topic Re: how to test a controller device in Disk Enclosures

how to test a controller device

Byron Fowlkes — Tue, 03 Jun 2008 19:50:30 GMT

I recently tried to mount one of the attached SCSI drives on our DS25 cluster. Most of our storage resides on EMC DMX san devices. The mount failed and I noticed a high error count on a device ( Adaptec AIC-7899 ). How can I test this device so as to determine if it within operating limits?

Re: how to test a controller device

Rob Leadbeater — Tue, 03 Jun 2008 20:25:40 GMT

Hi,

Giving us a clue what operating system you're running might help...!

Cheers,

Rob

Re: how to test a controller device

Byron Fowlkes — Tue, 03 Jun 2008 20:38:12 GMT

my apoogies... we are using OpenVMS 7.3-2

Re: how to test a controller device

Khairy — Wed, 04 Jun 2008 01:31:47 GMT

hi,

can you share with us more info about the disk.

Is the disk from an alphaserver too?

Are you attaching the disk into the internal disk backplanes of DS25?

Try mount the disk by overiding its ID. Assuming its dka300.
$ mcr sysman io autoconf/log
$ mount/ov=id dka300

Look for errors
$ show error
$ show dev/full dka300

If this adapter AIC adapter is attached to the backplanes, how about the other disks? Are there any errors too?

If no errors for the rest of the disks,try to swap the disk to different slot. This is just to isolate slots problem.

If it still not working, then maybe we can conclude something else about the disk. Post us the info. Thanks.

Rgds

Re: how to test a controller device

Byron Fowlkes — Wed, 04 Jun 2008 12:31:51 GMT

the disks are attached to the backplane and I believe that this device that is acting as controller is experiencing difficlty.. I am going to try to mount the disk using the SYSMAN facility as you have suggested and I will respond... thank you for this information..

Re: how to test a controller device

Khairy — Wed, 04 Jun 2008 23:45:11 GMT

hi byron,

i'm not entirely understands.

if you plan to take out the disks from EMC storage and put it into DS25 disk backplanes, i don't think its going to work. This is due to firmware difference and compatibility with scsi adapters and controllers. But you can give it a try.

Rgds

Re: how to test a controller device

Rob Leadbeater — Thu, 05 Jun 2008 05:24:05 GMT

Hi,

I'll guess that you're actually talking about the built in (6 slot) SCSI cage on the DS25, which is probably connected to an Adaptec controller on the system board...

The output of "SHOW DEV D" might be useful here.

If you can't mount the disk, please post the exact commands used, and the exact output. VMS is usually quite helpful in the responses that it gives...

Cheers,

Rob

Re: how to test a controller device

Byron Fowlkes — Thu, 05 Jun 2008 15:33:54 GMT

$1$DGA488: (ALPHA4) Mounted 0 $DISK488 42169758 7 2
$4$DKA0: (ALPHA4) Online 2
$4$DKA100: (ALPHA4) Online 2
$4$DKA200: (ALPHA4) Online 2
$4$DKA300: (ALPHA4) Online 2

PKA0: 260
$4$DKA0: (ALPHA4) 2
$4$DKA100: (ALPHA4) 2
$4$DKA200: (ALPHA4) 2
$4$DKA300: (ALPHA4) 2

first segment of display reflects "show device" output to include one volume with in the SAN (dga488)
and the next 3 scsi devices in the cage on this node $4$dka0, dka100, dka200 and dka300

next segment of display reflects "show error" output to include high error count on adaptec device PKA0:
and the 3 caged scsi devices

when trying to mount one of the scsi devices ($4$dka200:) the interactive command prompt session enters a sustained unresponsive state and must be terminated by revoking the telnet connecion to the server.. the exact command being used :: mount $4$dka200: alpha2 (alpha2 being the volume name is not ordinarily mounted)

after disengaging stalled session and using a responsive connection I measured the error count and found that it had increased for PKA0: