It’s Time to Become More Agile with Batch Performance Management
Agile is largely replacing the waterfall development approach mainframers are familiar with. In the past, large, multi-year development projects gave performance analysts time for loads of testing. You had ample time to plan your tuning efforts, and as production installs were few and far between, for the most part, you didn’t have to worry about code changes until the next waterfall.
Agile is a nimbler approach, delivering smaller, more frequent software releases. Agile is more responsive to business needs and ensures companies can maintain their competitive edge, delivering on customer demands. And Agile doesn’t just affect developers—it affects performance analysts too.
A few of the Agile principles that affect the mainframe performance analysts include:
- Satisfying customers through early and continuous delivery of features
- Frequent requirement changes throughout the process
- Delivering software frequently
Agile Batch Performance Management
When software changes were infrequent with waterfall development, you could apply the “tune one and done” approach. Batch would run fine until the next install and you’d have plenty of time to test and evaluate the new software.
That doesn’t work for Agile. Change is a constant, which means you need to be on top of batch performance at all times. And it isn’t just changes to mainframe batch applications you need to consider. Changes to online work can affect resource utilization, causing batch to wait. In addition, changes to work on distributed platforms (UNIX, Linux, Windows) can cause changes to batch due to file size or increased file activity. You probably won’t have any visibility into these changes.
You need to engage in continuous improvement, but it doesn’t have to be a manual process. In fact, using manual processes, you couldn’t possibly keep up, no matter how hard you worked. No one can work 24/7 anticipating the changes that impact daily, weekend, night and quarter-end work. And let’s be real; most performance analysts barely have the time to do the work necessary to keep online availability and performance within SLAs.
To move faster and keep up with your developer colleagues who are accelerating mainframe application development and delivery, you need to leverage automation through Agile tools for performance management.
Automated Performance Management Tools
Through Compuware Strobe, AutoStrobe can help with your critical batch jobs by automating the measurement request process instead of requiring you to select jobs and manually request a run. Since AutoStrobe works with various third-party monitors and/or a custom application through an API, you can start a Strobe measurement when these products detect a problem based on the criteria you establish.
AutoStrobe can also help with batch through Global Batch Monitoring, which is designed to find performance problems by looking at CPU rolling averages and job runtimes. Using data from previous runs, Global Batch Monitoring sets thresholds and reports on any job that exceeds these levels.
Automatic detection of problems is a terrific value, but you want more than that. Compuware ThruPut Manager is designed to optimize your batch workloads by balancing them and speeding batch throughput. Smarter prioritization, elimination of dataset contention and assurance that a job won’t be released until resources are available all ensure batch jobs will run better than any manual effort could achieve. In addition, through this tuning, ThruPut Manager delivers significant resource and software savings.
Don’t let the speed of change cause you to struggle with batch performance. Automation is the solution to overcoming the challenge of achieving agile batch performance management.
Latest posts by Denise Kalm (see all)
- How Developers Can Boost DevOps Through Automated Batch Processing - December 7, 2017
- How to Control Your CPU Demand and Costs at the Job Level - November 16, 2017
- Why ThruPut Manager Is the Real-time Cost-cutting Robot for You - November 9, 2017