Farren,
The planned feature 'collocation groups' will fix this problem.
You will be able to say in essence "collocate the files of nodes
<list of nodenames> together".
In the mean time you can do what I used to do to get by the
"small node on a large tape" problem.
1. mark all the tapes in the pool readonly, and don't let any more
scratches be added (this will be undone later).
("upd vol * acc=readonly wherestg=<poolname>")
2. mark a target filling tape readwrite (currently used by node A).
3. do move data on one or more of the low utilized tapes
(currently used by nodes B, C, D etc); the server
will move the data onto the target tape.
4. fill the target tape to 50%.
5. mark the first target readonly, mark a new target readwrite
6. loop back to 3 until you have consolidated all you low utilization tapes
onto a few 50% utilized tapes.
7. reverse what you did in step 1.
TSM will now target the same volumes for nodes A, B, C, D, but it will mount it
4 times.
Hope this helps,
Bill Colwell
At 10:17 AM 4/7/2003, you wrote:
>Hi All
>
>Running TSM 4.2.2.12 on Sol 2.7 server. At present we use collocation on
>the tapes within the TAPEPOOL (i.e. in the 3494 library). We are getting to
>a point where the number of available scratch tapes in the lib is getting
>small and I'm considering turning collocation off. We always seem to have
>about 25-30 tapes that are under 10% used which seems like a large waste of
>space.
>
>I understand that collocation speeds things up in the event of a system
>restore, but this is something that we rarely (in fact, never in the last
>two years) need to do and anyway, the library can handle all the tape
>mounts and stuff.
>
>This is all rather long winded, but my question is simple. Are there any
>things I should be aware of before setting collocation to NO. i.e., are
>there any possible problems I may encounter.
>
>Many thanks
>
>Farren Minns - John Wiley & Sons Ltd
>*****************************************************************************
>
>This email transmission is confidential and intended for the person or
>organisation it is addressed to. If you are not the intended recipient, you
>must not copy, distribute, or disseminate the information, open any
>attachment, or take any action in reliance of it. If you have received this
>message in error please notify the sender.
>
>Any views expressed in this message are those of the individual sender,
>except where the sender specifically states otherwise.
>
>Although this email has been scanned for viruses you should rely on your
>own virus check, as the sender takes no responsibility for any damage
>arising out of any bug or virus infection.
>*****************************************************************************
----------
Bill Colwell
C. S. Draper Lab
Cambridge Ma.
|
