The team is proud to introduce z/IRIS IronTap Cluster (Mode) for enterprise ready mainframe-inclusive observability. IronTap nodes now process telemetry in unison to provide Mainframe Workflow Tracing at scale.
Benefits of z/IRIS Mainframe Workflow Tracing
Mainframe Workflow Tracing, released in version 1.10.0, was an essential improvement to the powerful correlation capabilities provided by z/IRIS:
- Since version 1.0.0 – z/IRIS correlates and appends traces from mainframe to distributed applications to provide mainframe-inclusive observability.
- Since version 1.10.0 – Mainframe Workflow Tracing was introduced, which adds traces from downstream mainframe calls using a single IronTap server configuration.
- Now with 1.11.0 – Mainframe Workflow tracing is available using multiple IronTap nodes configured as a cluster.
z/IRIS IronTap cluster
Scaling the z/IRIS architecture is crucial to large enterprises that require higher availability and more resiliency from their observability solutions. That is why Mainframe Workflow Tracing is now supported in environments that require multiple IronTap servers for SMF processing and mainframe telemetry streaming.
An IronTap cluster is composed of 2 or more IronTap servers (i.e. nodes) configured to share information. As new APM spans for mainframe applications are processed, each IronTap node recognizes when it is processing a span that has a logical relationship to another mainframe span. The node will extract the required information from the cluster’s shared memory and add it to this span so that the association will be depicted on the target observability platform.
z/OS batch job traces with Mainframe Workflow Tracing
Using the Mainframe Workflow Tracing architecture, the team has improved z/OS Work tracing to provide users with a complete flame graph for batch jobs running on mainframe. Each span represents the various phases and steps executed in the batch job.
- Root Span – A batch job has started
- Child Spans – Each step executed in the batch job
- Final Span – The batch job has ended
Unifying z/OS batch job spans into a single trace tree not only establishes latency for each job step, but also simplifies analysis of job execution. Errors and events will now have context to the batch job rather than just the singular step. This improves impact analysis and facilitates alert configuration while boosting AIOps ingestion and learning.
Clock skew correction for testing
In sandbox and testing environments, time synchronization between systems is not always readily available. The system clock differences do not necessarily impact application integration, but can become evident on observability platforms where time-based information is collected from distributed and mainframe systems and displayed in a single pane of glass.

During z/IRIS testing, clock skewing can negatively impact the ways users perceive end-to-end workflows and complicates analysis of latency distribution. z/IRIS provides users with the ability to configure a clock skew correction so that IronTap can simulate synchronized system times by adjusting timestamps from mainframe spans so that the end-to-end traces are well aligned and easier to analyze.
