A co-worker of mine worked with IBM/Tivoli on this. The symptoms came during
the reclamation of the offsite copypool tapes for the DIRMC disk storage
pool. The answer he got was that reading from the disk storage pool was done
a file at a time and not buffered. They said it was because it was a random
access device and they had no plans to 'fix' it.
What he came up with was to create a sequential primary pool with a FILE
device class and make this the NEXTSTGpool for the DIRMC disk pool and
migrate daily. The reclamation and BA STG are alot faster. So what we do on
our server configurations:
DIRPOOL (disk) -> DIRFILE (file) then backup to offsite DIRCOPYPOOL (tape).
Our daily backup process is:
BA STG DIRPOOL DIRCOPYPOOL
BA STG DIRFILE DIRCOPYPOOL
UPD STG DIRPOOL HI=0 LO=0
The reclamation on this pool runs in about an hour (DIRFILE @450MB) instead
of a day. This is on OS/390 TSM 3.7.3.
Bill Boyer
DSS, Inc.
"I haven't lost my mind -- it's backed up on tape somewhere." - ???
-----Original Message-----
From: ADSM: Dist Stor Manager [mailto:[EMAIL PROTECTED]]On Behalf Of
nwindric
Sent: Friday, October 13, 2000 4:31 AM
To: [EMAIL PROTECTED]
Subject: long time move data for dirmc
Hi,
we made a move data of a dirmc copy pool. The move data process
needs 5:12 hours to move 70 823 172 bytes. Most of the data is read
from a disk storage pool. For database files we can move up to 10 GB
per hour.
How can we speed up the move data process for the dirmc pool?
Regards
Norbert Windrich
