The wonders of automated workload balancing
You may be familiar with the delays that occur when CPU utilization approaches 100%. While extended wait times are a common problem, there is a way to effectively minimize them without manual intervention.
The secret? Automated workload balancing. By balancing batch workload intelligently, a data center can benefit from increased throughput. This is accomplished thanks to the proper use of existing resources.
However, balancing in this context does not mean that an equal number of jobs are running on each system. Rather, each system in the JESplex runs the right amount of workload for the current conditions on each system. Furthermore, balancing means respecting any specific system affinity and resource requirements for individual batch jobs.
Automated workload balancing
ThruPut Manager specializes in automated balancing of batch workloads. It does so effectively by taking into consideration the performance and capacity of the entire environment. ThruPut Manager tracks the utilization of all LPARs and the required system and resource affinities of all batch workloads. It also requires that all installation-specified business priorities are defined so that it always selects the most urgent job.
The result is a system that can respond to rapidly-changing environments quickly, without increasing wait times. ThruPut Manager achieves this by considering the actual activity on each LPAR and re-evaluating CPC, LPAR and Service Class performance every 10 seconds. It also avoids overloading by responding to sudden changes in capacity (due to Capacity on Demand, LPAR weight changes, and soft capping), and rebalancing the batch workload as CPU demand and availability change.
Only automation can do this—and it’s quite an amazing feat. If busy restaurants implemented the same type of automated system, you would indeed be able to have your cake and eat it too—in an efficient amount of time.