Sunset or Replacement Plan¶
Transition plan — instantiates Technical Debt Containment
A staged plan to retire and replace a piece of debt too structural to clean up in place — sequencing migration, cutover, and decommission so the obligation is closed by replacement rather than endless patching.
Some debt cannot be repaid by cleanup because the thing itself is the debt. Sunset or Replacement Plan is the mechanism for that case: a staged plan to retire and replace a piece of structural debt too deep, too central, or too far gone for in-place repair, sequencing migration, cutover, and decommission so the obligation is finally closed by replacement rather than patched forever. Its defining move is reclassifying an item from "repair" to "replace" and routing it to a sequenced ending — recognizing that another round of local fixes only compounds the problem, and that the honest repayment path is to build the successor and switch. It does not merely fund a fix or nibble at symptoms; it commits to an endpoint, escalates the item above the routine cleanup queue because its compounding risk dwarfs the rest, and lays out how the old thing dies.
Example¶
An insurer runs its book on a decades-old policy-administration mainframe: undocumented, written in a language few remaining staff know, and dependent on one engineer a year from retirement. No refactor can save it — the debt is structural, and every year of patching raises the single-point-of-failure risk. The team writes a sunset and replacement plan. A new platform is stood up behind a façade that intercepts traffic; one product line is migrated at a time, running in parallel against the mainframe to prove equivalence; each successfully migrated module lets a slice of the mainframe be decommissioned; and a firm sunset date anchors the whole sequence. The plan escalates this work above the ordinary cleanup backlog precisely because its compounding, key-person risk outweighs everything else. The debt is closed by replacement on a sequenced path — not by a fourteenth year of "one more patch."
How it works¶
The plan begins with a classification judgment: is this item repairable in place, or must it be replaced? Once replacement is chosen, the plan sequences the transition — building the successor, migrating in slices, running in parallel to verify equivalence, cutting over, and decommissioning the old surface piece by piece — and fixes an endpoint so the effort commits rather than drifts. Because structural debt of this kind compounds faster and riskier than routine cleanup, the plan also elevates its priority, pulling it out of the ordinary backlog where it would perpetually lose to smaller, easier items. What distinguishes it from a cleanup effort is the endpoint: the plan's success condition is that the debt-bearing thing no longer exists, not that it has been improved. It connects to broader creative-destruction and modernization work without collapsing into it — the plan's job is the disciplined ending of one debt item.
Tuning parameters¶
- Replacement strategy — big-bang cutover versus incremental strangling. Big-bang is faster to reason about but riskier; incremental is safer but demands parallel-running overhead.
- Cutover sequencing — which slices migrate first (lowest-risk to learn, or highest-risk to de-risk early). The order shapes where surprises land.
- Parallel-run duration — how long old and new run side by side. Longer proves equivalence but doubles the carrying cost and can become permanent.
- Decommission triggers — what evidence lets a slice of the old system be switched off. Loose triggers leave zombies running; strict ones stall the sunset.
- Sunset-date firmness — whether the endpoint is a hard commitment or a soft aspiration. A firm date forces the ending; a soft one invites indefinite parallel life.
When it helps, and when it misleads¶
Its strength is that it is the right answer to a failure the archetype names directly: structural debt misclassified as local cleanup, where leaders keep treating an obsolete system as a list of small fixes it will never actually clear. The plan closes the debt for good instead of servicing it forever. A real, named approach is Martin Fowler's Strangler Fig pattern — incrementally growing a replacement around the edges of a legacy system until the old one can be removed.[n1]
Its failure mode is the replacement that never lands. Sunset plans are prone to second-system syndrome, where the successor accretes ambitions the original never had and slips endlessly; to parallel runs that quietly become permanent because no one commits to the cutover; and to sunk-cost patching that keeps postponing the replacement decision itself. The classic misuse is invoking "replacement" to justify a rewrite that was already decided for other reasons, dressing a want as a debt necessity. The guarding discipline is a firm sunset date, incremental cutover with explicit decommission triggers, and an honest up-front classification that this genuinely is replacement, not repair in disguise.
How it implements the components¶
closure_or_reclassification_rule— the plan closes the item by replacement, reclassifying it from "repair in place" to "retire and replace" and defining what decommission counts as done.repayment_priority_rule— it escalates structural, high-compounding debt above the routine cleanup queue, where it would otherwise perpetually lose to easier items.
The plan retires structural debt permanently but does not police temporary shortcuts — exception_record is the Exception Expiry Date, its nearest twin (the expiry date times out a temporary waiver; the sunset plan replaces a structural thing) — and it does not supply the recurring capacity to do the work: repayment_cadence is the Repayment Reserve.
Related¶
- Instantiates: Technical Debt Containment — the plan is the repayment path for debt too structural for local cleanup.
- Consumes: Debt Severity Rubric identifies the high-compounding structural item that warrants replacement over repair.
- Sibling mechanisms: Technical Debt Register · Debt Severity Rubric · Debt Budget Review · Debt-Service Dashboard · Architecture or Process Decision Record · Quality or Health Scan · Repayment Reserve · Exception Expiry Date · Refactoring or Cleanup Sprint
Editorial Notes¶
Form Classification¶
Form family: Representation, Specification & Plan
Rationale: Sunset Or Replacement Plan is defined in the frozen evidence as: A staged plan to retire and replace a piece of debt too structural to clean up in place — sequencing migration, cutover, and decommission so the obligation is closed by replacement rather than endless patching. Its operative deployed or enacted form is therefore Representation, Specification & Plan.
Nearest alternative: Rule, Policy & Commitment — Rule, Policy & Commitment can support this mechanism, but the evidence centers the concrete operation described above rather than the alternative family's defining operation.
Review outcome: Adjudicated after independent review; high confidence.
Origin Attribution¶
Primary origin: Computer Science & Software Engineering
Origin pattern: Cross-disciplinary synthesis
Present-day reach: Universal
Rationale: Staged migration, cutover, and decommission is the system-retirement pattern for structural technical debt.
Related originating lineages:
- Engineering & Design — Lifecycle engineering manages safe transition between assets.
- Organizational & Management Science — Portfolio governance authorizes replacement and closure.
Review resolution: The blind reviewers agree that computer_science is the primary origin and differ only on origin mode disagreement, domain reach disagreement, encyclopedia synthesis disagreement. I preserve every independently explained alternate from both records rather than imposing a numeric cap. I retain cross_disciplinary_synthesis because the combined evidence shows material contributions from several lineages. The broader reach of universal records portability separately from historical provenance; encyclopedia_synthesis=true preserves the affirmative synthesis judgment where either reviewer identified one.
Encyclopedia synthesis: The exact catalogued form synthesizes established practice rather than reproducing a single standard historical label.
Review outcome: Reconciled after independent review; high confidence.
Notes¶
[n1] The Strangler Fig pattern, named by Martin Fowler after the vine that grows around a host tree until it can stand on its own, describes incrementally building a replacement system around a legacy one and migrating functionality piece by piece until the old system can be decommissioned. It is a real approach to closing structural debt by staged replacement rather than in-place repair. ↩