Is there anything I can about this, other than throwing the disk away and
getting another?

Feb 4 15:06:42 apophis scsi: [ID 107833 kern.warning] WARNING: /pci@1f,4000/scsi@3/sd@0,0 (sd6):
Feb 4 15:06:42 apophis Error for Command: read(10) Error Level: Fatal
Feb 4 15:06:42 apophis scsi: [ID 107833 kern.notice] Requested Block: 65151308 Error Block: 65151415
Feb 4 15:06:42 apophis scsi: [ID 107833 kern.notice] Vendor: FUJITSU Serial Number: 0204X24837
Feb 4 15:06:42 apophis scsi: [ID 107833 kern.notice] Sense Key: Media Error
Feb 4 15:06:42 apophis scsi: [ID 107833 kern.notice] ASC: 0x15 (positioning error detected by read of medium), ASCQ: 0x2, FRU: 0x0
Feb 4 15:06:42 apophis md_stripe: [ID 641072 kern.warning] WARNING: md: d10: read error on /dev/dsk/c0t0d0s0

Fortunately, I mirrored it this very morning....

--
"Other people are not your property."
[email me at huge [at] huge [dot] org [dot] uk]

Huge wrote:
> Is there anything I can about this, other than throwing the disk away and
> getting another?
>
> Feb 4 15:06:42 apophis scsi: [ID 107833 kern.warning] WARNING: /pci@1f,4000/scsi@3/sd@0,0 (sd6):
> Feb 4 15:06:42 apophis Error for Command: read(10) Error Level: Fatal
> Feb 4 15:06:42 apophis scsi: [ID 107833 kern.notice] Requested Block: 65151308 Error Block: 65151415
> Feb 4 15:06:42 apophis scsi: [ID 107833 kern.notice] Vendor: FUJITSU Serial Number: 0204X24837
> Feb 4 15:06:42 apophis scsi: [ID 107833 kern.notice] Sense Key: Media Error
> Feb 4 15:06:42 apophis scsi: [ID 107833 kern.notice] ASC: 0x15 (positioning error detected by read of medium), ASCQ: 0x2, FRU: 0x0
> Feb 4 15:06:42 apophis md_stripe: [ID 641072 kern.warning] WARNING: md: d10: read error on /dev/dsk/c0t0d0s0
>
> Fortunately, I mirrored it this very morning....
>

Cables is another possibility of course.

You might want to look at the defects in the disk - have they grown?

format> defect


DEFECT MENU:
primary - extract manufacturer's defect list
grown - extract the grown defects list
both - extract both primary and grown defects lists
print - display defect list
dump - dump defect list to file
!<cmd> - execute <cmd>, then return
quit

defect> grown
Extracting grown defects list...Extraction complete.
Defect List has a total of 0 defects.


--
Dave K

http://www.southminster-branch-line.org.uk/

Please note my email address changes periodically to avoid spam.
It is always of the form: month-year@domain. Hitting reply will work
for a couple of months only. Later set it manually. The month is
always written in 3 letters (e.g. Jan, not January etc)

Since you know the bad block, in this case 65151308, you can go into format
and do a repair against that block on the disk.

Be sure there are no mounted file systems on the disk while doing this. Also
you should run fsck against all partitions on this disk before trying to
mount them back up again. If you can do all this in single-user mode all the
better.

If you're getting other errors or many different bad blocks, then replacing
the disk may be a better solution.

On Sat, 04 Feb 2006 14:33:32 -0500, Trinean wrote:

> Since you know the bad block, in this case 65151308, you can go into
> format and do a repair against that block on the disk.
>
> Be sure there are no mounted file systems on the disk while doing this.
> Also you should run fsck against all partitions on this disk before trying
> to mount them back up again. If you can do all this in single-user mode
> all the better.
>
> If you're getting other errors or many different bad blocks, then
> replacing the disk may be a better solution.

So far I've added 13 bad blocks, but it keeps throwing up new ones.

Anyone recommend which type & where to buy a replacement 36Gb disk?

For example, one of these?

http://www.misco.co.uk/productinform...068P%208MB.htm

--
"Other people are not your property."
[email me at huge [at] huge [dot] org [dot] uk]

Huge wrote:
> On Sat, 04 Feb 2006 14:33:32 -0500, Trinean wrote:
>
>
>>Since you know the bad block, in this case 65151308, you can go into
>>format and do a repair against that block on the disk.
>>
>>Be sure there are no mounted file systems on the disk while doing this.
>>Also you should run fsck against all partitions on this disk before trying
>>to mount them back up again. If you can do all this in single-user mode
>>all the better.
>>
>>If you're getting other errors or many different bad blocks, then
>>replacing the disk may be a better solution.
>
>
> So far I've added 13 bad blocks, but it keeps throwing up new ones.
>

I've had that but found it was a cable issue. Watch that.

You said in your original post "Fortunately, I mirrored it this very
morning...." Perhaps you have changed the scsi setup for the worst. I
note the disk you are considering is a 68-pin one, so you need to get
the termination right on them - unlike SCA disks where the backplane
usually sorts that out.

> Anyone recommend which type & where to buy a replacement 36Gb disk?
>
> For example, one of these?
>
> http://www.misco.co.uk/productinform...068P%208MB.htm
>


--
Dave K

http://www.southminster-branch-line.org.uk/

Please note my email address changes periodically to avoid spam.
It is always of the form: month-year@domain. Hitting reply will work
for a couple of months only. Later set it manually. The month is
always written in 3 letters (e.g. Jan, not January etc)

HI,

Huge wrote:
> On Sat, 04 Feb 2006 14:33:32 -0500, Trinean wrote:
>
>
>>Since you know the bad block, in this case 65151308, you can go into
>>format and do a repair against that block on the disk.
>>
>>Be sure there are no mounted file systems on the disk while doing this.
>>Also you should run fsck against all partitions on this disk before trying
>>to mount them back up again. If you can do all this in single-user mode
>>all the better.
>>
>>If you're getting other errors or many different bad blocks, then
>>replacing the disk may be a better solution.
>
>
> So far I've added 13 bad blocks, but it keeps throwing up new ones.
>
> Anyone recommend which type & where to buy a replacement 36Gb disk?
>
> For example, one of these?
>
> http://www.misco.co.uk/productinform...068P%208MB.htm
>
What machine is it to be installed in?, The link points to a no-hotswap
that is a 68 pol SCSI connector and separate power connector is that right?

Apart from that I would maybe get a bigger disk since the are usually
better price/GB!

/michael

On Sun, 05 Feb 2006 16:28:08 +0000, Dave (from the UK) wrote:

> Huge wrote:
>> On Sat, 04 Feb 2006 14:33:32 -0500, Trinean wrote:
>>
>>
>>>Since you know the bad block, in this case 65151308, you can go into
>>>format and do a repair against that block on the disk.
>>>
>>>Be sure there are no mounted file systems on the disk while doing this.
>>>Also you should run fsck against all partitions on this disk before
>>>trying to mount them back up again. If you can do all this in
>>>single-user mode all the better.
>>>
>>>If you're getting other errors or many different bad blocks, then
>>>replacing the disk may be a better solution.
>>
>>
>> So far I've added 13 bad blocks, but it keeps throwing up new ones.
>>
>>
> I've had that but found it was a cable issue. Watch that.

Well, it seems (what's the emticon for crossed fingers?!??!) OK now. It's
just rebuilding the mirrors.

> You said in your original post "Fortunately, I mirrored it this very
> morning...." Perhaps you have changed the scsi setup for the worst.

The thought crossed my mind, but the one with errors is the "live" one.

> I note
> the disk you are considering is a 68-pin one, so you need to get the
> termination right on them - unlike SCA disks where the backplane usually
> sorts that out.

That's the sort of question I meant - I've never looked inside a '60, so
they use SCA disks, yes?

--
"Other people are not your property."
[email me at huge [at] huge [dot] org [dot] uk]

On Sun, 05 Feb 2006 17:44:15 +0100, Michael Laajanen wrote:

> HI,
>
> Huge wrote:
>> On Sat, 04 Feb 2006 14:33:32 -0500, Trinean wrote:
>>
>>
>>>Since you know the bad block, in this case 65151308, you can go into
>>>format and do a repair against that block on the disk.
>>>
>>>Be sure there are no mounted file systems on the disk while doing this.
>>>Also you should run fsck against all partitions on this disk before
>>>trying to mount them back up again. If you can do all this in
>>>single-user mode all the better.
>>>
>>>If you're getting other errors or many different bad blocks, then
>>>replacing the disk may be a better solution.
>>
>>
>> So far I've added 13 bad blocks, but it keeps throwing up new ones.
>>
>> Anyone recommend which type & where to buy a replacement 36Gb disk?
>>
>> For example, one of these?
>>
>> http://www.misco.co.uk/productinform...068P%208MB.htm
>>
> What machine is it to be installed in?,

An Ultra 60.

> The link points to a no-hotswap
> that is a 68 pol SCSI connector and separate power connector is that
> right?

Dunno! That's what I'm asking!

>
> Apart from that I would maybe get a bigger disk since the are usually
> better price/GB!

It's handy to have a 36Gb 'cos it's part of a mirrored pair. If I need
more storage, that can go on my Ultra 2 "household server", which has a
tape drive so I can do proper backups.


--
"Other people are not your property."
[email me at huge [at] huge [dot] org [dot] uk]

Huge wrote:

>>I've had that but found it was a cable issue. Watch that.
>
>
> Well, it seems (what's the emticon for crossed fingers?!??!) OK now. It's
> just rebuilding the mirrors.
>
>
>>You said in your original post "Fortunately, I mirrored it this very
>>morning...." Perhaps you have changed the scsi setup for the worst.
>
>
> The thought crossed my mind, but the one with errors is the "live" one.

There are two controllers on that machine. One controllers does the two
internal disks, CD and tape if you have them.

The second controller controls anything attached to the rear SCSI
connector.

>>I note
>>the disk you are considering is a 68-pin one, so you need to get the
>>termination right on them - unlike SCA disks where the backplane usually
>>sorts that out.
>
>
> That's the sort of question I meant - I've never looked inside a '60, so
> they use SCA disks, yes?

If you mean an Ultra 60, it takes two internal SCSI disks with 80-pin
SCA connectors, not the 68-pin connectors you provided a link to. So
that disk you selected could not be fitted (easily at least) inside.

The dual channel SCSI controller is an Ultra Wide one, based on a 53c876
chip that is capable of a maximum of 40 Mb/s. So there is not really
anything to be gained by going for a Ultra 320 disk, although a new disk
is probably a bit quicker at 40 Mb/s than an old disk that is capable of
40 Mb/s.


--
Dave K

http://www.southminster-branch-line.org.uk/

Please note my email address changes periodically to avoid spam.
It is always of the form: month-year@domain. Hitting reply will work
for a couple of months only. Later set it manually. The month is
always written in 3 letters (e.g. Jan, not January etc)

HI,

Huge wrote:
> On Sun, 05 Feb 2006 17:44:15 +0100, Michael Laajanen wrote:
>
>
>>HI,
><snip>

>
>>What machine is it to be installed in?,
>
>
> An Ultra 60.
>
>
>>The link points to a no-hotswap
>>that is a 68 pol SCSI connector and separate power connector is that
>>right?
>
>
> Dunno! That's what I'm asking!
>
Then you should have a disk with a SCA connector ofter refered to as
"hot swap" or 80-pin SCSI drive.



/michael