Manufacturers carry distributed plants, large ERP estates, disaster recovery at many sites, and a long history of acquisitions. Each of those makes the certified count larger and harder to evidence, which is exactly why a careful, independent baseline pays for itself at exit.
Manufacturing estates are wide and old. Oracle runs in plants, in distribution centres, in disaster recovery sites, and in entities picked up through years of acquisition. That breadth grows the defensible count and complicates scope at the same time. Certify on an independent baseline that finds the deployments central IT cannot see and reconciles every site to your contract.
Three features of a manufacturing estate shape the certification. The first is geographic spread. Oracle is deployed across plants and sites that may run their own infrastructure and report into local IT, so the central view of where the software runs is usually incomplete. The second is the ERP and operations backbone, often a large Oracle Database and options estate behind manufacturing, supply chain, and finance systems, with test and disaster recovery copies multiplying the instance count. The third is acquisition history. Manufacturers grow by buying other manufacturers, and each acquisition brings plants and legal entities that may or may not sit inside the customer definition.
Taken together these mean the true count is almost always higher than the first internal estimate, and the scope is almost always messier. Both work in your favour if handled early and against you if discovered late. The breadth is free perpetual value waiting to be counted, and the scope is exposure waiting to be cleaned up before the certification letter freezes it.
In manufacturing the deployments you cannot see are the ones that decide the count. Find the plant and DR systems before Oracle's tooling does, and count every one you can defend.
Generally yes. Processor counting applies to production, test, and disaster recovery instances deployed within the term, and a database running at a plant or in a recovery site is as countable as one in the central data centre, provided you can evidence it. The evidence file is the gate: server lists, discovery tool output, and documented methodology that show the deployment existed and sat in scope. In a distributed manufacturing estate, the systems most often missed are exactly the ones that add the most, the standby copies, the plant level databases, and the test environments that proliferate around a production line.
This is where the breadth becomes an opportunity. A manufacturer that counts only what the central team already knows about leaves real entitlement behind. A manufacturer that runs proper discovery across every site routinely finds a count well above the starting estimate. The uplift is indicative and estate specific, but the pattern is consistent: the wider the footprint, the more there is to count, and the more a thorough baseline returns.
Manufacturers run large virtualized estates, and under Oracle's partitioning stance soft partitioning does not limit scope. That means an entire VMware cluster can be swept into the count wherever an Oracle instance runs on it. At exit this cuts both ways. As a risk, an uncontained cluster spanning many hosts can inflate exposure beyond what you intended to deploy. As an opportunity in a maximization context, the same rule lets a legitimately broad cluster contribute fully to your certified count. The lever is documentation and isolation: knowing which hosts run Oracle, which clusters are dedicated, and being able to evidence the boundary either way.
The clause that bites hardest in manufacturing is the customer definition. Years of acquisitions leave Oracle running in entities that joined the group at different times under different terms. Some are inside your ULA scope, some are not, and the difference decides whether their deployments can be certified or become a remediation demand. Before certifying, every plant and site should be reconciled to a legal entity and tested against the definition. An out of scope plant found now is fixable by migration, consolidation, or amendment. The same plant found in an audit two years later is a bill.
For a manufacturer the highest value first move is an independent baseline that runs discovery across every plant and site, reconciles each to a legal entity, and tests the VMware boundaries, so you enter the certification window knowing your number and its scope. Two companion playbooks sit alongside this one: Oracle ULA certification for media and Oracle ULA certification for travel and hospitality. The full method sits in our Oracle ULA certification guide.
Distributed plants, a long M&A history, large ERP estates, and DR sites at many locations. Oracle runs in places central IT does not always see, and acquired entities may sit outside the customer definition. The result is a count that is both larger and harder to evidence than first expected, which is why an independent baseline matters.
Production, test, and disaster recovery instances deployed within the term are generally countable, including those at plants and remote sites. The condition is evidence: server lists, discovery output, and methodology that prove the deployment existed in scope. Uncounted plant and DR systems are value left behind, so finding them is part of maximizing the count.
Acquired plants and entities may or may not be inside your customer definition. Oracle running in an entity outside scope cannot be certified and becomes a remediation risk. Manufacturers with a history of acquisitions should reconcile every site to a legal entity before certifying, because the scope questions bite hardest in this sector.
We run discovery across your full manufacturing estate, reconcile each site to its entity, and certify on a count that holds. Book a confidential assessment.