Google Groups no longer supports new Usenet posts or subscriptions. Historical content remains viewable.
Dismiss
MAN3367MC questions
20 views
Skip to first unread message
CIV
Oct 25, 2012, 6:58:53 AM10/25/12
to saw...@flashmail.com
Good morning:
I'd be very obliged if I could ask assistance from the more knowleadgeable members of the board.
I have tree (apparently identical) Fujutsu MAN3367MC SCSI drives.
They are unused spares from out of service servers.
Drives 1 and 2.
Labeled MAN3367MC ULTRA160 SCSI SCA-2 / LVD ID: JW
Date of manufacture: 03-2002
Bart's SCSITools IDs them as MAN3367M SUN36G Revision: 1502
Drive 3.
Labeled MAN3367MC ULTRA160 SCSI SCA-2 / LVD ID: JW
Date of manufacture: 10-2002
Bart's SCSITools IDs it as IBM-ESXS MAN3367MC F Revision: 5408
From what I see in the labels and the PCBs, all three drives would seem to be the 'same', but the one with the SUN36G ID will only work at 80mb/s and not at 160mb/s as the one with the IBM-ESXS ID. (Adaptec 29160 card)
The info I have found on the web makes no distinction between the MAN3367M and the MAN3367MC models, apparently they are both U160 LVD.
It would seem to me that it could be a firmware issue ie: the SUN36G would need the IBM-ESXS firmware to be able to work at 160mb/s.
I have attempted to find a firmware upgrade to no avail nor have I been able to find a utility to copy the IBM-ESXS to a file.
I'd appreciate any pointers towards being able to get the IBM-ESXS firmware on the SUN36G drive.
Thanks in advance.
CIV
I'd be very obliged if I could ask assistance from the more knowleadgeable members of the board.
I have tree (apparently identical) Fujutsu MAN3367MC SCSI drives.
They are unused spares from out of service servers.
Drives 1 and 2.
Labeled MAN3367MC ULTRA160 SCSI SCA-2 / LVD ID: JW
Date of manufacture: 03-2002
Bart's SCSITools IDs them as MAN3367M SUN36G Revision: 1502
Drive 3.
Labeled MAN3367MC ULTRA160 SCSI SCA-2 / LVD ID: JW
Date of manufacture: 10-2002
Bart's SCSITools IDs it as IBM-ESXS MAN3367MC F Revision: 5408
From what I see in the labels and the PCBs, all three drives would seem to be the 'same', but the one with the SUN36G ID will only work at 80mb/s and not at 160mb/s as the one with the IBM-ESXS ID. (Adaptec 29160 card)
The info I have found on the web makes no distinction between the MAN3367M and the MAN3367MC models, apparently they are both U160 LVD.
It would seem to me that it could be a firmware issue ie: the SUN36G would need the IBM-ESXS firmware to be able to work at 160mb/s.
I have attempted to find a firmware upgrade to no avail nor have I been able to find a utility to copy the IBM-ESXS to a file.
I'd appreciate any pointers towards being able to get the IBM-ESXS firmware on the SUN36G drive.
Thanks in advance.
CIV
Michael Baeuerle
Oct 25, 2012, 8:34:58 AM10/25/12
to
CIV wrote:
>
> [...]
originally intended for. But it is also possible that the higher
transfer rates don't work reliable enough.
> I have attempted to find a firmware upgrade
> [...]
I can't help you with the firmware update. But with OEM disks, the
controller hardware may have subtle differences to the "normal"
versions. Are the 80MByte/s really not enough for your purpose, the
sustained transfer rate of such a disk should be far below. Is it worth
the risk to create a brick?
Micha
>
> [...]
> only work at 80mb/s and not at 160mb/s as the one with
> the IBM-ESXS ID. (Adaptec 29160 card)
> [...]
> the IBM-ESXS ID. (Adaptec 29160 card)
> It would seem to me that it could be a firmware issue ie:
> the SUN36G would need the IBM-ESXS firmware to be able
> to work at 160mb/s.
Likely this limitation was required for the target host the disks was
> the SUN36G would need the IBM-ESXS firmware to be able
> to work at 160mb/s.
originally intended for. But it is also possible that the higher
transfer rates don't work reliable enough.
> I have attempted to find a firmware upgrade
I can't help you with the firmware update. But with OEM disks, the
controller hardware may have subtle differences to the "normal"
versions. Are the 80MByte/s really not enough for your purpose, the
sustained transfer rate of such a disk should be far below. Is it worth
the risk to create a brick?
Micha
CIV
Oct 25, 2012, 10:23:13 AM10/25/12
to
Hello:
Thanks a lot for your prompt reply.
Much appreciated.
> Likely this limitation was required for the target host the disks was originally intended for.
Surely.
And let Sun Microsystems use the same storage devices across different
spec hardware with just a firmware change.
> But it is also possible that the higher transfer rates don't work reliable enough.
It's SUN stuff ...
Heavy iron. =-)
But then, I'm no expert, it could well be as you say.
The PCBs seem to be exactly the same.
I will have a closer look and check if I can find any differences.
> I can't help you with the firmware update.
OK, thanks anyhow.
> ... OEM disks, the controller hardware may have subtle differences to the "normal" versions.
Yes, could be.
> Are the 80MByte/s really not enough for your purpose, the sustained transfer rate of such a disk ...
From what I have seen (SCSITool) the IBM disk has higher TRs, but not
too much.
I have an option to get another two drives and would like them all to
be U160.
I bought the SUN36 thinking they were 'identical' to the IBM one,
based on their labels.
> Is it worth the risk to create a brick?
Fortunately, I have a couple of PCBs from another two SUN36Gs that
went south, so that would not be an issue.
¿Do you know of any firmware extraction tool out there?
All the SCSI utilities I have found allow uploading new FW but not
making a back up copy of it, like you do when you flash a mobo BIOS.
For obvious reasons, of course.
Thanks for your input.
Best,
CIV
Thanks a lot for your prompt reply.
Much appreciated.
> Likely this limitation was required for the target host the disks was originally intended for.
And let Sun Microsystems use the same storage devices across different
spec hardware with just a firmware change.
> But it is also possible that the higher transfer rates don't work reliable enough.
Heavy iron. =-)
But then, I'm no expert, it could well be as you say.
The PCBs seem to be exactly the same.
I will have a closer look and check if I can find any differences.
> I can't help you with the firmware update.
> ... OEM disks, the controller hardware may have subtle differences to the "normal" versions.
Yes, could be.
> Are the 80MByte/s really not enough for your purpose, the sustained transfer rate of such a disk ...
From what I have seen (SCSITool) the IBM disk has higher TRs, but not
too much.
I have an option to get another two drives and would like them all to
be U160.
I bought the SUN36 thinking they were 'identical' to the IBM one,
based on their labels.
> Is it worth the risk to create a brick?
went south, so that would not be an issue.
¿Do you know of any firmware extraction tool out there?
All the SCSI utilities I have found allow uploading new FW but not
making a back up copy of it, like you do when you flash a mobo BIOS.
For obvious reasons, of course.
Thanks for your input.
Best,
CIV
Michael Baeuerle
Oct 25, 2012, 12:07:12 PM10/25/12
to
CIV wrote:
> �Do you know of any firmware extraction tool out there?
According to this manual:
http://www.andovercg.com/datasheets/fujitsu-MAN3xxxSCSI.pdf
your disks comply to the SCSI SPC-2 standard. SPC-2 specify a generic
way to write new firmware (named "microcode" there) to a device using
the WRITE BUFFER command modes 4 to 7. A READ BUFFER command also exist,
but there are no generic modes defined to read the microcode.
Because the firmware image is very likely compressed and/or encrypted,
it may not as trivial as it look like for the disk controller to provide
the data in the transfer format. Maybe this is not supported at all.
Micha
> Michael Baeuerle wrote:
> >
> > Is it worth the risk to create a brick?
>
> Fortunately, I have a couple of PCBs from another two SUN36Gs that
> went south, so that would not be an issue.
OK.
> >
> > Is it worth the risk to create a brick?
>
> Fortunately, I have a couple of PCBs from another two SUN36Gs that
> went south, so that would not be an issue.
> �Do you know of any firmware extraction tool out there?
>
> All the SCSI utilities I have found allow uploading new FW but not
> making a back up copy of it, like you do when you flash a mobo BIOS.
>
> For obvious reasons, of course.
I think you need a special tool from the manufacturer for this purpose.
> All the SCSI utilities I have found allow uploading new FW but not
> making a back up copy of it, like you do when you flash a mobo BIOS.
>
> For obvious reasons, of course.
According to this manual:
http://www.andovercg.com/datasheets/fujitsu-MAN3xxxSCSI.pdf
your disks comply to the SCSI SPC-2 standard. SPC-2 specify a generic
way to write new firmware (named "microcode" there) to a device using
the WRITE BUFFER command modes 4 to 7. A READ BUFFER command also exist,
but there are no generic modes defined to read the microcode.
Because the firmware image is very likely compressed and/or encrypted,
it may not as trivial as it look like for the disk controller to provide
the data in the transfer format. Maybe this is not supported at all.
Micha
CIV
Oct 25, 2012, 3:14:41 PM10/25/12
to
Hello again:
On Thursday, 25 October 2012 13:08:03 UTC-3, Michael Baeuerle wrote:
> I think you need a special tool from the manufacturer for this purpose.
Yes. I fear so.
Yes, I have the manual.
> ... generic way to write new firmware ...
Yes, I have three utilities for that: Bart's SCSITools, IBM's CodeUpdate and Fujitsu's siag, so that's taken care of.
But no way to copy it from another drive. =-/
> A READ BUFFER command also exist, but there are no generic modes defined to
> read the microcode.
I think I saw it in one or two of the utilities but am utterly ignorant as to how to use it.
> Because the firmware image is very likely compressed and/or encrypted ...
I don't think the manufacturers want the code out in the open.
It would be an 'easy peasy' matter then.
I have used these utilities before (codeupdt)to upgrade IBM SCSI drive firmware for a few disks a couple of years ago. It was kindly provided to me by a forum member by the name of Rienstra.
Thanks for the imput.
Cheers,
CIV
On Thursday, 25 October 2012 13:08:03 UTC-3, Michael Baeuerle wrote:
> I think you need a special tool from the manufacturer for this purpose.
Yes, I have the manual.
> ... generic way to write new firmware ...
Yes, I have three utilities for that: Bart's SCSITools, IBM's CodeUpdate and Fujitsu's siag, so that's taken care of.
But no way to copy it from another drive. =-/
> A READ BUFFER command also exist, but there are no generic modes defined to
> read the microcode.
> Because the firmware image is very likely compressed and/or encrypted ...
> Maybe this is not supported at all.
Most probably.
I don't think the manufacturers want the code out in the open.
It would be an 'easy peasy' matter then.
I have used these utilities before (codeupdt)to upgrade IBM SCSI drive firmware for a few disks a couple of years ago. It was kindly provided to me by a forum member by the name of Rienstra.
Thanks for the imput.
Cheers,
CIV
0 new messages