SEO

A friend has a Sun barcoded (540-4689) hard drive, made by Fujitsu.
Unless I am very much mistaken, the disk is half the capacity marked on
the drive.

When installing Solaris 10 it appeared to be 18 GB, not 36 GB as he
thought it was. After Solaris was installed I run format and see this
(ignore disk 1, only consider disk 0).

# format
Searching for disks...done

c1t1d0: configured with capacity of 68.36GB


AVAILABLE DISK SELECTIONS:
0. c1t0d0 <SUN18G cyl 7506 alt 2 hd 19 sec 248>
/pci@1f,0/pci@1/scsi@8/sd@0,0
1. c1t1d0 <SEAGATE-ST373307LC-0007 cyl 49780 alt 2 hd 4 sec 720>
/pci@1f,0/pci@1/scsi@8/sd@1,0

Note disk 0 is 18GB, which agrees with what the Solaris install program
thought.

Yet the disk is marked 36 GB and the P/N indicates it is 36 GB.

This is the markings on the disk. Is this a one-off?

540-4689-01
S/N 0066152-0328N03HYW.
Sun 36G MAP3367N Sun 36G rpm

Fujitsu Ltd model MAP3367NC
DATE 2003-07
Part number CA06200-B52200NB

His flash on the camera is not working, but here is a (poor) picutre

http://www.drkirkby.co.uk/disk.jpg



--
David Kirkby,
G8WRB

Please check out http://www.g8wrb.org/
of if you live in Essex http://www.southminster-branch-line.org.uk/

In comp.sys.sun.admin David Kirkby <david.kirkby@onetel.net> wrote:
> AVAILABLE DISK SELECTIONS:
> 0. c1t0d0 <SUN18G cyl 7506 alt 2 hd 19 sec 248>
> /pci@1f,0/pci@1/scsi@8/sd@0,0
> 1. c1t1d0 <SEAGATE-ST373307LC-0007 cyl 49780 alt 2 hd 4 sec 720>
> /pci@1f,0/pci@1/scsi@8/sd@1,0

Try relabeling the disk.

--
Daniel

Daniel Rock wrote:
> In comp.sys.sun.admin David Kirkby <david.kirkby@onetel.net> wrote:
>
>>AVAILABLE DISK SELECTIONS:
>> 0. c1t0d0 <SUN18G cyl 7506 alt 2 hd 19 sec 248>
>> /pci@1f,0/pci@1/scsi@8/sd@0,0
>> 1. c1t1d0 <SEAGATE-ST373307LC-0007 cyl 49780 alt 2 hd 4 sec 720>
>> /pci@1f,0/pci@1/scsi@8/sd@1,0
>
>
> Try relabeling the disk.
>
I've done that, to no effect.

I'm a bit reluctant to do anything more with it, as it now has a
complete install of Solaris 10 on it.

But it is suspicous that format shows the type number being "SUN18G". I
can't see where format can pick that up from, other than the disk
firmware. Surely Sun would not call a 36 GB disk "SUN18G" would they??

The other disk is shown as a SEAGATE-ST373307LC, and is 73GB, as the
part number would indicate.

Here's the partition data shown on format for the suspicous disk.

partition> print
Current partition table (original):
Total disk cylinders available: 7506 + 2 (reserved cylinders)

Part Tag Flag Cylinders Size Blocks
0 root wm 112 - 981 1.95GB (870/0/0) 4099440
1 var wm 982 - 1851 1.95GB (870/0/0) 4099440
2 backup wm 0 - 7505 16.86GB (7506/0/0) 35368272
3 swap wu 0 - 111 257.69MB (112/0/0) 527744
4 unassigned wm 0 0 (0/0/0) 0
5 unassigned wm 1852 - 3325 3.31GB (1474/0/0) 6945488
6 usr wm 3326 - 6201 6.46GB (2876/0/0) 13551712
7 home wm 6202 - 7505 2.93GB (1304/0/0) 6144448

--
David Kirkby,
G8WRB

Please check out http://www.g8wrb.org/
of if you live in Essex http://www.southminster-branch-line.org.uk/

In comp.sys.sun.admin David Kirkby <david.kirkby@onetel.net> wrote:
> Daniel Rock wrote:
>> In comp.sys.sun.admin David Kirkby <david.kirkby@onetel.net> wrote:
>>
>>>AVAILABLE DISK SELECTIONS:
>>> 0. c1t0d0 <SUN18G cyl 7506 alt 2 hd 19 sec 248>
>>> /pci@1f,0/pci@1/scsi@8/sd@0,0
>>> 1. c1t1d0 <SEAGATE-ST373307LC-0007 cyl 49780 alt 2 hd 4 sec 720>
>>> /pci@1f,0/pci@1/scsi@8/sd@1,0
>>
>>
>> Try relabeling the disk.
>>
> I've done that, to no effect.
>
> I'm a bit reluctant to do anything more with it, as it now has a
> complete install of Solaris 10 on it.

Try the following:

Start format with the -e option

format -e

select your disk, then relabel

label

choose "Auto configuration via generic SCSI-2"

Does it still show up as 18GB disk? If yes, go to the scsi menu

scsi

and do an inquiry

inquiry

The output from the inquiry command should enlighten what model is really
used.

BTW the "SUN18G" is just a magic string in the inquiry to "beautify" the
VTOC and make the real vendor invisible.

--
Daniel

Daniel Rock wrote:
> In comp.sys.sun.admin David Kirkby <david.kirkby@onetel.net> wrote:

> Try the following:
>
> Start format with the -e option
>
> format -e
>
> select your disk, then relabel
>
> label
>
> choose "Auto configuration via generic SCSI-2"

That was not an option - the options are:

format> label
[0] SMI Label
[1] EFI Label
Specify Label type[0]:

> Does it still show up as 18GB disk? If yes, go to the scsi menu

Can't tell at that point (I tried 0 and 1). I should add I am doing this
on the boot disk, there is no other boot disk and I am 100 miles from
the machine (the machine is located at a friend's house).

> scsi
>
> and do an inquiry
>
> inquiry

Interesting.
>
> The output from the inquiry command should enlighten what model is really
> used.
>
> BTW the "SUN18G" is just a magic string in the inquiry to "beautify" the
> VTOC and make the real vendor invisible.
>

Thanks for that.

"inquiry" that indicates it is 36 GB - see below. So the sticky label on
the disk is correct and the "SUN18G" shown by format is incorrect. But
how do I get access to the remaining 18 GB of disk space?

I expect it asking too much to get it without reinstalling Solaris? If
so, it might be less hassle to waste 18 GB. That said, if it is
recoverable, it might be worth a try, as only Solaris has been put on it
- no data, apart from gcc from Sunfreeware.

Any idea how this might have happened? The Sun had two identical disks
in it, which were mirrored I think.


Inquiry:
00 00 04 02 29 00 01 3a 46 55 4a 49 54 53 55 20 ....)..:FUJITSU
4d 41 50 33 33 36 37 4e 20 53 55 4e 33 36 47 20 MAP3367N SUN36G
30 33 30 31 30 33 32 38 4e 30 33 48 59 57 03010328N03HYW

Vendor: FUJITSU
Product: MAP3367N SUN36G
Revision: 0301
Removable media: no
Device type: 0
ISO version: 0
ECMA version: 0
ANSI version: 4
Async event notification: no
Terminate i/o process msg: no
Response data format: 2
Additional length: 41
Relative addressing: no
32 bit transfers: no
16 bit transfers: yes
Synchronous transfers: yes
Linked commands: yes
Command queueing: yes
Soft reset option: no

scsi>

--
David Kirkby,
G8WRB

Please check out http://www.g8wrb.org/
of if you live in Essex http://www.southminster-branch-line.org.uk/

David Kirkby <david.kirkby@onetel.net> writes:

>A friend has a Sun barcoded (540-4689) hard drive, made by Fujitsu.
>Unless I am very much mistaken, the disk is half the capacity marked on
>the drive.

Looks like it was labelled incorrectly by someone.


Clear the label and relabel (will lose all data except when the new label
matches the same partitions)

Casper

Casper H.S. Dik <Casper.Dik@Sun.COM> writes:

>David Kirkby <david.kirkby@onetel.net> writes:

>>A friend has a Sun barcoded (540-4689) hard drive, made by Fujitsu.
>>Unless I am very much mistaken, the disk is half the capacity marked on
>>the drive.

>Looks like it was labelled incorrectly by someone.


>Clear the label and relabel (will lose all data except when the new label
>matches the same partitions)


And you zap the old label using:


dd if=/dev/zero of=/dev/rdsk/..... count=16

which will overwrite the first 8K of the disk (this will leave ufs
filesystem intact but you can only recover them by repartition with
the exact same partition boundaries).

Casper

Casper H.S. Dik wrote:
> David Kirkby <david.kirkby@onetel.net> writes:
>
>
>>A friend has a Sun barcoded (540-4689) hard drive, made by Fujitsu.
>>Unless I am very much mistaken, the disk is half the capacity marked on
>>the drive.
>
>
> Looks like it was labelled incorrectly by someone.
>
>
> Clear the label and relabel

Hi Casper,

How is the best way to clear the label? You posted something some time
back in 2002

http://groups.google.co.uk/group/com...6bef3587322903

and said:

"What you could try is zap the label with dd, assuming you have no
data in the disk:

dd if=/dev/zero of=/dev/dsk/cX...... count=16

and then re-run format."

I would have thought one needs to write to the raw device
(/dev/rdsk/c?t?d?s2), but perhaps I am wrong.

Will it be necessary to boot from a CD to do this? The disk is the boot
disk.

The problem I have is explaining to a friend how to do it, so if I can
ssh to the box, and do it myself, rather than explain every single
command to him, it would be preferable! His UNIX knowledge is next to zero.

> (will lose all data except when the new label
> matches the same partitions)

Chris Lawson said:

# prtvtoc -s /dev/rdsk/{identical disk id}s2 > tmp_file
# fmthard -s tmp_file /dev/rdsk/{screwed up disk}s2

which looks attractive if it will work. We have a second identical disk,
which unlike this one does work at its full capacity.

>
> Casper

Dave Kirkby
--
David Kirkby,
G8WRB

Please check out http://www.g8wrb.org/
of if you live in Essex http://www.southminster-branch-line.org.uk/

In comp.sys.sun.admin David Kirkby <david.kirkby@onetel.net> wrote:
> Daniel Rock wrote:
>> In comp.sys.sun.admin David Kirkby <david.kirkby@onetel.net> wrote:
>
>> Try the following:
>>
>> Start format with the -e option
>>
>> format -e
>>
>> select your disk, then relabel
>>
>> label
>>
>> choose "Auto configuration via generic SCSI-2"
>
> That was not an option - the options are:

Sorry, meant "type" instead of "label".

--
Daniel

I dunno...
This sounds suspiciously like a mode page problem.
I wonder if perhaps the wrong mode page was downloaded to the drive...
i.e., the drive is 36G, but it has a mode page for an 18G drive.

Seagate makes depop versions of drives... that is, same drive with
fewer platters/heads. Everything else is the same, including firmware
(generally speaking). Only difference is in the mode page settings (and
the physicality of the drive, of course).

Downloading a proper mode page would (if this is the case) correct the
problem. But I wouldn't mess with a boot image. If it's working and
you like it, don't worry about it.