Insight Packs for metric anomaly detection

For metric anomaly detection, IBM Z® Anomaly Analytics includes the various servers and insight packs.

Depending on where metric-based machine learning is installed, you must enable the appropriate Insight Packs in the following components:
  • In the Problem Insights server, so that it can provide visualizations of the subsystem-specific insights
    • The Metric-Based Anomalies Insight Pack must be installed. The Metric-Based Anomalies Insight Pack provides the subsystem scorecard dashboard.
    • If metric-based machine learning is installed on IBM® z/OS® UNIX System Services, one or more of the following insight packs must be installed depending on, which subsystems you are analyzing:
      • CICS® Transaction Server for z/OS Insight Pack R2, which provides subsystem scorecard for IBM CICS Transaction Server for z/OS.
      • Db2® for z/OS Insight Pack R2 Insight Pack R2, which provides subsystem scorecard for IBM Db2 for z/OS.
      • IMS Insight Pack R2, which provides subsystem scorecard for IMS.
      • MQ for Insight Pack R2, which provides subsystem scorecard for IBM MQ for z/OS.
      • z/OS Insight Pack R2, which provides subsystem scorecard for IBM z/OS
    • If metric-based machine learning is installed on Linux®, the Subsystem Scorecard for z/OS Insight Pack R2 must be installed. The Subsystem Scorecard for z/OS Insight Pack R2 provides subsystem scorecards for IBM CICS Transaction Server for z/OS, IBM Db2 for z/OS, IMS, IBM MQ for z/OS, and IBM z/OS.

    • On Linux, these insight packs are deployed as OCI-compliant. On IBM z/OS UNIX System Services, they are deployed as standalone Java™ applications except for the Subsystem Scorecard for z/OS Insight Pack R2, which is not supported on IBM z/OS UNIX System Services.

  • In the metric-based machine learning system on IBM z/OS UNIX System Services, so that it can provide subsystem-specific machine learning capabilities. If the metric-based machine learning system is installed on Linux, no insight packs are needed.
For information about installing or updating Insight Packs in the Problem Insights server, see the following topics:

Key performance indicators (KPIs)

Each Insight Pack for metric anomaly detection analyzes key performance indicators (KPIs) for one or more subsystems to which it applies. See the following topics for descriptions of the KPIs that are analyzed: