problem with LSI SAS controller, spontaneous resets since Upgrade to 9.2

[hlinden]

Hi Forum,

It seems that our storages have the following problem since the recent upgrade to FreeNAS 9.2. They are all Intel SR2500ALLX(R) with a built-in LSI-controller.

mpt0@pci0:4:8:0: class=0x010000 card=0x35048086 chip=0x00541000 rev=0x02 hdr=0x00
vendor = 'LSI Logic / Symbios Logic'
device = 'SAS1068 PCI-X Fusion-MPT SAS'
class = mass storage
subclass = SCSI

We have removed the memory and the RAID-license-dongle and therefore are running the controller in IT-mode which should the best way for using ZFS. This has been working very nicely until the recent update from 8.3 to 9.2 - now the controllers get reset regularly which looks like this in dmesg:

mpt0: request 0xffffff80008be7d8:167 timed out for ccb 0xfffffe000e298800 (req->ccb 0xfffffe000e298800)
mpt0: attempting to abort req 0xffffff80008be7d8:167 function 0
mpt0: mpt_wait_req(1) timed out
mpt0: mpt_recover_commands: abort timed-out. Resetting controller
mpt0: mpt_cam_event: 0x80
mpt0: mpt_cam_event: 0x80
mpt0: completing timedout/aborted req 0xffffff80008be7d8:167
Lost cambio swi (2)!!!
Lost cambio swi (3)!!!
Lost cambio swi (4)!!!
Lost cambio swi (5)!!!
(probe0:mpt0:0:8:1): INQUIRY. CDB: 12 00 00 00 24 00
(probe0:mpt0:0:8:1): CAM status: Unexpected Bus Free
(probe0:mpt0:0:8:1): Retrying command
(probe0:mpt0:0:8:1): INQUIRY. CDB: 12 00 00 00 24 00
(probe0:mpt0:0:8:1): CAM status: Unexpected Bus Free
(probe0:mpt0:0:8:1): Retrying command
(probe0:mpt0:0:8:1): INQUIRY. CDB: 12 00 00 00 24 00
(probe0:mpt0:0:8:1): CAM status: Unexpected Bus Free
(probe0:mpt0:0:8:1): Retrying command
(probe0:mpt0:0:8:1): INQUIRY. CDB: 12 00 00 00 24 00
(probe0:mpt0:0:8:1): CAM status: Unexpected Bus Free
(probe0:mpt0:0:8:1): Retrying command
(probe0:mpt0:0:8:1): INQUIRY. CDB: 12 00 00 00 24 00
(probe0:mpt0:0:8:1): CAM status: Unexpected Bus Free
(probe0:mpt0:0:8:1): Error 5, Retries exhausted
(probe0:mpt0:0:8:2): INQUIRY. CDB: 12 00 00 00 24 00
(probe0:mpt0:0:8:2): CAM status: Unexpected Bus Free
(probe0:mpt0:0:8:2): Retrying command
[...repeat for all connected disks]
Does anybody have any idea what might be causing this behaviour and how we might start to debug this?

 

[dlavigne]

Please create a bug report at bugs.freenas.org and post the issue number here.

 

[hlinden]

https://bugs.freenas.org/issues/3882

 

[Hexland]

Confirmed here too...

Dell Poweredge 2950, 32Gb ECC RAM, IBM M1015 Controller (IT Mode)

Working fine under 9.1, but under 9.2 -- FreeNAS boots OK, but as soon as I try and write to the ZPOOL connected to the M1015, the system becomes unresponsive for 20 seconds or so, then spontaneously reboots.

I haven't had a chance to check the logs yet.