Google Groups no longer supports new Usenet posts or subscriptions. Historical content remains viewable.
Dismiss

an mps probelm

5 views
Skip to first unread message

Gregory Wright

unread,
Aug 3, 2011, 11:48:45 AM8/3/11
to

Hi,

I have an LSI SAS 9200-8e hba card. It's connected to a Dell
PowerVault 124T autoloader with an LTO4 drive.

I installed 9.0-BETA1 yesterday and was able to successfully
detect the tape drive (mps driver), but the hba doesn't find the changer.
I used camcontrol to force rescanning all luns but that
didn't help either.

Since I have just acquired this hardware I installed an OS
reported to work with this autoloader/hba combination,
ubuntu 11.04 server. On ubuntu, the changer is found:

gwright@greenhouse-george:~$ lsscsi -g
[0:0:0:0] disk AMCC 9650SE-12M DISK 3.06 /dev/sdb /dev/sg1
[0:0:0:1] disk AMCC 9650SE-12M DISK 3.06 /dev/sdc /dev/sg2
[0:0:1:0] disk AMCC 9650SE-12M DISK 3.06 /dev/sdd /dev/sg3
[1:0:0:0] tape IBM ULTRIUM-TD4 97F0 /dev/st0 /dev/sg5
[1:0:0:1] mediumx DELL PV-124T 0080 /dev/sch0 /dev/sg6
[2:0:0:0] disk ATA ST3160811AS 3.AA /dev/sda /dev/sg0
[6:0:0:0] cd/dvd MITSUMI CD-ROM SR244W T01A /dev/sr0 /dev/sg4

(On ubuntu the driver is mpt2sas.)

It seems as if the mps driver is not seeing the additional lun
hanging off the bus.

This server is not in production yet so I can re-install 9.0-BETA1
and run more tests, if that would help.

FWIW, I set hw.mps.0.debug_level=1. In the dmesg I get

greenhouse-george# dmesg | grep mps
mps0: <LSI SAS2008> port 0xa000-0xa0ff mem
0xff4f0000-0xff4fffff,0xff480000-0xff4bffff irq 28 at device 0.0 on pci5
mps0: Doorbell= 0x22000000
mps0: Doorbell= 0x12000000
mps0: IOCFacts :
mps0: Firmware: 02.15.63.00
mps0: IOCCapabilities:
185c<ScsiTaskFull,DiagTrace,SnapBuf,EEDP,TransRetry,IR>
mps0: Doorbell= 0x10000000
mps0: num_reqs 1024, num_replies 1055
mps0: Doorbell= 0x10000000
mps0: IOCInit status= 0x0
mps0: writing postindex 1
mps0: PortFacts :
mps0: writing postindex 2
mps0: writing postindex 3
mps0: mpssas_discovery_end: removing confighook
mps0: writing postindex 8
mps0: SAS Device Page 0 :
mps0: writing postindex 9
mps0: writing postindex 10
mps0: SAS PHY Page 0 :
mps0: Triggering rescan of 6:4:-1
mps0: writing postindex 11
mps0: writing postindex 12
mps0: writing postindex 13
mps0: (6:4:0) IOCStatus= 0x45, ScsiStatus= 0x2, SCSIState= 0x1
TransferCount= 0x0
mps0: writing postindex 14
mps0: (6:4:0) IOCStatus= 0x45, ScsiStatus= 0x0, SCSIState= 0x0
TransferCount= 0x11
mps0: writing postindex 15
mps0: (6:4:0) IOCStatus= 0x45, ScsiStatus= 0x0, SCSIState= 0x0
TransferCount= 0x56
mps0: writing postindex 16
mps0: (6:4:0) IOCStatus= 0x45, ScsiStatus= 0x0, SCSIState= 0x0
TransferCount= 0xe
mps0: writing postindex 17
mps0: (6:4:0) IOCStatus= 0x0, ScsiStatus= 0x2, SCSIState= 0x1
TransferCount= 0x0
mps0: writing postindex 18
(probe4:mps0:0:4:0): TEST UNIT READY. CDB: 0 0 0 0 0 0
(probe4:mps0:0:4:0): CAM status: SCSI Status Error
(probe4:mps0:0:4:0): SCSI status: Check Condition
(probe4:mps0:0:4:0): SCSI sense: NOT READY asc:3a,0 (Medium not present)
field rp


Please let me know what I can do to help debug this.

Best Wishes,
Greg

--
Gregory Wright
Antiope Associates LLC
18 Clay Street
Fair Haven, New Jersey 07733
USA

+1 732 924-4549 [office]
+1 732 345-8378 [fax]
gwr...@antiope.com

_______________________________________________
freebs...@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-scsi
To unsubscribe, send any mail to "freebsd-scsi...@freebsd.org"

--
Posted automagically by a mail2news gateway at muc.de e.V.
Please direct questions, flames, donations, etc. to news-...@muc.de

Kenneth D. Merry

unread,
Aug 3, 2011, 1:02:10 PM8/3/11
to
On Wed, Aug 03, 2011 at 11:48:45 -0400, Gregory Wright wrote:
>
> Hi,
>
> I have an LSI SAS 9200-8e hba card. It's connected to a Dell
> PowerVault 124T autoloader with an LTO4 drive.
>
> I installed 9.0-BETA1 yesterday and was able to successfully
> detect the tape drive (mps driver), but the hba doesn't find the changer.
> I used camcontrol to force rescanning all luns but that
> didn't help either.

Try rescanning the particular LUN for the changer device, like this:

camcontrol rescan 6:4:1

(That is assuming the tape drive is at 6:4:0.)

> Since I have just acquired this hardware I installed an OS
> reported to work with this autoloader/hba combination,
> ubuntu 11.04 server. On ubuntu, the changer is found:
>
> gwright@greenhouse-george:~$ lsscsi -g
> [0:0:0:0] disk AMCC 9650SE-12M DISK 3.06 /dev/sdb /dev/sg1
> [0:0:0:1] disk AMCC 9650SE-12M DISK 3.06 /dev/sdc /dev/sg2
> [0:0:1:0] disk AMCC 9650SE-12M DISK 3.06 /dev/sdd /dev/sg3
> [1:0:0:0] tape IBM ULTRIUM-TD4 97F0 /dev/st0 /dev/sg5
> [1:0:0:1] mediumx DELL PV-124T 0080 /dev/sch0 /dev/sg6
> [2:0:0:0] disk ATA ST3160811AS 3.AA /dev/sda /dev/sg0
> [6:0:0:0] cd/dvd MITSUMI CD-ROM SR244W T01A /dev/sr0 /dev/sg4
>
> (On ubuntu the driver is mpt2sas.)
>
> It seems as if the mps driver is not seeing the additional lun
> hanging off the bus.

Yeah, it looks like the path inquiry CCB in the driver is telling CAM not
to scan LUNs.

> This server is not in production yet so I can re-install 9.0-BETA1
> and run more tests, if that would help.

Try the attached patch, and let me know whether it works.

Ken
--
Kenneth Merry
k...@FreeBSD.ORG

mps_sas.c.max_lun.20110803

Gregory Wright

unread,
Aug 3, 2011, 2:46:34 PM8/3/11
to
Hi Ken,

On 8/3/11 1:02 PM, Kenneth D. Merry wrote:
> On Wed, Aug 03, 2011 at 11:48:45 -0400, Gregory Wright wrote:

>> Hi,
>>
>> I have an LSI SAS 9200-8e hba card. It's connected to a Dell
>> PowerVault 124T autoloader with an LTO4 drive.
>>
>> I installed 9.0-BETA1 yesterday and was able to successfully
>> detect the tape drive (mps driver), but the hba doesn't find the changer.
>> I used camcontrol to force rescanning all luns but that
>> didn't help either.

> Try rescanning the particular LUN for the changer device, like this:
>
> camcontrol rescan 6:4:1

The changer was found when I used that command.


> (That is assuming the tape drive is at 6:4:0.)
>

>> Since I have just acquired this hardware I installed an OS
>> reported to work with this autoloader/hba combination,
>> ubuntu 11.04 server. On ubuntu, the changer is found:
>>
>> gwright@greenhouse-george:~$ lsscsi -g
>> [0:0:0:0] disk AMCC 9650SE-12M DISK 3.06 /dev/sdb /dev/sg1
>> [0:0:0:1] disk AMCC 9650SE-12M DISK 3.06 /dev/sdc /dev/sg2
>> [0:0:1:0] disk AMCC 9650SE-12M DISK 3.06 /dev/sdd /dev/sg3
>> [1:0:0:0] tape IBM ULTRIUM-TD4 97F0 /dev/st0 /dev/sg5
>> [1:0:0:1] mediumx DELL PV-124T 0080 /dev/sch0 /dev/sg6
>> [2:0:0:0] disk ATA ST3160811AS 3.AA /dev/sda /dev/sg0
>> [6:0:0:0] cd/dvd MITSUMI CD-ROM SR244W T01A /dev/sr0 /dev/sg4
>>
>> (On ubuntu the driver is mpt2sas.)
>>
>> It seems as if the mps driver is not seeing the additional lun
>> hanging off the bus.

> Yeah, it looks like the path inquiry CCB in the driver is telling CAM not
> to scan LUNs.
>

>> This server is not in production yet so I can re-install 9.0-BETA1
>> and run more tests, if that would help.

> Try the attached patch, and let me know whether it works.
>
>

The patch works. Now I get

greenhouse-george# camcontrol devlist
<ST3160811AS 3.AAE> at scbus0 target 0 lun 0 (ada0,pass0)
<MITSUMI CD-ROM SR244W T01A> at scbus4 target 0 lun 0 (cd0,pass1)
<IBM ULTRIUM-TD4 97F0> at scbus6 target 4 lun 0 (sa0,pass2)
<DELL PV-124T 0080> at scbus6 target 4 lun 1 (ch0,pass3)
<AMCC 9650SE-12M DISK 3.06> at scbus7 target 0 lun 0 (da0,pass4)
<AMCC 9650SE-12M DISK 3.06> at scbus7 target 0 lun 1 (da1,pass5)
<AMCC 9650SE-12M DISK 3.06> at scbus7 target 1 lun 0 (da2,pass6)
greenhouse-george#

I hope the change makes it into BETA2/RC1. Thanks!

Best,

Gregory Wright

unread,
Aug 3, 2011, 3:34:57 PM8/3/11
to
On 8/3/11 2:46 PM, Gregory Wright wrote:
> Hi Ken,
>
> On 8/3/11 1:02 PM, Kenneth D. Merry wrote:
>> On Wed, Aug 03, 2011 at 11:48:45 -0400, Gregory Wright wrote:
>>> Hi,
>>>
>>> I have an LSI SAS 9200-8e hba card. It's connected to a Dell
>>> PowerVault 124T autoloader with an LTO4 drive.
>>>
>>> I installed 9.0-BETA1 yesterday and was able to successfully
>>> detect the tape drive (mps driver), but the hba doesn't find the
>>> changer.
>>> I used camcontrol to force rescanning all luns but that
>>> didn't help either.
>> Try rescanning the particular LUN for the changer device, like this:
>>
>> camcontrol rescan 6:4:1
> The changer was found when I used that command.
>> (That is assuming the tape drive is at 6:4:0.)
>>
>>> Since I have just acquired this hardware I installed an OS
>>> reported to work with this autoloader/hba combination,
>>> ubuntu 11.04 server. On ubuntu, the changer is found:
>>>
>>> gwright@greenhouse-george:~$ lsscsi -g
>>> [0:0:0:0] disk AMCC 9650SE-12M DISK 3.06 /dev/sdb /dev/sg1
>>> [0:0:0:1] disk AMCC 9650SE-12M DISK 3.06 /dev/sdc /dev/sg2
>>> [0:0:1:0] disk AMCC 9650SE-12M DISK 3.06 /dev/sdd /dev/sg3
>>> [1:0:0:0] tape IBM ULTRIUM-TD4 97F0 /dev/st0 /dev/sg5
>>> [1:0:0:1] mediumx DELL PV-124T 0080 /dev/sch0
>>> /dev/sg6
>>> [2:0:0:0] disk ATA ST3160811AS 3.AA /dev/sda /dev/sg0
>>> [6:0:0:0] cd/dvd MITSUMI CD-ROM SR244W T01A /dev/sr0 /dev/sg4
>>>
>>> (On ubuntu the driver is mpt2sas.)
>>>
>>> It seems as if the mps driver is not seeing the additional lun
>>> hanging off the bus.
>> Yeah, it looks like the path inquiry CCB in the driver is telling CAM
>> not
>> to scan LUNs.
>>
>>> This server is not in production yet so I can re-install 9.0-BETA1
>>> and run more tests, if that would help.
>> Try the attached patch, and let me know whether it works.
>>
>>
> The patch works. Now I get
>
> greenhouse-george# camcontrol devlist
> <ST3160811AS 3.AAE> at scbus0 target 0 lun 0 (ada0,pass0)
> <MITSUMI CD-ROM SR244W T01A> at scbus4 target 0 lun 0 (cd0,pass1)
> <IBM ULTRIUM-TD4 97F0> at scbus6 target 4 lun 0 (sa0,pass2)
> <DELL PV-124T 0080> at scbus6 target 4 lun 1 (ch0,pass3)
> <AMCC 9650SE-12M DISK 3.06> at scbus7 target 0 lun 0 (da0,pass4)
> <AMCC 9650SE-12M DISK 3.06> at scbus7 target 0 lun 1 (da1,pass5)
> <AMCC 9650SE-12M DISK 3.06> at scbus7 target 1 lun 0 (da2,pass6)
> greenhouse-george#
>
> I hope the change makes it into BETA2/RC1. Thanks!
>
> Best,
> Greg
>
As a further test I installed mtx and verified that the changer
works under software control:

greenhouse-george# mtx -f /dev/pass3 status
Storage Changer /dev/pass3:1 Drives, 16 Slots ( 0 Import/Export )
Data Transfer Element 0:Empty
Storage Element 1:Full
Storage Element 2:Full
Storage Element 3:Full
Storage Element 4:Full
Storage Element 5:Empty
Storage Element 6:Empty
Storage Element 7:Empty
Storage Element 8:Empty
Storage Element 9:Empty
Storage Element 10:Empty
Storage Element 11:Empty
Storage Element 12:Empty
Storage Element 13:Empty
Storage Element 14:Empty
Storage Element 15:Empty
Storage Element 16:Empty
greenhouse-george# mtx -f /dev/pass3 load 1
Loading media from Storage Element 1 into drive 0...done
greenhouse-george# mtx -f /dev/pass3 status
Storage Changer /dev/pass3:1 Drives, 16 Slots ( 0 Import/Export )
Data Transfer Element 0:Full (Storage Element 1 Loaded)
Storage Element 1:Empty
Storage Element 2:Full
Storage Element 3:Full
Storage Element 4:Full
Storage Element 5:Empty
Storage Element 6:Empty
Storage Element 7:Empty
Storage Element 8:Empty
Storage Element 9:Empty
Storage Element 10:Empty
Storage Element 11:Empty
Storage Element 12:Empty
Storage Element 13:Empty
Storage Element 14:Empty
Storage Element 15:Empty
Storage Element 16:Empty
greenhouse-george# mtx -f /dev/pass3 unload
Unloading drive 0 into Storage Element 1...done
greenhouse-george# mtx -f /dev/pass3 status
Storage Changer /dev/pass3:1 Drives, 16 Slots ( 0 Import/Export )
Data Transfer Element 0:Empty
Storage Element 1:Full
Storage Element 2:Full
Storage Element 3:Full
Storage Element 4:Full
Storage Element 5:Empty
Storage Element 6:Empty
Storage Element 7:Empty
Storage Element 8:Empty
Storage Element 9:Empty
Storage Element 10:Empty
Storage Element 11:Empty
Storage Element 12:Empty
Storage Element 13:Empty
Storage Element 14:Empty
Storage Element 15:Empty
Storage Element 16:Empty
greenhouse-george#

Looks like this issue is resolved.

Cheers,

Message has been deleted
0 new messages