Written by Brent Phillips on April 30, 2018.
The z/OS platform excels at efficiently executing transactions and is a critically important component for many business applications. Yet, application development teams rarely have an understanding of how application program changes impact the performance and cost efficiency of the underlying z/OS infrastructure.
Nearly all z/OS performance and capacity experts have stories about an application being launched into production and causing major service level delivery problems and/or unexpected z/OS infrastructure costs. These production problems happened despite existing DevOps test procedures for new application versions.
Last year IBM announced that z/OS mainframe clients could triple the capacity of their development environment with no increase in MLC. Being able to afford a properly sized Dev/Test environment is extremely valuable, but it is not enough to prevent these performance and cost issues.
Our past webinar, Easy z/OS Application Performance Testing Services from the Infrastructure Perspective, discussed an easy way to see variances in the “application performance signature” from an infrastructure perspective.
The webinar covered:
- Application Infrastructure Performance Signatures
- How to Automate the Calculation of Variances between Releases
- How to Increase Collaboration between DevOps areas
- How to Achieve Fewer Unexpected Infrastructure Cost Increases
- How to Have Fewer Production Performance Problems
In the webinar, we examined the historical reasons why the performance and financial cost impact on the z/OS infrastructure has been difficult to test in DevOps organizations and how utilizing a modernized approach with artificial intelligence can significantly improve release quality as it relates to the z/OS infrastructure’s ability to efficiently deliver the service levels required by specific applications.
Recommended Webinar
For insights on ensuring secure and efficient performance testing in z/OS environments, watch: zAcademy #7: Value of and Visibility into zERT Network Encryption Data