① Guide · Sub-capacity and SCRT
Your monthly Sub-Capacity Report decides what IBM bills you, and most of it is read by people who never ask whether the number should be lower. This is the buyer's read: where the billable MSU figure comes from, which product line is really driving the cost, and the levers sitting inside a report you already generate.
You send IBM a number every month. The technical team reads it for accuracy. Almost nobody reads it for leverage.
The Sub-Capacity Reporting Tool, SCRT, is a no charge IBM tool that processes your SMF and SCRT89 records and produces a monthly report of the peak rolling four hour average MSU consumption for each sub-capacity eligible product. You submit it to IBM through License Management Support, and IBM bills your monthly license charge products against it. The tool is accurate. The question buyers fail to ask is whether the inputs and the workload behind them are producing a number higher than they need to pay.
Reading the report like a buyer means treating each line as a cost you can interrogate, not a fact you report. The same file the operations team files for compliance is the file that tells you which product is carrying your bill, which LPAR peak is setting it, and where a defined capacity limit or a moved batch window would lower it. For the metric mechanics see the MIPS to MSU conversion question and batch window tuning to cut R4HA peaks.
Every billable figure traces to a single four hour window where consumption peaked. You are not billed on what the machine averages or what it can do at the top. Find the peak interval and you have found the one slice of the month that sets the whole charge.
Each product is reported at the peak of the LPARs it runs in, summed across the sysplex where the rules apply. A product that happens to be active during another workload's peak inherits that peak. This is why workload placement, not just workload size, drives the number.
If an LPAR runs under a defined capacity or a group capacity limit, SCRT honors it and your billable MSUs cannot exceed it. A limit set too high, or never set at all, leaves money on the table every month. This is the single most overlooked lever in the report.
Whether a product is reported as sub-capacity eligible, and how it maps to your entitlement, decides whether it counts at the LPAR peak or the full machine. Misclassification, common after migrations and acquisitions, quietly inflates the bill. Check the product list against what you actually license.
One report is a snapshot. Twelve are a negotiation exhibit. The trend shows whether peaks are creeping, whether a capping change worked, and what baseline you should carry into a renewal or a Tailored Fit Pricing conversation. Keep the series; it is your evidence.
| What the report shows | What it means for the bill | The buyer lever |
|---|---|---|
| Peak R4HA 1,200 MSU | Set by one four hour window | Identify and reshape that window |
| Product active in peak LPAR | Inherits the LPAR peak | Move or isolate the workload |
| No defined capacity limit | Billed to the full peak | Set a soft cap at the real need |
| Product on full capacity | Possible misclassification | Confirm sub-capacity eligibility |
| Peak rising 4% per quarter | Renewal baseline drifting up | Arrest the trend before renewal |
Figures above are illustrative of how a report reads, not a benchmark. SCRT behavior reflects current IBM tooling as commonly observed; your contracts and z/OS configuration govern the actual result.
SCRT reports the peak rolling four hour average MSU consumption of each sub-capacity eligible product over a reporting month, drawn from SMF and SCRT89 records. It is the number IBM bills monthly license charge products against, submitted through License Management Support.
Each product is billed on the LPAR peak it runs in, not the system average. A single product active during the four hour peak can carry a high MSU figure even when its real workload is modest, which is exactly where workload placement and capping become commercial levers.
The tool is accurate to its inputs, but the inputs and interpretation are where buyers lose money: misclassified products, defined capacity set too high, peaks driven by batch that could be moved. The report is the start of the analysis, not the verdict.
Audit notice or renewal under 18 months out? We mobilize within 48 hours. We read your SCRT series for leverage before the vendor reads it for revenue.
The report you file for compliance is the report that holds your leverage.
Related guides: reconciling MIPS and MSU against contracts, defending your SCRT position, and IBM MLC cost reduction without workload change. Explainers: batch window tuning to cut R4HA peaks and the MIPS to MSU conversion question. Commercial: mainframe cost optimization.