Accessing mass data under MVS

Our mass data already resides on MVS, see , so the access is local. Datasets can reside on a pool of disks or in a pool of tape cartridges. Data on cartridge has to be staged to disk before it can be used. This is managed automatically by the system, but may take some time. For reasons of performance RZ-files are created as so-called "VSAM clusters". They are made of two components, a control component which receives the name you chose and a data component with an appended ".DATA" qualifier.

PAW under MVS uses a special notation (a leading '.') different from the MVS notation ('<dataset-name>') to access another user's datasets:

raw/i/file .kp09.lmdv.s018.raw0815              # accessing raw data on the FOPI account
dst/scan   .kp09.dstx.s018.dst0815          ! X # overview  of DST data 
dst/i            dstx.s018.dst0815          ! X # DST input, your account
dst/o            dstx.s018.dst0815 <run-no> ! X # DST output,your account

Scanning a large amount of data requires a lot of valuable disk space as well as HSM recalling activities. The following scheme shows an example to accomplish this task effectively.

RA/I raw0815              # open a raw file
SHELL HMIG raw0814 NOWAIT # migrate the previous one
RA/A 100000               # in the mean time, go on with the analysis
RA/I raw0816              # open a raw file
SHELL HMIG raw0815 NOWAIT # migrate the previous one
RA/A 100000               # in the mean time, go on with the analysis