IBM WebSphere licensing and audit defense.

WebSphere is one of the products that most often draws an IBM audit, alongside Db2, Cognos, MQ, Maximo, and Tivoli. The reason is structural. WebSphere runs on shared virtual infrastructure, it is easy to deploy widely, and its license depends on a core count that is only defensible with continuous evidence. That combination, high deployment freedom plus an evidence-dependent metric, is exactly the profile IBM's audit motion is built to monetize.

The core metric is the processor value unit. PVU is a core-based measure: the number of cores allocated to WebSphere multiplied by a per core PVU rating, where an Intel core is commonly rated around seventy PVU. Without sub-capacity, that count is taken against the entire physical host. With sub-capacity, it is taken against only the virtual cores WebSphere actually uses, but only when ILMT proves it. The whole audit turns on which of those two counts applies.

The number that defines the risk

The standard illustration makes the stakes concrete. WebSphere using four of a host's thirty two cores is roughly 480 PVU under sub-capacity. The same deployment at full capacity is 3,840 PVU. That is the eightfold swing IBM is looking to apply wherever the sub-capacity evidence is incomplete, and it is why a finding against WebSphere can be so large relative to what the software actually consumed.

Common audit traps for WebSphere

• Edition confusion. Base, Network Deployment, and Liberty have different entitlements, and deployments drift between them without the license following.
• Sub-capacity not proven. ILMT missing, installed late, or not running continuously turns a small footprint into a full-host charge.
• VMware cluster reach. Where a WebSphere guest can move across hosts, the capacity count can expand to the cluster unless mobility is constrained and evidenced.
• Non-production in scope. Development, test, and disaster recovery copies counted as if they were licensed production.
• Bundled entitlements ignored. WebSphere entitlements that arrive bundled with other products, not credited against deployment.

How we defend it

We rebuild the WebSphere position from the metric up. We confirm the correct edition and entitlement for every instance, recalculate PVU independently using your own ILMT data, and validate that sub-capacity eligibility holds for each hypervisor and environment. Where IBM has defaulted an environment to full capacity, we produce the evidence that narrows it back to the virtual cores in use, and we separate non-production copies that should never have carried a production charge. The result is a documented count that meets each finding with a number you can stand behind.