Skip to content

monitoring ​

probes ​

endpointanswers 200 when
GET /healthzthe process runs
GET /readyzthe server is ready for traffic: recovery and the changes at startup are complete

metrics ​

GET /metrics gives the state of the server in the Prometheus text format. It is on the admin listener, and it needs the read token.

metrictypewhat it is
tidepool_build_infogaugethe running version, as the version label
tidepool_uptime_secondsgaugeseconds since the server started
tidepool_epochgaugethe newest published epoch
tidepool_config_versiongaugethe config version that the server runs
tidepool_table_rowsgaugelive rows, per table
tidepool_table_bytesgaugememory of the segments of each table
tidepool_table_segmentsgaugesegments, per table
tidepool_dictionary_bytesgaugememory of the dictionary values
tidepool_cache_bytesgaugememory of the caches of the query engine
tidepool_rows_written_totalcounterrows written or deleted through the ingest API, per table and credential
tidepool_ingest_requests_totalcounteringest requests, per table, operation and status. A table that does not exist is (none).
tidepool_ingest_refused_totalcounteringest requests that the limits of a credential refused, per credential and limit
tidepool_epoch_publish_secondssummarythe time to publish an epoch
tidepool_live_refresh_secondssummarythe time to update every live session after an epoch
tidepool_sessionsgaugelive dashboard sessions
tidepool_viewsgaugeviews in memory (POST /v1/views)
tidepool_query_secondssummarythe time to answer POST /v1/query
tidepool_query_errors_totalcounterqueries refused with 400
tidepool_mcp_tool_calls_totalcounterMCP tool calls, per tool
tidepool_rebuild_rowsgaugerows that a running model change rewrites. Only during a rebuild.
tidepool_rebuild_rows_donegaugerows rewritten so far. Only during a rebuild.
tidepool_wal_last_lsngaugethe last write in the log. Only with a data directory.
tidepool_checkpoint_lsngaugethe last write in the newest checkpoint. Only with a data directory.
tidepool_checkpoint_age_secondsgaugeseconds since the newest checkpoint. Only with a data directory.

the status page ​

Status in the web UI shows the same values for people: memory, tables, epochs, the epoch and refresh times, the log and the checkpoints. It updates every few seconds. GET /v1/status gives the same as JSON.

reference generated from the schemas, the endpoint table and the MCP tools: just docs