z/OS Work Observability
Created , Updated
z/IRIS creates z/OS Work traces for APM tools using SMF record type 30 records. z/OS Work traces provide visibility into critical batch jobs, processes, transactions and users that may impact business applications and services. Tags/attributes provide further information about the performance, the consumption of system resources and identifies any associated users and systems.
Trace design
z/OS Work initiated parent span
The parent span identifies the work and the start time of the unit of work that has begun processing on z/OS. The following types of work are supported:
TSO/E session
APPC/MVS transaction program
OMVS forked or spawned address space
Started task
Batch job
Batch Job steps completed child span(s)
These spans are only provided for Batch Jobs and describes the duration of the batch job step, the programs or utilities executed and resources consumed.
If an error occurs it will be indicated by the span status and an error message will provide the MVS completion code with any available operating system messages.
z/OS Work completed child span
This span indicates that the z/OS Work processing has ended and details the total duration of the unit of work. Additionally, trace attributes provide:
Batch job and step identifiers
Associated user identifiers
Relevant z/OS system
Duration of the step (i.e., the amount of time between the start and end of the step’s processing)
Performance (e.g., processor consumption)
Resource usage (e.g., storage consumption)
z/IRIS signals an error for a non-zero completion code and may provide error information using APM vendor events or using OpenTelemetry’s APIs to provide error descriptions in traces.
TSO Logon trace
Identifies the time at which a TSO user logged onto a TSO session on z/OS.
TSO Logoff trace
Identifies the time at which a TSO user ended their session and the duration of that session, as well as any resources consumed during the users' session.
z/OS Work Batch Job Workflow Tracing
Using the z/IRIS Mainframe Workflow Tracing feature users can see the breakdown of a z/OS Work batch job. This allows DevOps teams to cut through software complexity using observability to locate issues. Finding the origin of an issue becomes easier as z/IRIS highlights the impact of individual z/OS Work batch job steps.
