07-16-2012 12:06 PM
We just recently migrated a RAW DISK backup from an HPUX client to a Linux client. While the backup is running, we noticed memory utiliziation on the Linux client rise from about 68% to 100%. This 32% increase represents about an additional 100GB used during the backup. Nothing is failing and measurements are through GLANCE.
Details of the environment:
DP 6.20 patched as of Feb 2012.
Cell Manager is HPUX running 11iv2.
Client is RHEL 5.8 with 250 GB of memory.
Backup is RAWDISK using 2 to 3 tape drives, each drive is set to 12 concurrencies, 64K blocksize and DA buffer = 8.
We modified the DA buffer from 8 to 4 and then to 1. There was NO improvement in the memory util and no adverse affects on the backup timing.
Has anyone experieneced similiar situations?
07-22-2012 11:35 AM
Assign a kudo to this post, if you find it useful.