RMF
The Resource Management Facility (RMF™) collects system-wide data that describes the processor activity (WAIT time), I/O activity (channel and device usage), main storage activity (demand and swap paging statistics) and system resources manager (SRM) activity (workload).
RMF™ is a centralized measurement tool that monitors system activity to collect performance and capacity planning data. The analysis of RMF™ reports provides the basis for tuning the system to user requirements. They can also track resource usage.
- Processor usage
- Address space usage
- Channel activity:
- Request rate and service time per physical channel
- Logical-to-physical channel relationships
- Logical channel queue depths and reasons for queuing.
- Device activity and contention for the following devices:
- Unit record
- Graphics
- Direct access storage
- Communication equipment
- Magnetic tapes
- Character readers.
- Detailed system paging
- Detailed system workload
- Page and swap data set
- Enqueue.
- Evaluate system responsiveness:
- Identify bottlenecks
- The detailed paging report associated with the page and swap data set activity can give a good picture of the behavior of a virtual storage environment.
- Check the effects of tuning:
- Results can be observed dynamically on a screen or by postprocessing facilities.
- Perform capacity planning evaluation:
- The workload activity reports include the interval service broken down by key elements such as processor, input/output, and main storage service.
- Analysis of the resource monitor output (for example, system contention indicators, swap-out broken down by category, average ready users per domain) helps in understanding user environments and forecasting trends.
- The postprocessing capabilities make the analysis of peak load periods and trend analysis easier.
- Manage the larger workloads and increased resources that z/OS® can support.
- Identify and measure the usage of online channel paths.
- Optimize the usefulness of expanded storage capability.
- Monitor I measures and reports the use of system resources (that is, the processor, I/O devices, storage, and data sets on which a job can enqueue during its execution). It runs in the background and measures data over a time period. Reports can be printed immediately after the end of the measurement interval, or the data can be stored in SMF records and printed later with the RMF™ post-processor. The RMF™ post-processor can be used to generate reports for exceptions: conditions where user-specified values are exceeded.
- Monitor II, like Monitor I, measures and reports the use of system resources. It runs in the background under TSO or on a console. It provides snapshot reports about resource usage, and also allows data to be stored in SMF records. The RMF™ post-processor can be used to generate exception reports.
- Monitor III primarily measures the contention for system resources and the delay of jobs that such contention causes. It collects and reports the data in real time at a display station, with optional printed copy backup of individual displays. Monitor III can also provide exception reports, but its data cannot be stored in SMF records.
- The reports are written at the interval requested
- Other work is not delayed because of locks held by RMF™.
A report is generated at the time interval specified by the installation. The largest system overhead of RMF™ occurs during the report generation: the shorter the interval between reports, the larger the burden on the system. An interval of 60 minutes is recommended for normal operation. When you are addressing a specific problem, reduce the time interval to 10 or 15 minutes. The RMF™ records can be directed to the SMF data sets with the NOREPORT and RECORD options; the report overhead is not incurred, and the SMF records can be formatted later.