Processing external resources
A ThruPut Manager AE best practice is to use the time a job is awaiting selection in the queue, normally considered wasted time, to prepare the job to run. In effect, TM AE shifts the preparation of external resources from execution (initiator time) to queue time. Specifically, these activities include:
- Issuing HSM recalls and prioritizing them based on job importance. TM AE sets a maximum of 40 recalls per job, and avoids unnecessary recalls (IEFBR14’s).
- Checking virtual volumes for residency in cache, and staging them by importance, if necessary.
These requests are carried out in parallel rather than serially. Resources for a job are ready in less time than before, due to these parallel actions. Your jobs might spend a slightly longer time in the queue, but that will be more than compensated for by their decreased time in execution.
The overall effect is that your jobs spend less time in the system and overall throughput increases. In whatever way you choose to define “service”, this is clearly “better service”.
ThruPut Manager AE tracks the delay times involved in all of these external requests and makes them available for downstream reporting.