Skip to main content

​
Service health

  • Web process availability and page render success.
  • API /health for process liveness and /readyz for database readiness.
  • PostgreSQL and Redis container health checks.
  • Migration job completion before API/indexer startup.

​
Chain and indexer health

  • Current Base Mainnet head versus finalized indexer checkpoint.
  • Checkpoint age and hash continuity.
  • RPC latency, retry exhaustion, rate limiting, and chain-ID mismatch.
  • Reorg recovery activity and noncanonical row counts.
  • Launch discovery, trade projection, and matched graduation-settlement errors.

​
Data integrity

  • Fee/vault liabilities versus contract or database backing, depending on the layer being checked.
  • Canonical projection counts and exclusion behavior.
  • Metadata drafts that remain unfinalized after a confirmed launch.
  • Null OHLC fields treated as unavailable rather than zero.
  • Oracle freshness and oracle_unavailable rate.
An API or indexer alarm does not imply onchain financial state changed. Start incident tracing from receipt and block provenance, then checkpoint, canonical raw events, projections, API, and frontend.