Problem with mpt and Infortrend RAID Array

[Oliver Brandmueller]

Hi,

I've got a problem with using an Infortrend RAID array (actually EonStor 
A16U G1410) on an LSI Logic 1030 Controller on FreeBSD 5-STABLE.

First things first, here's my Environment:

mpt0: <LSILogic 1030 Ultra4 Adapter> port 0xb000-0xb0ff mem 0xf6820000-0xf683ffff,0xf6800000-0xf681ffff irq 24 at device 3.0 on pci2
mpt0: [GIANT-LOCKED]

FreeBSD zombie 5.3-STABLE FreeBSD 5.3-STABLE #5: Tue Jan 25 11:35:05 CET 2005     root at hudson:/usr/obj/usr/src/sys/NFS  i386

mpt is loaded as a module.

The LSI Controller and the RAID Array all have the latest firmware 
(meanwhile).


Now the problem:

If I only map map one ID/LUN with a drive (under 2 TB) everything works 
fine.

As soon as I map more drives (let's say 2 logical volumes of 800GB) be 
it to one ID and LUN 0 and 1 or on ID 0 and 1 both on LUN 0 I get error 
messages when scanning the SCSI bus (on boot or in multiuser when doing 
a "camcontrol rescan 0" after mapping the devices on the RAID):

(the messages here are while having mapped 2 800GB drives to SCSI ID 0 
on LUN 0 and 1):

da0 at mpt0 bus 0 target 0 lun 0
da0: <IFT A16U-G1410 341B> Fixed Direct Access SCSI-3 device 
da0: 320.000MB/s transfers (160.000MHz, offset 127, 16bit), Tagged Queueing Enabled
da0: 800000MB (1638400000 512 byte sectors: 255H 63S/T 101985C)
mpt0: time out on request index = 0xf2 sequence = 0x00000093
mpt0: Status 00000001; Mask 00000001; Doorbell 24000000
request state On Chip
SCSI IO Request @ 0xe4ad3c20
        Chain Offset  0x00
        MsgFlags      0x00
        MsgContext    0x000000f2
        Bus:                0
        TargetID            0
        SenseBufferLength   18
        LUN:              0x1
        Control           0x02000500  READ  UNTAGGED 
        DataLength      0x000000fb
        SenseBufAddr    0x3e8ea5e0
        CDB[0:6]        12 00 00 00 fb 00 
        SE32 0xe4bab430: Addr=0x2f1f7e84 FlagsLength=0xd10000fb
         LAST_ELEMENT END_OF_BUFFER END_OF_LIST
mpt0: time out on request index = 0xfe sequence = 0x00000094
mpt0: Status 00000001; Mask 00000001; Doorbell 24000000
request state On Chip
SCSI IO Request @ 0xe4ad3c20
        Chain Offset  0x00
        MsgFlags      0x00
        MsgContext    0x000000fe
        Bus:                0
        TargetID            0
        SenseBufferLength   32
        LUN:              0x0
        Control           0x02000500  READ  UNTAGGED 
        DataLength      0x00000008
        SenseBufAddr    0x3e8ebde0
        CDB[0:10]       25 00 00 00 00 00 00 00 00 00 
        SE32 0xe4bacc30: Addr=0x2b7960 FlagsLength=0xd1000008
         LAST_ELEMENT END_OF_BUFFER END_OF_LIST


The last messages keep slowly scrolling with ascending sequence numbers
until I reboot. camcontrol cannot be killed in this state.

A similar thing happens when I map the drives to different IDs:

mpt0: time out on request index = 0xf4 sequence = 0x00000088
mpt0: Status 00000001; Mask 00000001; Doorbell 24000000
request state On Chip
SCSI IO Request @ 0xe4ad3c20
        Chain Offset  0x00
        MsgFlags      0x00
        MsgContext    0x000000f4
        Bus:                0
        TargetID            0
        SenseBufferLength   32
        LUN:              0x0
        Control           0x02000500  READ  UNTAGGED 
        DataLength      0x00000014
        SenseBufAddr    0x3e8ea9e0
        CDB[0:6]        1a 00 0a 00 14 00 
        SE32 0xe4bab830: Addr=0x2dad80 FlagsLength=0xd1000014
         LAST_ELEMENT END_OF_BUFFER END_OF_LIST
mpt0: time out on request index = 0xf3 sequence = 0x00000089
mpt0: Status 00000001; Mask 00000001; Doorbell 24000000
request state On Chip
SCSI IO Request @ 0xe4ad3c20
        Chain Offset  0x00
        MsgFlags      0x00
        MsgContext    0x000000f3
        Bus:                0
        TargetID            1
        SenseBufferLength   32
        LUN:              0x0
        Control           0x02000500  READ  UNTAGGED 
        DataLength      0x00000014
        SenseBufAddr    0x3e8ea7e0
        CDB[0:6]        1a 00 0a 00 14 00 
        SE32 0xe4bab630: Addr=0x2dac80 FlagsLength=0xd1000014
         LAST_ELEMENT END_OF_BUFFER END_OF_LIST

You might notice, that in the first case at least one drive appeared. 
This is not predictable: Sometimes this happens, sometimes it doesn't. 
This seems in no way connected to the one of the configurations, but 
seems to appear randomly on either config.

I did NOT yet:

  - boot the machine with Linux/Knoppix (the machine is off site,
      I've not been going there, will be by end of the week)
  - change the SCSI controller (I will get  a Adaptec U320 controller
      for testing purposes in a few weeks, so I hope to solve the
      problem earlier)

I will NOT have the chance to:

  - replace the RAID array by another one
  - use another operating system :-)

Thanx for assistance,

			Oliver


-- 
| Oliver Brandmueller | Offenbacher Str. 1  | Germany       D-14197 Berlin |
| Fon +49-172-3130856 | Fax +49-172-3145027 | WWW:   http://the.addict.de/ |
|               Ich bin das Internet. Sowahr ich Gott helfe.               |
| Eine gewerbliche Nutzung aller enthaltenen Adressen ist nicht gestattet! |
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 187 bytes
Desc: not available
Url : http://lists.freebsd.org/pipermail/freebsd-scsi/attachments/20050125/be15b6c8/attachment.bin