{"schema_version":1,"experiment_id":"eoa_inverse_innovation_exp09_archetype_breadth150_20260804","research_id":"eoa_inverse_innovation_exp09_light_prior_art_20260804","cell_id":"constraint_propagation_and_decoupling__computer_science","search_lanes":{"direct_problem_and_intervention":{"queries":["zero downtime database schema migration compatibility multiple application versions rollback expand contract","database schema evolution model checking deployment migration state space compatibility rollback","schema migration dependency graph constraint solver service versions backfill"],"source_ids":["SRC2","SRC3","SRC4"],"no_result_note":null},"synonyms_and_historical_terms":{"queries":["database evolution dependency analysis parallel independent schema changes","Evolutionary Database Design parallel change schema migration","database migration planner formal verification application schema compatibility paper"],"source_ids":["SRC1","SRC2","SRC4"],"no_result_note":null},"products_practices_and_standards":{"queries":["site:docs.aws.amazon.com database schema changes backward compatible deployments expand contract","site:docs.gitlab.com database migration multiple versions backward compatibility zero downtime","site:github.com schema migration planning dependency graph compatibility versions rollback tool"],"source_ids":["SRC1","SRC2","SRC3"],"no_result_note":null},"component_combination":{"queries":["database schema evolution dependency impact automatic ordering constraints","schema migration constraint rollback deployment orchestration","parallel database migrations independent tables deployment orchestration"],"source_ids":["SRC3","SRC4"],"no_result_note":"No retained source implemented the complete combination of cross-service fixed-point state reduction, separator-derived work cells, and global recomposition checking; this bounded phrase and component search does not establish novelty."}},"sources":[{"source_id":"SRC1","title":"Shared-database-per-service pattern","publisher":"Amazon Web Services","url":"https://docs.aws.amazon.com/prescriptive-guidance/latest/modernization-data-persistence/shared-database.html","source_type":"OFFICIAL_GUIDANCE","claims_supported":["Sharing one database across microservices creates development-time and runtime coupling.","Schema changes must remain compatible with current and previous versions of every affected microservice.","Shared writers and long-running transactions can create cross-service table and lock dependencies."]},{"source_id":"SRC2","title":"Backwards compatibility across updates","publisher":"GitLab","url":"https://docs.gitlab.com/development/multi_version_compatibility/","source_type":"FIRST_PARTY_PRODUCT","claims_supported":["Non-atomic deployments create prolonged mixed-version intermediate states across application nodes, migrations, and the database.","GitLab documents ordering-sensitive compatibility failures, including old writers violating a newly introduced database constraint.","Expand-migrate-contract and separate pre-, post-, and background-migration phases are established practices for preserving compatibility."]},{"source_id":"SRC3","title":"Schema Migration: Safe database schema migrations with automated constraint enforcement","publisher":"Ryvn","url":"https://ryvn.ai/docs/guides/examples/schema-migration-constraints","source_type":"FIRST_PARTY_PRODUCT","claims_supported":["Ryvn represents each release's supported schema versions and automatically constrains deployment paths.","Its orchestration forces migration-capable intermediate releases, blocks advancement until migration completion, and rejects rollback targets incompatible with current data.","This is close product prior art for constraint-driven migration ordering and rollback validation."]},{"source_id":"SRC4","title":"Automatic Recommendations for Evolving Relational Databases Schema","publisher":"arXiv","url":"https://arxiv.org/abs/2404.08525","source_type":"PRIMARY_RESEARCH","claims_supported":["The research models dependencies among database entities and iteratively derives additional operations needed to maintain consistency.","It orders derived operations into a valid patch using a dependency graph and reports experiments on real database evolutions.","Its scope excludes external application programs, live data migration, and multi-version deployment compatibility."]}],"problem_evidence":{"status":"SUPPORTED","finding":"The problem is directly visible. AWS identifies shared-database development, runtime, writer, and lock coupling across microservices. GitLab documents long-lived mixed-version states and concrete incidents where locally plausible schema or component changes failed when combined. Ryvn describes corruption, outage, out-of-sequence deployment, and unsafe-rollback risks addressed by automated schema-version constraints.","source_ids":["SRC1","SRC2","SRC3"]},"closest_prior_art":[{"name":"Ryvn schema-migration constraint enforcement","source_ids":["SRC3"],"overlap":"Declares release-to-schema compatibility, derives allowed deployment transitions, forces migration-capable intermediate versions, gates on migration completion, and validates rollback targets.","remaining_difference":"The documented mechanism is principally a service/schema-version transition check. It does not show a full cross-service constraint network over readers, writers, backfills, routing, locks, and rollback targets; fixed-point domain reduction; separator-derived migration cells; or recomposition of permitted interleavings."},{"name":"Automatic Recommendations for Evolving Relational Databases Schema","source_ids":["SRC4"],"overlap":"Builds a dependency meta-model, iteratively propagates change impacts into additional required operations, orders the resulting operations, identifies disconnected dependency structure, and retains DBA control.","remaining_difference":"It operates inside the database schema, expressly excluding external application programs and data. It does not prove live coexistence across service versions, derive owner-addressable cells by separator analysis, or recompose local deployment plans against global availability and rollback invariants."},{"name":"GitLab mixed-version compatibility and expand-migrate-contract practice","source_ids":["SRC2"],"overlap":"Models deployment as successive schema and component-version states, requires compatibility throughout non-atomic transitions, and separates phases according to ordering dependencies.","remaining_difference":"The guidance and migration framework prescribe and test known phases but do not document automatic fixed-point implication propagation, residual-graph separator analysis, or a general recomposition checker for independently prepared cells."}],"prior_art_disposition":"ADJACENT_PRIOR_ART","contrastive_claim_remaining":"For a bounded shared-schema migration, a unified constraint model spanning all participating services and operational states can first derive forced and impossible states, then use only the residual dependency graph to form explicit-boundary migration cells, and finally reject locally valid cell plans whose combined sequence or permitted interleavings violate an original compatibility, availability, data, or rollback invariant. The retained prior art covers substantial components, but no retained source demonstrates this complete procedural chain.","contrastive_claim_falsifier":"The claim is falsified if the replay misses any known critical dependency, proposes a boundary crossed by a verified hard dependency, accepts an interleaving that violates a declared invariant, or produces no auditable reduction in feasible states or coordination scope beyond Ryvn-style compatibility gates and an ordinary expand-migrate-contract runbook.","gates":{"adequate_source_search":{"status":"PASS","rationale":"The bounded search covered the proposal directly, expand-contract and parallel-change terminology, first-party deployment practices and products, official shared-database guidance, and research combining dependency modeling with iterative impact propagation and operation ordering. Exactly four opened sources from four publishers were retained.","source_ids":["SRC1","SRC2","SRC3","SRC4"]},"supported_problem":{"status":"PASS","rationale":"Official and first-party sources directly show cross-service database coupling, mixed-version compatibility windows, ordering failures, writer conflicts, lock coupling, and rollback constraints.","source_ids":["SRC1","SRC2","SRC3"]},"distinct_testable_claim":{"status":"PASS","rationale":"Despite close component-level prior art, the remaining claim is contrastive and observable: residual-dependency cells produced after global propagation must preserve every declared invariant when recomposed and must reduce coordination beyond existing gates and runbooks.","source_ids":["SRC2","SRC3","SRC4"]},"bounded_next_test":{"status":"PASS","rationale":"A two-hour, read-only replay of one completed migration with at most six services, three schema transitions, one rollback target, and two alternative interleavings is bounded. Dependency recall, boundary violations, state reduction, explanation traceability, and invariant preservation are recordable outcomes.","source_ids":["SRC2","SRC3","SRC4"]},"no_obvious_safety_or_authority_stop":{"status":"PASS","rationale":"The proposed first test uses archived artifacts, is advisory, is disconnected from production and deployment control, and leaves classification and approval with existing human authorities. The main false-independence risk is contained by halting on untraceable implications or contradictions and discarding the advisory partition.","source_ids":["SRC3","SRC4"]}},"screen_survival":true,"world_novelty_boundary":"This bounded public-web screen found adjacent prior art and supports only a narrow, testable procedural distinction. It cannot establish world novelty, patentability, market size, expert acceptance, completeness of the dependency inventory, or realized operational value."}