This slowness can happen when reclamation is done from primary storage pool
tapes also. Assume content of the offsite tape is scattered over multiple
on site tapes and tape drive has to seek to multiple location on tape to
pick up required piece of data can add up to significant tape
loading/unloading and tape seek operation. So it is not unusual for an old
offsite tape  to take long time for reclamation. You should actually allow
the reclamation process to finish. If you feel that reclamation process is
keeping your tape drives engaged for long, then you can stop reclamation of
offsite tapes alltogather and allow the offsite tapes to become empty due
to natural expiration process.

Samiran Das




                    "Rushforth, Tim"
                    <[EMAIL PROTECTED]        To:     [EMAIL PROTECTED]
                    PEG.MB.CA>                  cc:
                    Sent by: "ADSM: Dist        Subject:     Re: Reclamation is not 
happening -
                    Stor Manager"               HELP!
                    <[EMAIL PROTECTED]
                    U>


                    05/22/2002 01:13 AM
                    Please respond to
                    "ADSM: Dist Stor
                    Manager"





Offsite reclamation runs very slow when processing input data from disk.
(see http://msgs.adsm.org/cgi-bin/get/adsm0203/501.html)

Also with offsite reclamation, TSM will process all the eligible offsite
tapes at once (it does not process one offsite tape then the next).

Have you tried a higher value, eg how many tapes are eligible over 95%, try
this value etc.

Tim Rushforth
City of Winnipeg
-----Original Message-----
From: Coats, Jack [mailto:[EMAIL PROTECTED]]
Sent: Tuesday, May 21, 2002 2:25 PM
To: [EMAIL PROTECTED]
Subject: Re: Reclamation is not happening - HELP!


Currently, I have done the following:
  Allowed expiration to complete fully.
  Started reclamation again.
  It is running now but very slowly
        In the last 4+ hours, moved 55K files and 1.45G of data (from q
proc)
        or if I do a
        "query volume * access=readw,reado status=full,filling
stgpool=copypool"
        it says my LTO is 17% full.

  The machine does nothing but TSM, and is not under a load from I/O,
networking,
virtual memory, storage, or processor as far as I can tell.

  In any case it isn't filling quickly at all.
Suggestions?

PS... Thanks for all the comments & suggestions sofar! ... Jack

Reply via email to