Dear list readers,
I have trouble with an Mylex raid controller, which has two
BT-952 on it. Whenever a scsi device misbehaves, the complete
controller seem to hang and I have to reset the machine to
get them alive back. kernel's testet go's from 2.0.32 to 2.0.38,
seems to make no difference. The Buslogic driver is alwas 2.0.15,
see attached startup output.
I'm also attaching one of the failure reports - as far as I
can find them after a crash in syslog.
Maybe the main problem is related to the shared interrupt of
the controller's. (By accident I misconfigured the system
in the last days, so that the also used adaptec scsi controller
also shared the interrupt with the Buslogic's - and after
a scsi problem on the adaptec scsibus (bad cdrom) the complete
system hangs)
The driver has some startup options for error handling, maybe
there is one magic setting, so that the system doesn't completely
hang after a scsi reset - please let me know.
Trying many different option's is difficult, as the machine is
heavily used as server.
Upgrading to the 2.2.x series should be avoided at the moment, until
sombody can tell me, that this would solve this problem (also the
new kernel should be stable enough, with the 2.0.x series I have
had problems until 2.0.27 or so (monthly crashes!).)
And a last question: Where is an archive of this list available
on the web (on my favorite linuxhq it is gone) ?
Thanks for any help/suggestions etc. I can send more syslog output if
needed, please tell me what is needed.
Thanks for any reply (please also via direct email, as I'm not on the list).
Greetings
Hermann
--
Bildverarbeitungsgruppe des Interdiziplinaeren Zentrums fuer
wissenschaftliches Rechnen, Universitaet Heidelberg
INF 368; 69120 Heidelberg; Tel: (06221)54-8826, -6314 Fax: -8850
Email: [EMAIL PROTECTED]
Sep 1 13:21:08 klimt kernel: scsi: ***** BusLogic SCSI Driver Version 2.0.15 of 17
August 1998 *****
Sep 1 13:21:08 klimt kernel: scsi: Copyright 1995-1998 by Leonard N. Zubkoff
<[EMAIL PROTECTED]>
Sep 1 13:21:08 klimt kernel: scsi0: Configuring BusLogic Model BT-952 PCI Wide Ultra
SCSI Host Adapter
Sep 1 13:21:08 klimt kernel: scsi0: Firmware Version: 5.02, I/O Address: 0xC800,
IRQ Channel: 10/Level
Sep 1 13:21:08 klimt kernel: scsi0: PCI Bus: 1, Device: 4, Address: 0xF69FE000,
Host Adapter SCSI ID: 7
Sep 1 13:21:08 klimt kernel: scsi0: Parity Checking: Enabled, Extended Translation:
Enabled
Sep 1 13:21:08 klimt kernel: scsi0: Synchronous Negotiation: UUUUUFF#FUUFFFFF, Wide
Negotiation: Enabled
Sep 1 13:21:08 klimt kernel: scsi0: Disconnect/Reconnect: Enabled, Tagged Queuing:
Enabled
Sep 1 13:21:08 klimt kernel: scsi0: Driver Queue Depth: 255, Scatter/Gather Limit:
128 segments
Sep 1 13:21:08 klimt kernel: scsi0: Tagged Queue Depth: Automatic, Untagged Queue
Depth: 3
Sep 1 13:21:08 klimt kernel: scsi0: Error Recovery Strategy: Default, SCSI Bus
Reset: Enabled
Sep 1 13:21:08 klimt kernel: scsi0: SCSI Bus Termination: Both Enabled, SCAM:
Disabled
Sep 1 13:21:08 klimt kernel: scsi0: *** BusLogic BT-952 Initialized Successfully ***
Sep 1 13:21:08 klimt kernel: scsi1: Configuring BusLogic Model BT-952 PCI Wide Ultra
SCSI Host Adapter
Sep 1 13:21:08 klimt kernel: scsi1: Firmware Version: 5.02, I/O Address: 0xCC00,
IRQ Channel: 10/Level
Sep 1 13:21:08 klimt kernel: scsi1: PCI Bus: 1, Device: 8, Address: 0xF69FF000,
Host Adapter SCSI ID: 7
Sep 1 13:21:08 klimt kernel: scsi1: Parity Checking: Enabled, Extended Translation:
Enabled
Sep 1 13:21:08 klimt kernel: scsi1: Synchronous Negotiation: FFUFFFF#FUUFFFFF, Wide
Negotiation: Enabled
Sep 1 13:21:08 klimt kernel: scsi1: Disconnect/Reconnect: Enabled, Tagged Queuing:
Enabled
Sep 1 13:21:08 klimt kernel: scsi1: Driver Queue Depth: 255, Scatter/Gather Limit:
128 segments
Sep 1 13:21:08 klimt kernel: scsi1: Tagged Queue Depth: Automatic, Untagged Queue
Depth: 3
Sep 1 13:21:08 klimt kernel: scsi1: Error Recovery Strategy: Default, SCSI Bus
Reset: Enabled
Sep 1 13:21:08 klimt kernel: scsi1: SCSI Bus Termination: Both Enabled, SCAM:
Disabled
Sep 1 13:21:08 klimt kernel: scsi1: *** BusLogic BT-952 Initialized Successfully ***
Sep 1 13:21:08 klimt kernel: (scsi2) <Adaptec AHA-294X Ultra SCSI host adapter> found
at PCI 14/0
Sep 1 13:21:08 klimt kernel: (scsi2) Wide Channel, SCSI ID=7, 16/255 SCBs
Sep 1 13:21:08 klimt kernel: (scsi2) Warning - detected auto-termination
Sep 1 13:21:08 klimt kernel: (scsi2) Please verify driver detected settings are
correct.
Sep 1 13:21:08 klimt kernel: (scsi2) If not, then please properly set the device
termination
Sep 1 13:21:08 klimt kernel: (scsi2) in the Adaptec SCSI BIOS by hitting CTRL-A when
prompted
Sep 1 13:21:08 klimt kernel: (scsi2) during machine bootup.
Sep 1 13:21:08 klimt kernel: (scsi2) Cables present (Int-50 YES, Int-68 NO, Ext-68 NO)
Sep 1 13:21:08 klimt kernel: (scsi2) Downloading sequencer code... 419 instructions
downloaded
excerpt from crash:
Sep 1 12:32:13 klimt kernel: scsi1 channel 0 : resetting for second half of retries.
Sep 1 12:32:13 klimt kernel: SCSI bus is being reset for host 1 channel 0.
Sep 1 12:32:13 klimt kernel: scsi1: Sending Bus Device Reset CCB #399902 to Target 2
Sep 1 12:32:15 klimt kernel: scsi1: Bus Device Reset CCB #399902 to Target 2 Completed
Sep 1 12:32:15 klimt kernel: SCSI disk error : host 1 channel 0 id 2 lun 0 return
code = 18000002
Sep 1 12:32:15 klimt kernel: extra data not valid Current error sd08:30: sense key
Aborted Command
Sep 1 12:32:15 klimt kernel: Additional sense indicates Scsi parity error
Sep 1 12:32:15 klimt kernel: scsidisk I/O error: dev 08:30, sector 2, absolute sector
2
Sep 1 12:32:18 klimt kernel: SCSI disk error : host 1 channel 0 id 2 lun 0 return
code = 18000002
Sep 1 12:32:18 klimt kernel: extra data not valid Current error sd08:30: sense key
Aborted Command
Sep 1 12:32:18 klimt kernel: Additional sense indicates Scsi parity error
Sep 1 12:32:18 klimt kernel: scsidisk I/O error: dev 08:30, sector 8585308, absolute
sector 8585308
Sep 1 12:32:18 klimt kernel: SCSI disk error : host 1 channel 0 id 2 lun 0 return
code = 18000002
Sep 1 12:32:18 klimt kernel: extra data not valid Current error sd08:30: sense key
Aborted Command
Sep 1 12:32:18 klimt kernel: Additional sense indicates Scsi parity error
Sep 1 12:32:18 klimt kernel: scsidisk I/O error: dev 08:30, sector 8590526, absolute
sector 8590526
Sep 1 12:32:18 klimt kernel: SCSI disk error : host 1 channel 0 id 2 lun 0 return
code = 18000002
<truncated>
Sep 1 12:32:35 klimt kernel: scsi : aborting command due to timeout : pid 420844,
scsi1, channel 0, id 2, lun 0 Write (6) 00 00 24 02 00
Sep 1 12:32:35 klimt kernel: scsi1: Aborting CCB #399906 to Target 2
Sep 1 12:32:35 klimt kernel: scsi : aborting command due to timeout : pid 420845,
scsi1, channel 0, id 2, lun 0 Write (6) 02 00 52 02 00
Sep 1 12:32:35 klimt kernel: scsi1: Aborting CCB #399908 to Target 2
Sep 1 12:32:35 klimt kernel: scsi : aborting command due to timeout : pid 420846,
scsi1, channel 0, id 2, lun 0 Write (10) 00 00 35 c0 50 00 00 02 00
S
<truncated>
short excerpt from an older crash:
Jun 16 19:14:07 klimt kernel: scsi1: Unable to Abort Command to Target 0 - CCB Reset
Jun 16 19:14:09 klimt kernel: SCSI host 1 channel 0 reset (pid 3376033) timed out -
trying harder
Jun 16 19:14:09 klimt kernel: SCSI bus is being reset for host 1 channel 0.
Jun 16 19:14:09 klimt kernel: scsi1: Resetting BusLogic BT-952 due to Target 0
Jun 16 19:14:09 klimt kernel: scsi1: *** BusLogic BT-952 Initialized Successfully ***
Jun 16 19:14:31 klimt kernel: scsi : aborting command due to timeout : pid 3376039,
scsi1, channel 0, id 2, lun 0 Read (6) 10 76 f8 22 00
<truncated>
