Mainframe Performance Management: Still Important!
Last year I realized I had to update my bio to read that I have over three decades of experience in the mainframe environment, most of that dealing in some way with performance. Where does the time go?
There have been huge changes in mainframe performance over the past 30 years: processor, memory and storage are now orders of magnitude larger in terms of both capacity and performance. However, performance management still is an important part of a well-run mainframe environment because we still have performance problems to deal with.
We’re still working to squeeze performance out of available hardware capacity. But capacity issues are usually not because of physical hardware constraints but rather due to financial concerns that prevent adding capacity. Other performance problems emerge from today’s more complicated interconnected environment or from newly introduced technologies whose implementation best practices may not be well understood.
Meanwhile, business demands have increased as well, and today poor performance is even less tolerable than in the past.
Today’s performance analysts need usable and useful performance reporting that allows them to quickly investigate a problem, find the cause and determine how to prevent future similar problems.
Usable performance reporting is immediately at hand without having to figure out where the raw data resides or the arcane batch job that needs to be run to access it.
Useful performance reporting transforms performance data into performance information that helps the analyst understand the context and meaning and draw conclusions.
Managing performance involves regular performance reporting, and somebody should be periodically reviewing it, even in an absence of an obvious issue. Over my career, there have been many times where chasing an interesting “rabbit trail” led to changes to mitigate potential future problems. And staying on top of small performance issues can help avoid large ones in the future.
Organizations that have let performance management slip by the wayside are increasing their business risk because when a problem does occur, the analysis and resolution will take longer than it should. Mitigating that risk involves investing in tools and staff training to make performance analysts effective and giving them the time to regularly evaluate system and application performance.
Is your organization managing performance or merely hoping for it?
Scott Chapman is the CIO for Enterprise Performance Strategies, Inc and director of Software Design and Development for Pivotor z/OS performance reporting.
See more by Scott Chapman
About Enterprise Performance Strategies, Inc.
Enterprise Performance Strategies, led by Peter Enrico, has assisted hundreds of installations with their z/OS performance needs. Learn more about our solutions →