I downloaded and read through the "IBM TotalStorage Tape Device
Drivers - Installation and Users's Guide" and then ran some tests on our
supposedly upgraded 3590J cartridges.
First, I created a 10.1 gig file and then wrote it to the suppsedly
upgraded cartridge with compression turned off. It worked fine so then I
ran the append command and tried writing the same file which should have
started writing where the first file left off. It failed as it ran out of
tape but looks like it was aproximately a 20 gig tape. A "qrypart" command
says that the volume has 1 SDP parttition and it is 19531 meg in size. This
should be expected for a normal 3590J cartridge in a 3590E tape drive as
3590E drives can write as much as 20 gig on 3590J cartridges with
compression turned off.
Second, I took another one of the supposed upgraded cartridges and
did just a "qrypart" command on it and it showed as 1 SDP partition that
was 9765 megs in size. After running the "erase" command, it now shows as
19531 meg in size. It looks like our vendor sold us 3590J cartridges that
had never been used in a 3590E drive as the partition size on them was
still set at 10 gig instead of the expected 20 gig.
So in conclusion, I highly suggest you thoroughly check what your
vendors are selling you. Since TSM can use compresson and the tape drives
use their own compression, it may be hard to tell just exactly how much
capacity a cartridge can hold without running some tests yourself.
We're still waiting for an answer on what they think was done to
upgrade the 3590J cartridges but so far, out vendor says they bought them
from someone else and were told the cartridges were upgraded to hold the
equivalent of a 3590K cartridge. Unless I did something wrong, they are
ripping people off and I'm sure we went the first. As you can see above, I
left out the vendors names.
John
Forum: ADSM.ORG - ADSM / TSM Mailing List Archive
Date: Feb 05, 20:44
From: Richard Sims <nobody at nowhere.com>
>We are in the midst of replacing our existing 3590J cartridges with 3590J
>cartridges that have been (supposedly) upgraded to allow 20/40 gigabyte
>capacity instead of of 10/20 capacity. I can successfully checkin the new
>libvolumes but until they actually get used, I have no way of even
guessing
>how much data they can hold. I want to verify that the new 3590J
cartridges
>we just bought have really been upgraded to support 20/40 capacity. Is
>there a way to force data to goto my new libvolume? Maybe there is a
atape
>(or whatever the AIX command is) command to do this?
>Since we are replacing existing 3590J cartridges that are labeled
>100000-100300 with (supposedly upgraded) 3590J cartridges with the same
>labels, do I have any potential problems in store other than removing a
>libvolume that still has data on it? They will all stay in the same
device
>class.
I've never heard of "upgrading a tape cartridge", and would be curious to
have some references to what that's supposed to be about. (It sounds like
something one would be offered in spam.) The way we mortals get more data
on a tape is by upgrading our drives to higher density; or we switch to
longer
tapes, such as the 3590K (in concert with drive upgrading to accommodate
the
handling of thinner tape).
To prove the existence of magic in this universe you need to perform a
case
study: as with any new technology being introduced to your shop, you
should
prove its capabilities before putting it into production. If you're a
3590 user
you need to be well acquainted with the "IBM TotalStorage Tape Device
Drivers:
Installation and User's Guide" manual (see topic "tapeutil" in ADSM
QuickFacts).
Give a tape drive over to the test, turn off tape drive compression, and
write
a controlled amount of data to the tape using the tapeutil command, or
even an
OS utility like the Unix 'dd' command, trying to fill it according to the
capacity you expect it should hold.
You will need to consider Devclass values according to what you really
have
there, which is a mystery to me. There are READMEs and List postings
regarding
the handling of drive upgrades vs. existing tapes.
Richard Sims http://people.bu.edu/rbs
