Many components are monitored in isolation, and some are even missed. An E2E health check would allow us to capture errors anywhere in the QueryAPI pipeline.
We could achieve this by creating a * indexer which writes the block height to Postgres on every block. We could then monitor this block height via Hasura GraphQL queries, and assert that it is always increasing. If any component within QueryAPI is failing, i.e. Block Streamer/Runner/Postgres/Hasura the block height will not increase.
We may be able to query Hasura directly from Grafana with: Infinity
Many components are monitored in isolation, and some are even missed. An E2E health check would allow us to capture errors anywhere in the QueryAPI pipeline.
We could achieve this by creating a
*indexer which writes the block height to Postgres on every block. We could then monitor this block height via Hasura GraphQL queries, and assert that it is always increasing. If any component within QueryAPI is failing, i.e. Block Streamer/Runner/Postgres/Hasura the block height will not increase.We may be able to query Hasura directly from Grafana with: Infinity