Xpediter: a developer debugger priced on the LPARs it sits on.

Xpediter is Compuware's family of interactive, source level debuggers for mainframe applications, with editions for CICS, for TSO and IMS, and for Db2, covering COBOL, Assembler, PL/I, and C. Developers use it to step through code line by line, inspect data, and find defects in production grade programs, which makes it a daily tool for the teams that maintain core systems. BMC acquired Compuware in 2020 and has rebranded the product as BMC AMI DevX Code Debug, though the underlying debugger and the Xpediter name are still widely used in the field. Because it is embedded in everyday developer workflow, it is rarely removed, and that stickiness is exactly why its licensing deserves scrutiny.

Xpediter is licensed on capacity. BMC uses MSU, the IBM measure of processor capacity, as the unit of measurement, so the entitlement is built on the MSU rating of the machines or LPARs where the debugger is authorized, carried under a multi year agreement rather than a charge per developer seat or per debug session. The CICS, TSO, IMS, and Db2 editions are distinct, so which variants you are entitled to matters. BMC also offers zConsumption Licensing (zCL), a consumption based model that prices against prior year measured MSU utilization with a year end true up. Because the branding moved from Compuware Xpediter to BMC AMI DevX Code Debug, your schedules may carry either name.

Directional, pattern level. Confirm your own metric, editions, and authorized capacity against the contract schedules before modeling a renewal.

The dominant driver is the MSU capacity of the development and test LPARs where Xpediter is authorized. Because the price tracks partition capacity rather than developer count or debug activity, a debugger sitting on an oversized or shared LPAR can carry an entitlement far larger than the team that uses it would imply. The second driver is edition scope: the CICS, TSO, IMS, and Db2 debuggers are separate, so paying for editions you do not use is a real and common cost. The third is the BMC agreement structure and any uplift or escalator terms, which BMC can apply across a bundled Compuware and AMI portfolio in ways that obscure what Xpediter alone actually costs.

Xpediter exposure sits in the partition footprint and the edition set. Common traps we see at pattern level:

The entitlement is built on partition capacity and edition scope, so the levers work both, plus the timing. The five that pay:

Mainframe debugging has credible alternatives: IBM Debug for z/OS and other source level debuggers cover overlapping ground, and a shop can evaluate a switch. But a debugger is woven into daily developer habit and into established test procedures, so displacing it carries retraining cost and friction that often outweighs the licensing saving on a tool that is usually a modest line relative to the database and systems products around it. That makes a switching threat real but secondary. For most estates the more reliable wins are containing the debugger to right sized partitions, trimming unused editions, and pricing it inside the broader BMC negotiation.