Re: Quantum XP39100S Atlas II and 8GB limit (not int 13h ext'd related?)

"Michael Baeuerle" <michael.baeuerle@xxxxxxxx> wrote in message news:jd75j4-cu3.ln1@xxxxxxxxxxxxxxxxxxx
da9000 wrote:

[...]
Now a couple more general questions:

1) I spent hours trying to find out how to get the true max LBA limit
with SCU, but no luck :( "show capacity" returns the imposed limit.
Showing the mode pages doesn't have anything either... Anyone know?

I haven't found a standard method to read out the maximum supported
number of LBs. As I have written in the other posting, the capacity can
be changed using a block descriptor. You can try to set the new capacity
to 0xFFFFFFFF because the SCSI spec says in SBC2 document chapter
6.3.2.2:

If the NUMBER OF BLOCKS field is set to a value greater than the
maximum capacity of the device and less than FFFFFFFFh, then the MODE
SELECT command shall be terminated with CHECK CONDITION status
[...]

This means you can try to increase the capacity in intervals, if you get
a CHECK CONDITION you have crossed the "max LBA" boundary.

But the standard has one exception that offers a more easy way:

It's not an exception, it's an option.


If the NUMBER OF BLOCKS field is set to FFFFFFFFh,
the logical unit shall be set to its maximum capacity.

[...]
If the content of the BLOCK LENGTH field is the same as the current
block length this capacity setting shall take effect on successful
completion of the MODE SELECT command.

With this method you still cannot read the maximum value

Yes, you can. This is what allows you to.
What is doesn't do is show it to you without also setting it.
So you will have to set it back afterwards.

but you can order the disk to use it.

And that's why you can. You even said so already in the other post.

I would try someting like:
scu -f /dev/sdx set device capacity 0xFFFFFFFF

A LL-format should not be necessary

Period.

as long as you don't change the blocksize too ...

And why the hell would you.
Apparently, changing both will make the unit 'Format Corrupted',
until you do a Format Unit. So this is to be strongly avoided.
Unfortunately it doesn't say whether setting the value back to
current blocksize will revert the Format Corrupted state or not.


[...]
2) According to the mode pages (and from the noise of this beast),
it's a 7200RPM drive,

Yes.

yet doing a low-level format takes 5-6 (maybe more, not sure, I fell
asleep) hours. Isn't this way too slow?

No, I have seen format times of several hours on Seagate disks with many
heads too.

Large # of heads has nothing to do with it other than that it makes for a very
large drive and large drives obviously take longer to format than smaller ones.
So capacity divided by throughput decides the time it takes to write it all.
At ~7.5 MB/s avg this drive should (in theory) be able to complete it in
9,100,000,000/7,500,000/60 = ~20 minutes, without verification.
(Of course that's counting on that the drive formats efficiently and is
tracking well).


3) After the low-level format is done, and I use dd if=/dev/sda of=/
dev/null bs=16k (Linux, MacOSX, *BSD) to exercise the drive, I only
get 1.1 - 1.5MB/sec. Tad bit low for a 7200RPM drive with 10 (!)
platters and with a FastSCSI card (either 10MB/sec or 20MB/sec
settings without complaints), in synchronus mode, isn't it?

Are the values higher if you use more than 16k transfersize?

Oops, missed that.
Fortunately 16kB isn't half bad and should still give you about 2/3
(6-7MB/s) of what the drive can do at 64kB or higher transfer lengths.
1-1.5 MB/s is still way below that.

I'm still counting on Cache being disabled. That or physical shock.



Micha
.

Relevant Pages

  • Re: The end of the world as we know it
    ... we build to meet peak hour - capacity sufficient to serve peak hour ... no matter - the taxes are expended to meet the needs of maximum ... access and mobility for a basic price and more intensive users pay ... that drives 40 miles a day in the height of rush hour. ...
    (misc.transport.road)
  • Re: External Large Capacity Drive Formatting
    ... My USB enclosure is a Noontec Gigasave--documentation says it can suport ... When I format I only see 38GB. ... It seems I can't get to see the full capacity unless I format this ... capacity drives with XP? ...
    (microsoft.public.windowsxp.hardware)
  • Re: XP does not see all of a WD 160GB drive that was added as a slave
    ... > and binary measurements of a drive's capacity. ... > as 1,048,576 bytes and a gigabyte as 1,073,741,824 bytes. ... >>| My Device Manager/Disk Drives sees it as 152625MB. ...
    (microsoft.public.windowsxp.hardware)
  • Re: External Large Capacity Drive Formatting
    ... partition in Drive Manager that was created with Fat 32 format and recreate ... When I format I only see 38GB. ... It seems I can't get to see the full capacity unless I format this ... capacity drives with XP? ...
    (microsoft.public.windowsxp.hardware)
  • Re: ANSWER
    ... flash drive was apparently mislabeled re capacity. ... USB 8GB drive, 2Gb limit ... disk drives over 2 GB. ... is the only format that will accommodate more than 4 GB file sizes. ...
    (microsoft.public.windowsxp.hardware)