MASH is definitely an area of the DFSMShsm documentation that we need to improve. We have this as a 'To Do'.
The primary intent of the MASH function is to improve DFSMShsm throughput. It can be shown that as the number of tasks within a single HSM address space increases that the throughput per task decreases. For example, it is more efficient to run 8 recall tasks across two HSM address spaces than 16 recall tasks on a single HSM address space. Multiple Address Space Support (MASH) enables multiple HSM address spaces to run within the same LPAR such that the level of HSM activity on a single LPAR can remain the same while distributed the work. Other benefits are those that have been discussed within this thread. For SHARE attendees, at the August 2008 conference I presented Session 3032 "DFSMS Basics: A DFSMShsm Configuration How To" which provides more details on setting up MASH. ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to [email protected] with the message: INFO IBM-MAIN
