{"slug":"business/50-troubleshooting","title":"Troubleshooting","section":"business","sectionLabel":"Business","status":"unclassified","type":"document","path":"business/50-troubleshooting.mdx","bytes":3180,"description":null,"updated":null,"source":"integratewise/Integratewise-docs-wiki","content":"\nUse this page for common integration and operator issues. For step-by-step resolution, also see:\n\n- `docs/internal/runbooks/`\n- `docs/internal/operations/DEPLOYMENT_RUNBOOK.md`\n- `docs/internal/runbooks/sync-failure-and-dlq-redrive.md`\n\n## Common symptoms and causes\n\n### Projection is stale\n\nLikely causes:\n\n- sync is lagging behind provider changes.\n- cache refresh has not occurred.\n- provider change did not pass promotion checks.\n- tenant policy windows delay refresh.\n\nActions:\n\n- inspect provider health and sync state.\n- confirm whether cache or Spine source is stale.\n- do not assume stale projections are safe to act on without disclosure.\n\n### Capability is unavailable\n\nLikely causes:\n\n- provider degraded or disconnected.\n- user authority is insufficient for the current context.\n- organization policy restricts the capability shape.\n- required fields or state prerequisites are missing.\n- feature policy is disabled for the tenant.\n\nActions:\n\n- use available provider health indicators.\n- review authority and role scope.\n- verify capability contract requirements before retrying.\n\n### Execution proposal is denied\n\nLikely causes:\n\n- pre-proposal governance denied authority or scope.\n- post-proposal governance denied risk, evidence, or policy alignment.\n- confidence is irrelevant if authority is missing.\n- execution plan is malformed or uses stale configuration.\n\nActions:\n\n- review governance denial reason and affected fields.\n- correct proposal inputs, evidence, or authority source.\n- recompile configuration if plan version is stale.\n\n### Approval is stuck\n\nLikely causes:\n\n- assigned approver is absent.\n- approval policy requires multiple actors.\n- policy changed after proposal creation.\n- provider or connectivity issue blocks dispatch after approval.\n\nActions:\n\n- use delegation when policy permits it.\n- escalate through defined approval policy channels only.\n- do not bypass governance for expedience.\n\n### Integration webhook failures\n\nLikely causes:\n\n- signature validation failed.\n- tenant correlation did not match any active workspace.\n- payload exceeded schema constraints.\n- inbound path was rate-limited or unavailable.\n- queue backpressure delayed ingestion.\n\nActions:\n\n- verify signing secrets and rotation schedule.\n- inspect correlation and tenant scope in webhook logs.\n- use retry with backoff only where idempotency is guaranteed.\n\n### Twin output is incorrect or stale\n\nLikely causes:\n\n- active context is stale.\n- recommended action is based on outdated evidence.\n- memory promotion rules have not applied.\n- provider state changed after context assembly.\n\nActions:\n\n- refresh context before acting.\n- inspect evidence and provenance.\n- use correction and reporting paths to reduce similar incorrect suggestions.\n\n## Information to gather before escalation\n\n- tenant ID, workspace ID, user ID\n- correlation ID or request ID\n- proposal ID and approval token if present\n- execution-plan version and timestamp\n- affected context identifiers\n- provider identifiers and connection states\n- exact error phase: observe, orient, decide, act, sync, promotion\n- whether customer or external execution endpoints are involved\n"}