How to recover DB2 DPF database with all nodes simultaneously ?

Hello,

we have one Linux client (and S/N) which has DB2 multipartition DB, DPF. This client is backed up through all nodes at the same time - NODE0000 to NODE0008 and the backup is really fast on DataDomain with all nodes in parallel.

Restore is also on one Linux client over NW server which has access to DD devices.

When set nmda_restore.cfg file and start db2 command as it is stated in NMDA admin guide - the client restores only NODE0000 and the stops. All other nodes have to be restored manually using db2_all <<+... command as in old NMDB module (i.e. NODE0007 restore is done via db2_all "<<+7< db2 restore db DWPROHSE load /usr/lib/libnsrdb2.so options @nsr/apps/config/nmda_DWPROHSE_restore.cfg") which lasts 2-3 times more than with all nodes in parallel.

Does anyone know what could be the cause?

This is nmda restore file:

NSR_SERVER=i560004b
DB2INSTANCE=db2inst1
DB2_NODE_NAME=db2inst1
NSR_CLIENT=i560c891
DB2_ALIAS = DWPROHSE
NSR_DATA_VOLUME_POOL=dd860HSE
DB2_PARTITION_LIST = all
DB2_VENDOR_LIB_PATH = /usr/lib/libnsrdb2.so
NSR_DEBUG_LEVEL=1
USER_PSWD="$1$KPHXKFNZPCHCNZPN$0077467348cd4ef1845c129d8ffb880b"

and DB2 restore command:

db2 restore db DWPROHSE load /usr/lib/libnsrdb2.so options @nsr/apps/config/nmda_DWPROHSE_restore.cfg

NW server is 8.1.0.5, Source SN/Client is on 8.1.0.5 and target client is on 8.1.0.1 version of software.

Please, if someone has same/similar issues and how we can resolve it!

Thanks,

Andrej