Step 3: For a colocated solution, WLM, RMF and SMF track and record solution usage data
Once the WLM service definition is activated with tenant resource group definitions and associated classification rules, WLM tracks the processor consumption on behalf of the tenant resource group-specified workload. RMF then writes each tenant resource group's consumption data to SMF Type 70, Subtype 1 records every recording interval. SCRT analyzes the recorded data and produces a report for consumption by IBM's fulfillment systems, removing the workload's MSU utilization from the R4HA.
- By z/OS system programmers or performance analysts to monitor workload resource usage and performance
- For other metering purposes, such as chargeback related to the specific workload.
The system programmer can instruct WLM to limit (cap) processor consumption related to that workload, specified in the tenant resource group definitions. Similarly, the cap can be removed (or not specified), allowing processor consumption for that workload to continue unrestrained based on expected workload consumption, without the workload behavior modification that occurs when a CPU cap is reached.
Tailored Fit Pricing for IBM Z generally requires WLM work qualifiers that correspond to address space or independent enclaves. For specific details on qualification rules, see the terms and conditions of the Tailored Fit Pricing Solution.
Tailored Fit Pricing
Many IBM subsystems and sub-capacity products record their own CPU usage directly to SMF using IFAUSAGE or IFAEDREG, which is written as SMF Type 89 Subtypes 1 and 2 records, respectively. In both Type 70 and Type 89 records, data is written in tenant resource group segments, defining the relevant data for each tenant resource group. The example in Figure 2 shows the SMF data recorded on behalf of CICS regions and other information.
