How a broken ILMT becomes a multi year back payment.

ILMT rarely fails loudly. An agent stops reporting after a server rebuild, a new cluster is provisioned without the agent, a report is not generated for a quarter, and the tool keeps showing green for everything it can still see. The problem is invisible until an audit looks at the period the tool could not measure, and by then the gap is historical and cannot be filled.

Why a gap voids the claim

Sub-capacity is a privilege you earn by continuously measuring virtual capacity. When the agent on a host stops reporting, the virtual capacity for that host is no longer evidenced for that period. Without evidence, IBM is entitled to fall back to full-capacity charging for the affected host, counting every physical core whether or not the IBM software ever used them. One unmonitored host can move a single product from hundreds of PVU to thousands.

The lookback multiplies a small gap

The reason a quiet failure becomes a large number is the audit lookback. An IBM audit does not only assess today. It can reach back across multiple years, and the back-payment can run two to five years at full-capacity rates for any period the sub-capacity claim cannot be supported. A broken agent that went unnoticed for eighteen months is not an eighteen month problem if the lookback extends further and the surrounding periods are also thin.

• A stopped or missing agent voids virtual capacity evidence for that host
• Full-capacity charging applies to every physical core for the unmeasured period
• The lookback can extend the exposure across two to five years
• Misconfigured or uncorrected installs can fail silently while the dashboard looks healthy

Healthy dashboard, unhealthy evidence

ILMT also miscategorizes installs and needs manual correction, so the absence of red on the console is not proof of a defensible position. A bundle counted as a standalone product, a product version the catalog reads incorrectly, or an environment the agent never reached can all leave the tool reporting confidently on the wrong picture. The audit tests the evidence, not the dashboard color.

Catching it before IBM does

The defensible posture is to verify ILMT the way an auditor would: confirm every eligible host has a reporting agent, that coverage has no gaps across the retained quarters, and that catalog categorization has been reviewed and corrected. Found internally, a gap is a remediation task with a bounded period. Found by IBM in an audit, the same gap arrives already priced at full-capacity across the lookback.