DBA_FEATURE_USAGE_STATISTICS is the publisher's feature telemetry inside every Oracle Database. The LMS audit script reads it. The buyer-side discipline is to read it first, reconcile it against entitlement, and resolve the false-positive patterns before the publisher does.
DBA_FEATURE_USAGE_STATISTICS is a system view inside every Oracle Database instance. The view captures feature telemetry against an internal catalogue of feature codes; the catalogue runs to several hundred entries covering options, management packs, advanced security features, partitioning, in-memory, Real Application Clusters, and a long tail of internal database features.
The view's principal columns are NAME (the feature code), CURRENTLY_USED, DETECTED_USAGES, FIRST_USAGE_DATE, LAST_USAGE_DATE, AUX_COUNT and FEATURE_INFO. Each row records a single feature's lifetime telemetry inside that instance.
The wider editorial sits in the Oracle pillar; the option / pack inventory context sits at Database options and management packs; the LMS audit anatomy that reads the view sits at the LMS audit, anatomised.
The LMS Database script extracts the view as part of the wider audit-collection package. The script does not interpret the view; it captures the rows for offline analysis. The interpretation happens inside the LMS engagement, against the publisher-side feature-to-option mapping.
The script also captures DBA_REGISTRY (installed database components), V$OPTION (linked options at instance level), V$LICENSE (the session and CPU watermarks), the parameter values for partitioning and compression, and the deployment topology against the Oracle home. The combined extract is the audit-quality read against the database instance.
The buyer-side discipline is to run the same script, read the same view, and prepare the same reconciliation before the publisher's audit team arrives. The wider LMS scripts read closes the script-mechanics question; the wider LMS scope-control read closes the engagement-boundary question.
Several feature codes in the view carry well-documented false-positive patterns. The feature is marked used because an internal Oracle process referenced the feature once, not because an end-user workload consumed the feature in a way that requires an option licence.
The classic patterns are: Advanced Compression marked used because of an internal index-rebuild operation; Partitioning marked used because an internal partitioned object exists (often a system-generated partition for a particular feature); Diagnostic Pack and Tuning Pack marked used because an AWR snapshot was captured or because OEM polled the instance; Real Application Testing marked used because an internal capture / replay test was run by a DBA against a development instance.
Each pattern has a reconciliation path. The discipline is to resolve the pattern on the buyer side, document the resolution, and present the resolved row at audit. The detail sits in the Database options and management packs read; the audit-quality language for presenting the resolution sits at the LMS settlement position.
The buyer-side run is the discipline of reading the view on a periodic cadence (quarterly at minimum, monthly during a renewal or audit window), reconciling each used row against the option / pack inventory, and resolving the false-positive patterns. The output is the audit-grade feature inventory.
The reconciliation has three columns per row. First, the raw view output (NAME, CURRENTLY_USED, DETECTED_USAGES, FIRST_USAGE_DATE, LAST_USAGE_DATE). Second, the entitlement read against the cumulative Schedule A (does the buyer hold the option / pack against the database server in question). Third, the resolution (a confirmed deployment requiring an option licence, a false positive resolved with a documented explanation, or a remediation in flight).
The reconciliation runs against the same database instances the publisher will read in an audit. The audit-grade inventory is the artefact the buyer holds when the publisher's read arrives; the wider Schedule A map sits at Schedule A, anatomised.
When a confirmed use is found without an entitlement, the remediation moves are: stop the use (decommission the feature, refactor the workload), buy the option (acquire the missing entitlement), narrow the deployment (move the workload to a database server already authorised for the option), or formalise the exception (negotiate a written carve-out from the publisher).
The remediation is best executed before the publisher reads the view. A remediation in flight during a contract-cycle event (renewal, audit, BYOL move to cloud) reads as a documented programme; a remediation surfaced for the first time inside an audit reads as a finding. The cycle context sits at the Oracle renewal cycle.
The publisher-side language for the remediation matters. The buyer presents the remediation as a managed programme with a date and an owner, not as an open finding. The LMS settlement position read frames the language; the Audit Defence programme applies it during an active audit moment.
The buyer-side artefacts to hold against database feature use are: the periodic raw DBA_FEATURE_USAGE_STATISTICS extract per instance, the reconciled audit-grade inventory per instance, the option / pack entitlement read against the cumulative Schedule A, the false-positive resolution log, and the remediation log indexed by date and owner.
The audit-quality posture is then a documented programme: regular reads, regular reconciliations, regular remediations, regular review against the cumulative Schedule A. The publisher's read on the same view arrives, finds the same telemetry, and meets a managed response rather than a surprised one.
The wider engagement sits in the Oracle practice; the aggregated reading list sits in the Oracle knowledge hub; active renewal moments route to the Renewal Programme; active audit moments route to Audit Defence.
The option / pack inventory and the entitlement map read in plain language.
How the publisher reads the same view and the wider audit collection.
The script-mechanics question that frames the database extract.
A senior Admodum Oracle advisor will read your DBA_FEATURE_USAGE_STATISTICS extract against the cumulative Schedule A on a private call. Active audit moments route to Audit Defence.