A Tier 1 industrial received an LMS finding tied to undeclared Partitioning, Advanced Compression, and Diagnostics Pack usage. Three quarters of the claim dissolved on documentary grounds.
The client operated a global manufacturing estate built on Oracle Database Enterprise Edition across forty seven production and non production instances. Oracle License Management Services opened a formal audit and ran its measurement scripts against the full estate. The output captured every database option that had ever been touched, including Partitioning, Advanced Compression, and the Diagnostics and Tuning Packs.
The finding was blunt. Oracle asserted that option usage was detected on instances that held no entitlement, and priced the shortfall at list, with back support added across the asserted usage period. The opening number was $24.0M. The script made no distinction between options enabled deliberately for production workloads and options that had been switched on once by a previous administrator, used briefly, and never run in anger.
This is the central problem with an options audit. The Oracle measurement collects the fact that a feature was instrumented. It does not, by itself, establish who enabled it, under what authority, or whether the usage carried any commercial benefit. The buyer side burden is to reconstruct that history with evidence Oracle cannot dismiss. Without that reconstruction, the list price claim stands and the support multiplier compounds it every year the dispute drags on.
The client had no internal record of option activation events, no central change log tying activations to approved projects, and no measurement baseline of its own. The estate had grown through two acquisitions, each carrying its own contract with different entitlement definitions. The audit therefore arrived against a fragmented contractual position with no buyer side measurement to argue from.
We took the engagement under the Contain phase first. The immediate priority was to stop the flow of additional data to Oracle until the buyer side had reviewed every script output and understood exactly what each line item asserted. We established a single channel for all Oracle correspondence, paused the audit calendar to a defined response window, and began an independent measurement in parallel.
The Measure phase reconstructed option activation history forensically. We pulled the database feature usage statistics, the DBA_FEATURE_USAGE_STATISTICS views, and cross referenced first and last usage timestamps against the client change management system, project records, and administrator access logs. The objective was to separate three categories of usage. First, options genuinely run in production against live workloads, which the client would license. Second, options enabled during evaluation or migration and never used commercially. Third, options that the script flagged on the basis of a single incidental touch, often a default behaviour of a packaged tool or a one time administrative action.
The evidence was decisive. The majority of the flagged Partitioning and Advanced Compression usage fell into the second and third categories. Diagnostics Pack activations traced to a monitoring tool that had been installed with default settings and produced feature usage entries without any human decision to license the pack. We documented each line, attached the supporting evidence, and built a finding response that addressed the claim instance by instance rather than as a single aggregate number.
In the Negotiate phase we argued the response line by line. Oracle conceded the incidental activations once the documentary trail was placed in front of it, because the script output alone could not survive scrutiny against timestamped change records. The genuine production usage was acknowledged and licensed. The remaining exposure was converted into forward commercial value rather than a backward penalty payment.
The Convert phase rebuilt the contract. Rather than pay a settlement that vanished into back support, the client redirected the spend into a restructured support agreement with clean option pack entitlements, a defined activation governance process, and an audit cure provision giving the client a window to remediate accidental activations before they became findings.
The script captured everything that had ever been touched. The buyer side burden was to prove who touched it, when, and whether it mattered.
The $24.0M opening claim settled at $5.8M, a reduction of 76 percent. The settlement licensed only the options genuinely run in production, and the spend was structured as forward entitlement value rather than backward penalty. The table below shows how each exposure category resolved.
| Exposure category | Oracle claim | Settled |
|---|---|---|
| Partitioning (incidental) | $9.4M | $0.0M |
| Advanced Compression | $6.1M | $1.6M |
| Diagnostics & Tuning Packs | $5.3M | $1.1M |
| Production option usage | $3.2M | $3.1M |
The renegotiated support agreement carried three protections the original lacked. An option activation governance process required documented approval before any database option could be enabled in any environment, removing the ambiguity that created the audit. An audit cure provision gave the client a defined window to disable accidentally activated options before they could be cited as a finding. And the fragmented entitlement definitions inherited from two acquisitions were harmonised into a single consistent set of metrics across the estate.
The result reset the client position for the next cycle. The estate is now measured continuously against the contract on the buyer side, so the next audit, if it comes, opens against clean data and a documented governance trail rather than a script output the client cannot explain.
For the buyer side perspective on this product line, our Oracle Database licensing team and our Oracle audit defence practice work the same playbook on every engagement. Compare outcomes across the full case study library.
If Oracle License Management Services has opened a Database options audit, the first response shapes the outcome. We measure the estate independently before any further data leaves the perimeter.
Request a Consultation