feat(reliability): complete tranche two controls #64
Merged
wasabi opened 4 days ago

Summary

Implements Tranche 2 reliability and correctness controls:

  • Adds transactional PostgreSQL-to-ClickHouse outbox delivery, leases, retries, dead letters, reconciliation state, and producer integration.
  • Adds durable KGP worker leases, cursor failure handling, and dead-letter metadata.
  • Adds scheduler lease-token ownership guards.
  • Adds explicit remote-sensing provider fallback and durable provider attempt history.
  • Adds scoped telemetry freshness results and durable alert history, cooldown, escalation, and recovery state.
  • Adds bounded/stable exports, atomic output handling, and durable report-generation run status.
  • Adds generation-safe RDF/graph/IRI rebuild coordination and IRI provenance.
  • Adds migrations 335-338 and 340-343 without editing applied migrations.

Verification

  • Focused Tranche 2 reliability tests: 114 passed.
  • Fast-fail CI checks: 65 passed.
  • Security Ruff gate: passed.
  • Migration validation: 451 sources passed.
  • Full Compose suite was run with a 15-minute timeout. It stalls in test_e2e_platform.py::test_all_tables_exist because the test launches hundreds of sequential docker compose exec checks against the existing database; no new assertion failure was reported after the focused fixes.
Commits were merged into target branch
1/1
Submitter wasabi
Target main
Source feat/tranche2-reliability-correctness
Jobs
Merge Strategy
Create Merge Commit
Watchers (1)
Reference
pull request KI-64
Please wait...
Connection lost or session expired, reload to recover
Page is in error, reload to recover