What real-time monitors do
SMF and RMF record what happened; a real-time monitor shows what is happening now. Products such as IBM OMEGAMON, BMC AMI Ops and Broadcom SYSVIEW sample the system every few seconds, let you drill down from the whole system to one address space or transaction, and raise alerts when a condition is met.
Now, recent and history
| Time frame | Typical source | Used for |
|---|---|---|
| Now (seconds) | Real-time monitor screens | What is slow or stuck right now, and why |
| Recent (minutes to hours) | The monitor's short-term history, RMF Monitor III | What changed just before the problem started |
| History (days to years) | SMF records and RMF Postprocessor reports | Trends, capacity planning, chargeback, audit |
The products you will meet
- IBM OMEGAMON: a family with monitors for z/OS, CICS, Db2, IMS, MQ, networks, storage and Java. It has a classic 3270 interface, an enhanced 3270 user interface, and a portal and APIs for graphical dashboards.
- BMC AMI Ops (formerly MainView): monitors and automation for z/OS and its subsystems.
- Broadcom SYSVIEW: performance management for z/OS, CICS, Db2, MQ and more.
- RMF Monitor III: part of z/OS RMF, showing who is delayed and by what (processor, I/O, storage, enqueues).
Sites usually have one main monitoring family, and its name appears in job adverts. The ideas are the same in all of them, so skills transfer.
Drill down
Monitors are built for drill-down: start from a system summary (CPU busy, paging, the worst workloads), select the workload that is missing its goal, then the address space, then the transaction or thread, then the reason it is waiting. Each step narrows the question until the cause is clear.
Situations and alerts
A monitor checks conditions on a schedule, for example "CICS average response time above 1 second for 3 samples". OMEGAMON calls these situations. When one is true, the monitor can show it on a dashboard, send an event to automation or the enterprise event console, open a ticket or page someone.
Which part of z/OS RMF shows near real-time delay analysis? (for example Monitor I)
Show a hint
RMF has three monitors; it is the third.
Show the solution
RMF Monitor III.
Common mistakes
SMF is written at intervals and processed later. For now, use the real-time monitor or RMF Monitor III.
Start at the system summary and drill down; the cause is often in a different subsystem from the symptom.
An alert describes a symptom, such as a long response time. Drill down to find what it is waiting for.
What you will see at work
- Operations analysts often keep the monitor's system summary open on one screen all shift.
- Monitor data can also be streamed to enterprise observability platforms, covered in the Mainframe observability course.
- The Command Center's monitoring wall drill trains the same habit: read several graphs together and find the change that started it.
Key terms
Check your understanding.
Take this lesson's quiz and save your progress. Free.