The hidden cost of migration.
The hidden cost of Red Hat migration is the cost the published comparison models systematically leave off the page. The headline arithmetic compares subscription line against subscription line; the realised arithmetic is materially different. The realised cost is rarely smaller than the model suggests; in observed engagements it is meaningfully larger. This note sets out the cost lines published models leave out, the structural reason each one is left out, and the discipline that produces a defensible cost model for the buyer side decision.
The headline number, read in isolation.
The hidden cost of Red Hat migration begins with the structure of the published comparison. Vendor and alternative vendor literature compare the line price per subscription against the alternative line price per subscription, sometimes weighted by a coarse estimate of estate size and a generic estimate of support overhead. The arithmetic that reaches the steering committee from this comparison is typically a recurring saving line of meaningful size, presented as the cost differential between staying and moving. The realised cost differential, in the practice reading, is rarely the number presented in this form.
A buyer side reading separates the recurring subscription delta, which is what the published comparison measures, from the one time execution cost and the durable operational cost, which are the lines the published comparison does not measure. The recurring subscription delta is real and it matters. The full picture also includes a non recurring execution cost that lands in the first one or two years and a durable operational cost that runs for the lifetime of the alternative posture. Reading all three together is the discipline that produces a defensible decision rather than a regrettable one.
The companion notes on the exit planning service, on Rocky Linux migration economics, and on AlmaLinux migration timeline and risk cover the destination specific arithmetic. This note covers the cost lines that recur across every named destination and that the headline number systematically leaves out.1
Certification and ISV support, the durable line.
The first hidden cost is certification and ISV support. Many enterprise applications carry a published support matrix that names a specific operating system distribution and a specific version. Where the application is a commercial database, a middleware product, a workload from a specialised industry vertical, or a vendor managed appliance, the support matrix may name RHEL specifically and may withhold supported status on any alternative distribution. The buyer side decision to move off RHEL has to be read against the support matrix for every load bearing application in the estate. Each unsupported application becomes a cost line or a constraint.
The cost line takes one of three forms. The first is an explicit support fee uplift charged by the ISV when running on a non certified distribution, where the ISV will provide best effort support. The second is a parallel run in which the application stays on RHEL while the rest of the estate moves, which means the RHEL line stays on the contract at a smaller scale and the migration savings are correspondingly reduced. The third is a replacement of the application itself, which is a project of its own with its own cost line and its own calendar.
In observed engagements, the certification surface costs between five and twenty percent of the headline subscription saving once the support matrix has been read line by line and each unsupported application has been priced. The variance is wide because the cost depends on which ISVs are in the estate, how strict the support matrix is in each case, and how willing each ISV is to extend support to the named alternative in writing. The companion notes on the practice hubs cover the ISV crossover surface in product specific detail.2
| Category | Type | Range against headline saving |
|---|---|---|
| Certification and ISV support | Recurring | 5% to 20% |
| Parallel run | One time | 10% to 30% |
| Training and operational tooling | One time and recurring | 5% to 12% |
| Contractual overhang on RHEL | One time | Variable, frequently 10% to 25% |
| Internal program cost | One time | 8% to 18% |
| Post migration audit posture | Contingent | Significant if mismanaged |
Parallel run and operational tooling, the project line.
The second hidden cost is the parallel run. Every credible migration carries a window during which the source distribution and the destination distribution coexist in the estate. The window may be as short as a sprint or as long as the full term of the existing Red Hat contract; whichever the case, both distributions are running and both incur cost during the parallel run. The Red Hat line does not come off the contract on the day the first host migrates. The new line is on the contract from the day the first host moves. The compounded cost of the overlap is non trivial and frequently underestimated in the published model.
The third hidden cost is operational tooling. Configuration management content, monitoring agents, backup integration, security scanning, hardening baselines, and compliance reporting pipelines may all carry RHEL specific assumptions that have to be rewritten or replaced for the destination. The tooling rewrite is rarely budgeted on the migration line because it does not appear on the published comparison; the rewrite happens regardless, paid out of the platform team operating budget, and reduces the realised saving in the first year by an amount frequently equal to a quarter of the headline number.
The fourth hidden cost is training. The operating system on the disk may be functionally similar across distributions but the support relationship, the documentation, the patch cadence, the issue triage process, and the escalation path are all materially different on the alternative side. The platform team has to learn the new vendor relationship; the application teams have to learn the new support process; the security team has to learn the new advisory cadence. Each learning curve carries time, and time on the platform team is the most expensive currency in the migration.3
Contractual overhang on RHEL, the term line.
The fifth hidden cost is contractual overhang on the Red Hat side. The existing Red Hat contract typically carries a term of one, three, or five years. The migration cannot complete inside the residual term unless the Red Hat contract permits a partial term exit or a quantity reduction inside the term. The default Red Hat enterprise agreement does not permit either; the contract runs to the named end date at the named quantity and the named price, regardless of whether the buyer is still using the entitlements.
The cost line that results is the cost of paying for the unused balance of the existing Red Hat contract while the destination is being executed. In observed engagements that placed migration on an aggressive calendar against the residual term, the overhang cost reached between ten and twenty five percent of the headline saving across the first year of the alternative posture. The number is materially smaller where the migration calendar was aligned to the renewal cycle and the alternative was named ahead of the renewal sheet; the companion note on migration timing relative to the renewal cycle covers the timing surface in detail.4
The cost line is materially larger where the migration calendar was set first and the renewal cycle was an afterthought. The mechanism is straightforward: the buyer side is paying twice during the parallel run, once on the residual Red Hat term and once on the alternative. Buyers who reach the realisation late frequently discover that the realised first year cost of the migration is greater than the cost of holding the Red Hat contract through one more renewal and migrating across the next term break. The migration arithmetic is unforgiving on this line; the timing decision frequently matters more than the destination decision.
Internal program cost, the staffing line.
The sixth hidden cost is the internal program cost. The migration project itself is a project with named staff, a governance overhead, a steering committee, a status report, and a budget. None of this appears on the published vendor comparison; all of it appears on the buyer side internal budget. In observed engagements, the staffing line costs between eight and eighteen percent of the headline saving, depending on the scale of the estate, the maturity of the platform team, and the use of external implementation partners.
The internal program cost is the line where the cost model most often gets second order effects wrong. The staff time consumed by the migration is staff time not consumed by the platform roadmap, by the security backlog, by the application onboarding queue, or by the next renewal cycle. The opportunity cost of platform team attention during a migration is frequently larger than the explicit staff cost on the budget, and frequently does not appear on the cost model at all.
The seventh hidden cost, named here for completeness, is the audit posture on the residual Red Hat estate during and after the migration. A credible migration disposes of the RHEL binaries cleanly, documents the disposition in writing, and closes out the entitlement record with the seller side in a way that does not invite a compliance letter. A migration that does not do this work invites the audit profile the practice sees most often after a CentOS Stream migration: lapsed entitlement, binaries in place, compliance letter from the Red Hat account team within twelve to eighteen months. The companion note on how leaving Red Hat changes the current contract covers the disposition discipline.5
When the arithmetic still works.
None of the lines named above is an argument against migration. Each is an argument for honest cost modelling. The migration arithmetic does work for many estates, particularly across the second and third years of the alternative posture; the hidden costs are concentrated in the first year and tail off as the platform team absorbs the new operating posture. A defensible cost model reads three years rather than one, names every category of hidden cost rather than the headline alone, and prices each category against the specific estate rather than against a generic comparison.
The cost model that produces a defensible decision typically lands the buyer in one of three outcomes. The first is a full migration on the strength of the three year arithmetic, accepting the first year cost as the price of admission. The second is a hybrid posture in which load bearing applications stay on RHEL and the broader estate moves, which preserves the certification surface and captures a meaningful share of the saving. The third is the named destination held in posture, with the renewal contract rewritten under the leverage and no migration executed.
The companion notes on when not to migrate off Red Hat, on hybrid RHEL and alternative posture, on the credible migration plan as leverage, and on communicating the migration plan internally sit adjacent to this surface. The contact form at the engagement section returns a desk response on the cost model question typically inside the business day.6
The buyer side line worth holding through every cost modelling exercise is the line the practice repeats in every exit planning discovery call. The published comparison is a starting point. The honest model is the starting point plus six lines the published comparison leaves out. The decision that defends in writing is the decision drawn from the honest model. The decision that does not defend is the decision drawn from the headline alone.
Notes & references
- 1. Published vendor and alternative vendor comparison models typically reduce the migration arithmetic to a subscription line delta. The realised arithmetic in observed engagements includes one time execution cost and durable operational cost that the line delta does not capture. See practice notes on cost model construction.
- 2. ISV support matrix observations reflect engagement experience across enterprise database, middleware, and vertical application vendors. Variance is wide; the matrix has to be read line by line for each load bearing application rather than estimated at the estate level.
- 3. Training cost and operational tooling rewrite cost are frequently absorbed into the platform team operating budget rather than budgeted on the migration line. The realised first year saving is correspondingly reduced relative to the published comparison.
- 4. Contractual overhang on the Red Hat side depends on the residual term of the existing contract and the absence of partial term exit language in the default Red Hat enterprise agreement. Migrations aligned to the renewal cycle materially reduce this cost line.
- 5. Audit posture during and after migration is the contingent cost line that becomes material when binary disposition is mishandled. The practice sees the audit profile most often after CentOS Stream migrations where the entitlement record was not closed with the seller side in writing.
- 6. The three year cost model produces a materially different decision from the one year model. The honest decision frequently is to migrate; the honest decision frequently is to hold the destination in posture; the honest decision frequently is the hybrid. The cost model produces the decision; the published comparison does not.
Preparing a response? The practice keeps a one-page Red Hat audit response checklist — what to acknowledge, what to preserve, and what not to volunteer in the first fourteen days after the letter arrives.