How to parallelize a GPFS policy driven migration process on a single TSM client node?

tivolimistery

ADSM.ORG Member
Joined
Oct 10, 2009
Messages
58
Reaction score
0
Points
0
PREDATAR Control23

Dear Users,

I would like to increase the TSM server mount points used (I'm referring to the number of drives/tapes used) during a GPFS policy driven migration.
Some basic information about software version before go ahead:

TSM server v.5.5
TSM client v.6.4
GPFS v.3.5

The hsm-managed filesystem name is "/archive"

At this moment, the data migration starts automatically respecting the following work flow:

1. GPFS I/O server (NSD server) detect the lowDiskSpace event and launch a call-back script;
2. the call-back script launchs the mmapplypolicy command against one of the TSM client node (tsm1-ib) as follows:

/usr/lpp/mmfs/bin/mmdsh -v -N tsm1-ib "/usr/lpp/mmfs/bin/mmapplypolicy /archive -B 3000 -I yes --single-instance"

3. on the target node (tsm1-ib) I can see the migration process running correctly:

/usr/bin/dsmmigrate -detail -premigrate -filelist=/tmp/mmPolicy.ix.19893.66EE428F.2.hsm

My question is: is there a way to parallelize, on the same target node (tsm1-ib) the migration process increasing also the number of tapes and drives used by TSM server?

I need of this improvement in order to make the migration process faster.

Thank you in advance,
Mauro
 
Top