Howard,
I don't see how that helps. So you have 10 drives, and you set
the mount maximum to 9 instead of "DRIVES" so you always have 1 free.
When a restore comes along, won't it also be bound by the rule to not
have more than 9 mounts at any given time? Wouldn't it also cause
another process to get canceled in order to free up tape drives to get
down to a maximum of 9?
In reading the Help I don't see any details that explains why
restores get to use that tenth drive. I don't mean to doubt your word, I
would just like to see this explained in the manual, and the Help
doesn't seem to mention it.
Best Regards,
John D. Schneider
Phone: 314-364-3150
Cell: 314-750-8721
Email: John.Schneider AT Mercy DOT net
-----Original Message-----
From: ADSM: Dist Stor Manager [mailto:ADSM-L AT VM.MARIST DOT EDU] On Behalf Of
Howard Coles
Sent: Tuesday, November 04, 2008 1:02 PM
To: ADSM-L AT VM.MARIST DOT EDU
Subject: Re: [ADSM-L] Offsite reclamation problem
Yes, look up the "mount options" option in the device class (In ISC).
You can set this to a hard number, such as in the case we've been
speaking of to 8 or 9 (for 2 or 1 drive free), or allow TSM to reserve a
drive by setting it to "Up to the number of online drives in the
library".
See Ya'
Howard
> -----Original Message-----
> From: ADSM: Dist Stor Manager [mailto:ADSM-L AT VM.MARIST DOT EDU] On Behalf
> Of Schneider, John
> Sent: Tuesday, November 04, 2008 12:35 PM
> To: ADSM-L AT VM.MARIST DOT EDU
> Subject: Re: [ADSM-L] Offsite reclamation problem
>
> All,
>
> I have often heard people say, as Linday just did, "keep at least one
> drive free for user restores". Is there in fact any way to cause TSM
> to
> leave one tape drive always free, so no other tape process besides a
> restore can access it?
>
> In our environment, we have 10 TSM servers sharing a tape library with
> 24 LTO4 tape drives. Each day they all have to accomplish the entire
> daily cycle of backup stgpools, db backups, migrations, reclamations,
> etc. and these are driven by perl scripts on each of the 10 instances.
> I cannot think of any reasonable way for all of these to get their
jobs
> done as quickly as possible, but somehow remain sensitive to what all
> the other instances need to do so that they won't use all the tape
> drives available. We use most of the 24 tape drives most of the time
> as
> it is.
>
> I can think of some extremely inefficient ways, like making all the
> backup storage pools wait until all of them are finished, then
> everybody
> goes on to the next step, etc. But this will inevitably cause long
> delays where half the tape drives are sitting idle waiting for one or
> two slow instances to finish before going on to the next step. We
> can't
> afford that sort of waste in the schedule.
>
> Is there some feature to do this that I missed?
>
> Best Regards,
>
> John D. Schneider
> Phone: 314-364-3150
> Cell: 314-750-8721
> Email: John.Schneider AT Mercy DOT net
>
>
> -----Original Message-----
> From: ADSM: Dist Stor Manager [mailto:ADSM-L AT VM.MARIST DOT EDU] On Behalf
> Of
> lindsay morris
> Sent: Tuesday, November 04, 2008 11:57 AM
> To: ADSM-L AT VM.MARIST DOT EDU
> Subject: Re: [ADSM-L] Offsite reclamation problem
>
> Lots of people have a best-practice of always keeping at least one
> drive free for user restores.
> That minimizes the problem.
>
> It makes users happy too, because
> even though a restore does pre-empt another task, it may take TSM 40
> minutes to finish the reclamation it was working on and give up the
> drive.
>
> So the user has to sit waiting for far too long (in some cases).
>
>
> ------
> Lindsay Morris
> Principal
> www.tsmworks.com
> 919-403-8260
> lindsay AT tsmworks DOT com
>
>
>
> On Nov 4, 2008, at Nov 4, 12:21 PM, Bos, Karel wrote:
>
> > No :)
> >
> >
> > -----Original Message-----
> > From: ADSM: Dist Stor Manager [mailto:ADSM-L AT VM.MARIST DOT EDU] On
> > Behalf Of
> > Thomas Denier
> > Sent: dinsdag 4 november 2008 18:14
> > To: ADSM-L AT VM.MARIST DOT EDU
> > Subject: Offsite reclamation problem
> >
> > We have a 5.4.2.0 TSM server running under mainframe Linux. We have
> > five
> > tape drives available for our primary tape storage pool and five
tape
> > drives available for our copy storage pool. We run offsite tape
> > reclaimation with 'maxproc=5'. If a client runs a restore while off
> > reclamation is going on, TSM will take a tape drive away from
> > reclamation. This is done by cancelling a reclamation process,
rather
> > than having a process go into mount point wait. TSM does not start a
> > replacement process when this happens. A restore that runs for a
> > couple
> > of minutes can leave a pair of tape drives sitting idle for hours.
Is
> > there any configuration setting or release level upgrade that will
> > cause
> > TSM to handle this situation more intelligently?
> >
> > <disclaimer.txt>
> This e-mail contains information which (a) may be PROPRIETARY IN
NATURE
> OR
> OTHERWISE PROTECTED BY LAW FROM DISCLOSURE, and (b) is intended only
> for the
> use of the addressee(s) named above. If you are not the addressee, or
> the
> person responsible for delivering this to the addressee(s), you are
> notified
> that reading, copying or distributing this e-mail is prohibited. If
you
> have
> received this e-mail in error, please contact the sender immediately.
This e-mail contains information which (a) may be PROPRIETARY IN NATURE OR
OTHERWISE PROTECTED BY LAW FROM DISCLOSURE, and (b) is intended only for the
use of the addressee(s) named above. If you are not the addressee, or the
person responsible for delivering this to the addressee(s), you are notified
that reading, copying or distributing this e-mail is prohibited. If you have
received this e-mail in error, please contact the sender immediately.
|
