White Box Analysis Vs. Black Box Statistical Performance Analysis

This short video discusses the differences between white box analysis and black box analysis for RMF/SMF performance management and is excerpted from a webinar. Watch the full webinar, “RMF/SMF Analysis – Get Ahead of the Curve with Statistical Analysis” here.

Many new z/OS performance analytics solutions talk about black box analysis for their statistical approaches. This offers advantages such as not needing prior expert knowledge about the mainframe infrastructure as the black box statistical approach is platform agnostic.

However, white box analysis utilizes built-in expert knowledge about z/OS metrics and the z/OS infrastructure to detect absolute issue violations and derive true intelligence for predictive and prescriptive insights.

Black box statistical approaches often trigger alerts after significant changes to key application workloads. Again, this can be useful, but the alerts are both reactive, and do not consider whether the infrastructure can properly service the new workload pattern.

White box analysis, or Availability Intelligence, issues alerts only when the underlying infrastructure is at risk of not delivering workload needs. This means the alerts are proactive and will not send alerts about a doubled workload if the infrastructure can handle it.

Oftentimes both black box and white box approaches are necessary for a complete z/os performance management solution. Watch the short video above to learn more.