt

[fatsindey]

Ok, well since the problem persists, but my kids are still nagging me, I have decided to keep tinkering.  I have found that if just after I reboot it will start a sync which will fail at some point.  So this time I was able to cancel sync via web gui and now I have media again which is good.  I will be able to watch a chick flick with the wifee tonight after all.  Dont know about tomorrow though.

 

The bad is I had to reboot a couple times as the first one had a kernel error initiating the mdrecovery script.  I was not able to get the syslog as everything halted during boot and luckily I was watching from the console. 

 

The last or top line of the traceback was unraid_sync ....

 

I hope this sheds some light as I am very nervous to do any maintenance on it now.

 

Thanks and Happy Valentines Day.

[BRiT]

There was a kernel oops SMP bug fix in 4.6.