Rehearse metadata-schema rollback across old and new API consumers
The team wants to remove legacy JSON fields immediately after switching reads, which could break old clients during a rollback.
- Focused work estimate
- 5h + prerequisites
- Priority in the scenario
- Medium
- Engineering practice
- Schema evolution · Recovery
Estimated field mix
- Database engineering60%
- API design20%
- Platform engineering20%
Field percentages are editorial estimates of the ticket's engineering focus. They total 100%; they are not measured time, proficiency scores, or ownership evidence.
Review it, then add it to your workspace.
The board opens an editable draft; nothing is saved until you confirm it. Sign-in and workspace permissions apply, and Demo boards remain ephemeral.
Project context
A fictional records portal stores every property in one JSON column. Queries disagree about missing values, and malformed records make ordinary filters fail.
Setup prerequisites
- Create synthetic document metadata with malformed and legacy versions.
- Use migrations and a local query API.
Preceding work
Complete these dependencies, or supply their agreed outputs before taking this ticket.
- AMETA-101 · Separate stable document identity fields from flexible attributes
- AMETA-102 · Define versioned metadata schemas with explicit size and depth limits
- AMETA-103 · Choose distinct semantics for absent, null and empty metadata values
- AMETA-104 · Extract stable document fields through an additive migration
- AMETA-106 · Keep document metadata updates under optimistic revision control
- AMETA-107 · Build parameterized metadata filters with an allowlisted operator set
- AMETA-108 · Verify metadata filter results against a simple reference evaluator
- AMETA-105 · Backfill document timestamps without guessing missing offsets
- AMETA-109 · Enforce required relational document fields after a clean backfill
Acceptance criteria
- Test old and new clients against the expanded schema.
- Document the compatibility window and contraction prerequisites.
- Keep versioned metadata and relational values coherent during rollback.
Implementation constraints
- Use synthetic records from every supported version.
Verification to include
- Switch readers and restore the prior compatible client with unchanged results.
- Introduce an unsupported version and verify a clear failure rather than lossy conversion.
Deliverables
- Compatibility matrix and rollback drill
Rollout and recovery
Retain bridge columns through the reviewed window; contract only after old writers are retired.
Value of the work
For the engineer: Practice relational/JSON boundaries, versioned validation and query semantics.
For the team: Review a data model that stays inspectable while allowing controlled variation.
Evidence boundaries
Outcome Evidence: Tests, patches, and runbooks are requested deliverables. They become Outcome Evidence only through a qualified Mission and immutable Evidence IDs.
Ownership Evidence: Independent adaptation must be observed under a declared verification policy and cite immutable Evidence IDs. Completing a planning ticket establishes no Ownership Evidence.