MQ appliance vs software licensing.

IBM MQ is the messaging backbone that moves transactions between systems, and it is one of the high-risk products that draw audits. Part of what makes MQ tricky under audit is that it is not a single licensable thing. The software queue manager that you install on your own servers and the MQ Appliance that IBM ships as physical hardware are different products with different metrics, and an estate that runs both has to keep two distinct accounting models straight. When they blur together, the audit reconciliation surfaces gaps that were never intended.

MQ software: counted by cores

The software edition of MQ is a processor based product. It is licensed in value units tied to the cores allocated to the software on each server where a queue manager runs, the same core based model that governs WebSphere and Db2. That means the same disciplines apply: every server running a queue manager has to be in the count, the value units per core have to match the processor type, and where MQ runs virtualized on eligible technology, sub-capacity licensing can limit the count to the virtual cores actually allocated rather than the full physical host. Sub-capacity only holds with the approved measurement tool deployed within ninety days, running continuously, and the quarterly reports retained. Without it, IBM defaults to full capacity.

MQ Appliance: licensed as hardware

The MQ Appliance is a purpose built hardware unit that runs MQ in a sealed form. Because it is a device rather than software you install on a general purpose server, it is licensed and supported on a per appliance basis rather than by counting cores underneath it. The appliance does not draw a processor core count the way the software does, and it should not be reconciled against a core based entitlement. Its entitlement and support are tied to the unit itself. Treating an appliance as if it were a software install, or a software install as if it were an appliance, produces a mismatch the audit will flag.

Where the two models collide

• Counting an MQ Appliance against a processor based entitlement, or the reverse
• Software queue managers on additional servers that were never added to the core count
• Sub-capacity claimed on software MQ without the measurement evidence to support it
• Development and test queue managers assumed to be free when they require entitlement
• Appliances retired or redeployed without the records following them

How we defend it

The defense separates the two models cleanly. We inventory every software queue manager and count cores against the correct value units, holding the sub-capacity position where the measurement evidence supports it, and we account for each MQ Appliance as a device against its own entitlement. Where IBM has counted an appliance under the software metric or applied full capacity to software that qualified for sub-capacity, we correct it with your own records. The result is an MQ position where each form is measured by its own rules rather than collapsed into one inflated number.