Counting and Evidence · 10 min read

Discovery tooling for a ULA count

Three kinds of tool measure an Oracle estate for certification: Oracle LMS scripts, third party asset management discovery, and direct database queries. Each sees a different slice. The choice of tool shapes both the number you can certify and the risk you carry, so it deserves analysis rather than reflex.

By the Meridian advisory team, former Oracle LMS and GLAS licensing analysts. Updated 4 June 2026.

What tools measure an Oracle ULA estate?

Three main options exist, and they are not interchangeable. Oracle LMS measurement scripts produce Oracle's own view of your deployment. Third party software asset management discovery tools scan the estate and infer Oracle usage from what they find. Direct database queries read the configuration and feature usage from inside the databases themselves. Each captures a different slice of reality, and each has blind spots. A defensible count usually combines independent discovery with targeted verification rather than trusting a single source, because the certified number has to be both complete and provable, and no one tool delivers both on its own.

The tooling principle

The tool you choose shapes the number you can declare and the risk you take to declare it. LMS scripts are a choice, not an obligation, unless your contract requires them. Measure independently first, then decide.

Option one: Oracle LMS scripts

LMS scripts are the measurement Oracle provides and trusts. They read database configuration, options, and management pack usage, and they produce output in the form Oracle expects. The advantage is alignment: a count built from LMS output speaks Oracle's language. The risk is control. The scripts produce a particular view, and that view is not always the one most favourable or most defensible for you. More importantly, handing full script output to Oracle before you understand your own position can surface exposures, an option enabled in a pack you did not know was in use, for example, that you were not ready to discuss. Running the scripts is a choice unless your contract requires it, and even when you do run them, the timing is yours to control. Measure independently first, understand what the scripts will show, then decide whether and when to produce them.

Option two: software asset management discovery

Third party asset management tools scan the network and the estate, inventory the hosts, and infer Oracle deployments and configurations. Their strength is breadth: a good discovery tool finds servers and instances that manual processes and partial records miss, which matters for a maximised count because every legitimate deployment you fail to find is entitlement you fail to certify. Their weakness is precision on Oracle specifics. Feature and option usage, the detail that drives much of the licensing risk, is not always something a general discovery tool reads accurately. Discovery tools are excellent for finding the estate and weaker for measuring the nuance inside it, which is why they pair well with verification rather than standing alone.

Option three: direct database queries

Querying the databases directly reads configuration and feature usage from the source. This is the most precise view of what a given database is actually running, including which options and packs are enabled and used. The trade off is reach: queries only see the databases you point them at, so they depend on a complete inventory to be complete themselves. Direct queries are the verification layer that confirms what discovery found and corrects what discovery inferred. Used together, broad discovery finds the estate and database queries measure it accurately.

What each tool sees

ToolStrengthBlind spotBest used for
LMS scriptsAligned to Oracle's viewControl and timing of disclosureFinal reconciliation, where required
Asset management discoveryFinds the full estateOption and feature precisionInventory and completeness
Database queriesPrecise feature usageOnly sees databases it targetsVerification of the count

This comparison is indicative. The right combination for your estate depends on your products, your environments, and what your agreement requires, which is the analysis that comes before any tool is run.

Should you run Oracle LMS scripts before certifying?

Only after you understand your own position. The sequence matters more than the tool. If you run LMS scripts as the first act, you have produced Oracle's view of your estate before you know your own, and any surprise in that output is now a surprise you discover in front of the vendor. If you measure independently first, with discovery and database queries, you build your own picture, identify any exposures while you still control the conversation, and then decide what the contract requires and what you choose to disclose. Where the agreement mandates LMS measurement, you run it, but you run it informed. Where it does not, the scripts are one option among several, not a default.

Tooling and the count

Discovery is upstream of everything else in the count. The core factor math we set out in core factor math in a ULA exit is only as good as the hardware and deployment data the tools feed it. The risk of getting discovery wrong is not abstract: an estate measured incompletely produces an undercount, and an undercount is permanent entitlement surrendered, a cost we quantify in the undercount risk and its cost. Good tooling, used in the right order, is what turns a scattered estate into a complete, verified, defensible inventory.

A worked illustration

Consider an indicative retail group preparing to certify. Its first count, drawn from internal records, lists 280 database servers. Asset management discovery across the full network finds 340, surfacing a disaster recovery site and a cluster of non production instances that records had missed. Database queries against the additional hosts confirm which options are actually in use, removing a handful of false positives the discovery tool had inferred. The verified inventory settles at 326 servers, higher than the starting figure and now backed by evidence on every line. Had the group run LMS scripts first and stopped there, it would have measured a different and narrower slice, and it would have done so on Oracle's timetable rather than its own. The numbers here are indicative, but the pattern, discover broadly, verify precisely, disclose deliberately, is the one that holds.

The takeaway

No single tool gives you a complete and defensible ULA count. Use broad discovery to find the whole estate, database queries to measure it precisely, and LMS scripts where the contract requires them or where you choose to run them with your position already understood. Above all, control the order: know your own number before Oracle sees it. The full process sits in our pillar, the Oracle ULA certification guide. When you want the discovery run independently and in the right sequence, we can help.

Questions

Quick answers.

Three main options exist: Oracle LMS measurement scripts, third party software asset management discovery, and direct database queries. Each sees a different slice of the estate. LMS scripts are a choice unless the contract requires them, and a defensible count usually combines independent discovery with verification rather than relying on a single source.

Only after you understand your own position. LMS scripts produce Oracle's view of your deployment, and handing that output over before you have measured independently can surface exposures you were not ready to discuss. Measure your estate first, decide what the contract requires, then choose whether and when to run the scripts.

Strictly confidential

Measure independently first.

Book a confidential assessment and we will run discovery and verification in the right order, so you know your number before Oracle does.

Book a ULA assessment