Leaving WebLogic replaces a $5,500 per Processor Oracle support bill with either a Red Hat core subscription at roughly $766 per core or a zero-fee community runtime you support yourself. This page prices both routes honestly, including the Java SE bill and the lifecycle clocks that decide when you must move again.
Leaving WebLogic replaces a $5,500 per Processor Oracle support bill with either a Red Hat core subscription at roughly $766 per core or a zero-fee community runtime you support yourself. This page prices both routes honestly, including the Java SE bill and the lifecycle clocks that decide when you must move again.
The clock is not yours to set. Oracle's Fusion Middleware Statement of Direction puts Premier Support for 12c (12.2.1.4) at December 2026, with Extended Support closing December 2027. Beyond that, Oracle has signaled a Market Driven Support program, reportedly extending 12.2.1.4 coverage on a yearly basis through 2030. Read that for what it is: an unpriced line item, renegotiated every twelve months, with Oracle holding the only pen. In 25 years of these conversations, I have never seen an annually renewed bespoke support program get cheaper in year two. If you plan to stay on 12.2.1.4 past 2027, you are not avoiding a project, you are buying an option whose strike price Oracle sets after you have committed.
The stay-on-Oracle path is 14.1.2, released December 2024, with Premier Support planned through 2030. That sounds like breathing room until you read the rest of the roadmap. Oracle intends to ship the next major WebLogic and Coherence release in calendar 2026, renumbered as version 26, carrying Java 21 and 25, Jakarta EE 11, and post-quantum crypto. So the 2030 date you are budgeting against gets a successor before your 14.1.2 upgrade finishes testing. Two structural traps sit underneath this. First, AIX: 14.1.2 did not support it, and AIX coverage only returns with the 26 and FMW 27 releases, which leaves AIX shops on 12.2.1.4 with no in-place upgrade at all. Second, error correction on non-LTS releases: 14.1.1 was never a terminal release, and once a later 14.1.x shipped, error correction on 14.1.1 began winding down after a grace period. Customers who wanted LTS-only deployment were told to wait.
Version currency, not contract status, determines whether you get fixes, so a stay-put decision has its own migration project attached.
That is the point buyers miss when they model this. Paying support does not entitle you to patches for the binary you are running; it entitles you to patches for a release Oracle still corrects. A 12.2.1.4 estate in 2028 under Market Driven Support, or a 14.1.1 estate after 14.1.2 shipped, is a maintained contract sitting on an unmaintained runtime. Compare that honestly against the middleware alternatives and licensing exit options before you assume staying is the low-effort choice. Every route on the table, EAP, WildFly, or 14.1.2, is a migration. You are choosing which one, and who prices it.
Set the number the alternatives must beat. Per Oracle's Technology Price List effective April 16, 2026, WebLogic Standard Edition lists at $10,000 per Processor, Enterprise Edition at $25,000, and WebLogic Suite at $45,000. Premier Support runs 22%, so $2,200, $5,500, and $9,900 per Processor per year. The Suite premium over EE is $20,000 per Processor, which is the single largest downgrade lever in most middleware estates. Named User Plus follows the one-fiftieth rule across WebLogic, Coherence, and Internet Application Server lines, so crossover sits near 50 users per Processor. If SOA Suite is in the same contract, note it lists at $57,500 per Processor with $12,650 support, and its NUP crossover is 47.9 users, not 50.
| Line item | Figure |
|---|---|
| Estate: 4 nodes x 2 sockets x 12 cores | 96 cores |
| x86 core factor 0.5 | 48 Processor licenses |
| EE license at list ($25,000) | $1,200,000 |
| EE Premier Support at 22% of list | $264,000 per year |
| Same estate at Suite support ($9,900) | $475,200 per year |
| Same estate at SE support ($2,200) | $105,600 per year |
Now the correction that kills most business cases. Support is 22% of net, not list. Oracle discounts off TGPL, and the support base is list minus the negotiated discount. Across Oracle technology quotes we reviewed from 2024 to 2026, first-time Database purchases landed 50 to 80 percent below list. Middleware lines were consistently under-negotiated, because buyers fought over Database and signed whatever the reseller typed for WebLogic. So your actual EE exposure may be $264,000, or it may be $80,000, and the difference decides whether a Red Hat subscription is a saving or a cost increase. Pull the CSI support renewal lines and read the per-line net amounts before you model anything, and check whether Java SE is already coupled into your WebLogic entitlement, because that bill travels with you on exit.
Red Hat does not publish JBoss EAP pricing, which is deliberate and it works in their favor. Quotes arrive priced either on physical core count or on virtual instances, with two support tiers: Standard (web and phone during business hours) and Premium (24/7 for severity 1 and 2). There are no meaningful add-on SKUs to unbundle, so the negotiation is entirely about the counted quantity, the tier, and the term. Because you have no list price to anchor against, you need an external reference point before the first call. The most useful public data point available is a secondary market figure of roughly $12,261 per year for a 16-core Premium subscription, about $766 per core per year, with additional named support contacts at roughly $3,065 per year each. Treat that as a sanity check, not a target: in our experience Red Hat quotes at enterprise volume land well below single-unit web pricing, and if your quote arrives materially above $766 per core you are being sized for peak, not steady state.
| Support spend comparison | WebLogic EE (Oracle) | JBoss EAP Premium (reference) |
|---|---|---|
| Licensing unit | Processor (2 x86 cores at 0.5 core factor) | Core or virtual instance |
| Annual support per unit | $5,500 per Processor | ~$12,261 per 16 cores |
| Annual support per core | ~$2,750 | ~$766 |
| Ratio on support spend | 3 to 4x higher | Baseline |
| Extra support contacts | Included in CSI structure | ~$3,065 per year each |
The 3 to 4x gap on support spend alone is the headline, but it evaporates if you let Red Hat count badly. Push on five points. First, pin the counting definition in writing: socket pairs versus physical cores versus virtual instances produce wildly different totals on the same hardware, and hyperthreaded logical processors must be excluded explicitly. Second, force non-production into scope or into a separate zero-cost or discounted tier; paying Premium rates for dev and test is the most common overspend we see. Third, buy a three-year term with pricing locked for all three years, not a one-year deal with a renewal conversation. Fourth, cap renewal uplift at a stated percentage, ideally CPI-linked, because the second renewal is where the savings against Oracle quietly erode. Fifth, refuse to size to peak virtual instance count in an elastic environment: agree an average or a committed floor with true-up, and document the measurement method. Read this alongside the broader middleware alternatives and licensing exit analysis before you commit to a runtime.
The 3 to 4x gap on support spend is real, but it evaporates if you let Red Hat count badly.
WildFly carries no license fee and no subscription, and that is a genuine, bankable saving against a $5,500 per Processor Oracle bill. It is not free. WildFly ships on a fast community cadence (standard WildFly moved to Jakarta EE 11 in WildFly 40, released May 2026) with short community support windows per release. You are not inheriting a seven-year certified lifecycle; you are inheriting an upgrade obligation measured in months. Somebody on your payroll now owns the work that a subscription outsources: monitoring CVE feeds for the runtime and its transitive dependencies, deciding whether to upgrade or backport, building and regression-testing your own patch stream, and standing up an internal escalation path for production incidents at 2am with no vendor to call.
The subtler cost is paperwork. ISVs certify against JBoss EAP versions, not WildFly builds, and auditors and regulators ask for a named vendor support statement. If your estate carries third-party application certifications or sits under financial services or healthcare examination, community WildFly will generate friction that no engineering effort resolves. Where WildFly holds up well: internal-facing applications, non-regulated workloads, and DevOps-mature teams that already own container image pipelines and patch automation.
The pattern that survives procurement review most reliably is hybrid. Run JBoss EAP subscriptions on production nodes where you need the certification statement, the seven-year lifecycle, and the 24/7 escalation, and run WildFly across development, test, and staging where you need none of those and where per-core subscription costs multiply fastest. That structure typically removes 40 to 60 percent of the countable cores from the Red Hat quote based on the estates we have modeled, and it gives you a credible walk-away position at renewal: if Red Hat pushes uplift too hard, you already run WildFly competently and can extend it upward. Model the runtime decision alongside the Java SE exposure inside the migration, because both EAP and WildFly still require a JDK decision that can cost more than either runtime.
Every migration business case that ignores the next migration is wrong. JBoss EAP runs a seven-year long-life cycle: four years of Full Support, where Red Hat delivers minor releases and qualifies new platforms, then three years of Maintenance Support, where you get security and severe-defect fixes only. Extended Life Support is a separate purchase on top, and it is split into ELS 1 and ELS 2, with critical fixes ceasing after ELS 1. That last detail matters more than the headline seven years, because an ELS 2 subscription buys you an audit answer, not a patched runtime. EAP 7 already went end of life on June 30, 2025, and EAP 8.1.x carries support to February 5, 2031, so a 2026 landing gives you roughly five years of runway, not seven. Compare that to Oracle, where WebLogic 14.1.2 has planned Premier Support through 2030 and the version 26 release renumbers and resets the roadmap again in calendar 2026. WildFly has no dated cliff at all: the community stream moves continuously (WildFly 40 already shipped Jakarta EE 11), which converts a renewal negotiation into a standing upgrade obligation on your own engineers. Before signing anything, demand the exact end-of-Full-Support date for your specific EAP minor version in writing, plus indicative ELS 1 and ELS 2 pricing, so year five is not a blank check. Our experience across middleware renewals is that Red Hat will quote ELS only when the clock is nearly expired, and that is precisely when your leverage is gone. Model the same discipline you would apply to a WebLogic exit generally, as covered in our middleware alternatives and licensing exit analysis.
| Runtime | Support horizon | Renewal risk to price now |
|---|---|---|
| JBoss EAP 8.1.x | Full plus Maintenance to Feb 5, 2031 | ELS 1 and ELS 2 uplift, critical fixes stop after ELS 1 |
| JBoss EAP 7 | Ended June 30, 2025 | Already unsupported, no in-place option |
| WildFly (community) | Continuous, no dated EOL | Internal upgrade labor every release cycle |
| WebLogic 14.1.2 | Premier planned through 2030 | Version 26 resets roadmap, 22% support continues |
| WebLogic 12.2.1.4 | Premier ends Dec 2026, Extended Dec 2027 | Unpriced Market Driven Support, renegotiated yearly |
Neither EAP nor WildFly is certified to run on the Oracle JDK as your standard, and both are designed around OpenJDK builds, so the app server move only delivers savings if the Java runtime moves with it. Leaving an Oracle JDK in place while you congratulate yourself on killing a $5,500 per Processor WebLogic support line is the single most expensive mistake we see in this project. Java SE is now sold on the Universal Subscription, an Employee metric starting at $15 per employee per month with published tiers falling to $5.25. Employee is not a user count: it includes full-time, part-time and temporary staff plus the employees of your agents, contractors, outsourcers and consultants supporting internal operations, whether or not any of them ever touch Java. Oracle's own worked example in the global price list is 28,000 employees at $6.75 per month, which is $2,268,000 per year. A 10,000-employee enterprise sits at roughly $1.8M to $2.4M annually at list, dwarfing the middleware line you set out to cut. Sequence accordingly. Start the JDK swap ahead of, or in parallel with, the app server work, because OpenJDK migrations at large enterprises have run nine to fourteen months, and the blocker is almost never application code: it is discipline over developer workstations and CI/CD pipelines that quietly pull an Oracle build. Read our detail on the Java SE bill hiding inside your WebLogic migration and the practical steps in exiting Oracle Java SE. Action: inventory every Oracle JDK binary before you scope EAP cores, and make removal of Oracle Java a gating milestone in the migration plan rather than a cleanup task at the end.
Killing a $5,500 per Processor WebLogic support line means nothing if an Oracle JDK stays behind and triggers a headcount-priced Java bill.
Use the 48-Processor, 96-core reference estate (four nodes, two sockets, twelve cores, 0.5 x86 core factor) and price three years honestly. Stay-put costs $264,000 per year in WebLogic Enterprise Edition support at list, plus a mandatory 14.1.2 or version 26 upgrade project because 12.2.1.4 Premier Support ends December 2026 and Extended Support ends December 2027, plus an unpriced Market Driven Support line if you slip past 2027. EAP Premium at the ITQlick reference point of roughly $766 per core per year is about $73,500 annually for 96 cores, so the subscription pays back in year one; the cost is migration labor. WildFly removes the subscription entirely and moves the money into headcount, which in our experience means two to three loaded FTEs to own patching, CVE triage, and version currency for an estate this size.
| Three-year line item | Stay on WebLogic EE | JBoss EAP Premium | WildFly |
|---|---|---|---|
| Support or subscription | $792,000 | ~$220,500 | $0 |
| Forced upgrade or migration project | 14.1.2 or v26 upgrade, unavoidable | one-time migration labor | one-time migration labor |
| Internal support headcount | existing | existing | 2 to 3 loaded FTEs |
| Lifecycle reset | next release renumbering (v26) | seven-year window plus ELS | continuous, community cadence |
| Java SE exposure | Oracle JDK likely in scope | non-Oracle JDK | non-Oracle JDK |
Three items decide the outcome regardless of runtime. First, whether those Oracle licenses can actually be terminated or only shelved: perpetual licenses do not refund, and shelved licenses still carry compliance obligations. Read the licensing exit mechanics for middleware migrations before you assume the $264,000 disappears. Second, partial termination and repricing clauses, which let Oracle recalculate the unit price on retained lines when you drop part of a CSI, quietly erasing the savings you booked. Third, audit exposure during the parallel-run window, when both estates are live and your deployed processor count temporarily exceeds entitlement.
Run this as a ninety-day sequence, in order, and do not compress it.
One discipline matters more than any of the above. Raise no termination or reduction request with Oracle until the replacement platform is contracted and the migration plan is funded. Signalling an exit before you hold a signed alternative invites two predictable responses: an audit letter timed to your parallel-run window, and a bundled WebLogic Suite renewal offer priced to look cheap against a migration you have not yet costed. Hold the exit signal until your leverage is real.
On support spend, almost always yes. WebLogic Enterprise Edition support runs $5,500 per Processor per year, roughly $2,750 per x86 core at the 0.5 core factor, while a Red Hat EAP Premium subscription referenced at about $12,261 for 16 cores works out near $766 per core per year. The caveat is that your Oracle support is 22% of the net licence fee, so a deeply discounted original purchase narrows the gap, and migration labor plus the OpenJDK swap consume most of year one savings.
Legally yes, WildFly is open source with no licence fee. Operationally you take on CVE tracking, patch backporting, and production escalation yourself, and you accept a release cadence where community support windows are measured in months rather than the seven-year EAP lifecycle. Most enterprises land on EAP in production and WildFly in development and test, which keeps the certified support statement auditors and ISVs ask for.
EAP follows a seven-year long-life cycle: four years of Full Support with enhancements, bug and security fixes, then three years of Maintenance Support limited to bug and security fixes. Beyond that, Extended Life Support is a separate subscription split into ELS 1 and ELS 2, and critical fixes are only provided during ELS 1. EAP 8.1.x currently carries support to February 5, 2031, so a 2026 landing gives roughly five years before the next decision.
Premier Support for Fusion Middleware 12c (12.2.1.4) ends in December 2026 and Extended Support ends in December 2027. Oracle has signalled a Market Driven Support program beyond 2027, expected to run yearly through 2030, but it is unpriced and renegotiated annually, which is the weakest possible buyer position. The alternative in-place path is 14.1.2 with Premier Support planned through 2030, and a renumbered version 26 release expected in calendar 2026.
Not if you also move the runtime. Both EAP and WildFly run on non-Oracle OpenJDK builds, including Red Hat's own, so the Oracle Java SE Universal Subscription becomes avoidable. That subscription is priced on total employee count including contractors and outsourcers, from $15 down to $5.25 per employee per month, so a 10,000-employee firm faces $1.8M to $2.4M per year at list, which is often larger than the WebLogic bill itself.
It is a common trigger, particularly when a support renewal shrinks or a termination request appears. During migration you typically run both platforms in parallel, which inflates deployed Processor counts and creates real exposure. Keep the exit quiet until the replacement platform is contracted, freeze new WebLogic deployments, and document decommissioning dates before you approach Oracle about reducing the renewal.
Oracle Fusion Middleware is licensed per processor with the core factor. WebLogic editions from $17,500 to $120,000, the SOA Suite drag, Coherence, and how to license middleware to
Gated with a work email on the download page. No sales follow up you did not ask for.
Get the White Paper →
500+ enterprise clients. 11 vendor practices. Industry recognized. One conversation can change what you pay for the next three years.
One buyer side briefing a week. Renewal signals, audit moves, and the levers that work. No vendor spin.
