Our new z/OS Infrastructure dashboard is also available in our Datadog offering, collecting the same metrics as in Grafana, keeping with our goal to provide users with the equivalent mainframe-inclusive observability value, regardless of which APM tool is used.
Join our mailing list and stay up to date!
Roadmap excerpt for IBM MQ and CICS TS
The team’s first goal of 2022 is to provide MQ-CICS observability support to mainframe-backed organizations, which will enable enterprise monitoring teams to:
- identify resources utilization and constraints that impact the performance of MQ and CICS on z/OS.
- troubleshoot MQ messages to CICS transactions that report high latency or errors.
- integrate key performance indicators for MQ and CICS into AI-Ops software to enable proactive workflows that prevent downtime and customer dissatisfaction.
Currently, we are working on transforming MQ statistics records into metrics. Next up will be converting MQ accounting records into APM traces, followed by the transformation of CICS SMF records into metrics and APM traces.
New Grafana and Datadog Dashboards will also be introduced that highlight the key measurements of performance using z/IRIS MQ and CICS metrics.
Improved user experience
z/IRIS enriches mainframe traces with attributes to facilitate searching, filtering and analyzing traces in APM tools. Overarching attributes have been selected to be part of the common span attributes to permit a consistent and predictable user experience. Common attributes are present in every z/IRIS span and describe the relevant mainframe infrastructure, host (a.k.a. LPAR) and subsystem.
Additionally, some of these attributes are used by APM products to draw correlations between mainframe hosts and the dependent requests/transactions, resulting in enhanced service to infrastructure mapping and alerting capabilities.
Db2 Accounting telemetry
Db2 traces have also been enhanced. Historically, parent spans were influenced when the duration of the Db2 Accounting child span was set. To avoid impacting APM users, the duration of a JDBC transaction was provided in Db2 Accounting tag: zos.db2.elapsed.time_ms. Changes to the APM integration have resolved the parent span impact, permitting the duration of JDBC transactions to be set at the Db2 Accounting span level. JDBC call duration is calculated via:
zos.db2.end.timestamp – zos.db2.start.timestamp (QWACESC-QWACBSC)
As a result, the duration was added to the call hierarchy. The parent and child spans are displayed with the appropriate lengths in the flame graphs
See our example representation in Instana.
New z/OS Infrastructure Dashboard
z/IRIS RMF metrics provide data about your z/OS resource usage that help you identify possible contention or constraints which impact service-level objectives (SLOs), such as response times or application availability.
The z/OS Infrastructure dashboard, visualizes our RMF Metrics to give you a high-level view and understanding of your mainframe performance and resource utilization and track your CPU contention and mean CPU usage by LPAR and MVS.
Additionally, the Ready-Work-Unit-Queue Distribution depicts how much work is waiting for processors to become available. The system contention rises as the amount of work waiting increases and is an important signal that increased processing capacity is required to meet SLOs.
If you want further information about how RMF metrics have been overhauled, take a look at the RMF metrics streaming documentation.
Datadog z/OS Infrastructure dashboard
Grafana z/OS Infrastructure Dashboard
Grafana 8 support
z/IRIS z/OS Connect Dashboards
z/OS Connect dashboards are now available for Grafana 8. Enterprise Performance Management teams and DevOps engineers can monitor and analyze how z/OS Connect servers and mainframe-based service providers perform for REST API requests by monitoring load, latencies and error rates which correlate to the availability and reliability of business applications.