Lakebase system table reference

Important

The Lakebase system tables are in Beta.

Lakebase captures its observability telemetry into the system.lakebase schema. When the Lakebase Observability in System Tables preview is on for your workspace, Lakebase writes telemetry to these tables automatically. You don't create a configuration or choose a destination for it. Advanced Postgres telemetry is the first data to land here. The tables below list what each one holds.

For the detailed column reference for each table, see Lakebase telemetry in system tables.

Requirements

  • The Lakebase Observability in System Tables preview enabled for your workspace. A workspace admin can turn it on from the Previews page.
  • A Lakebase project on AWS or Azure.

Available tables

The system.lakebase schema contains the following tables:

Table Description
active_session_history Sampled snapshots of active Postgres backend state, including what each session is running and waiting on.
wait_event_counters Per-window counts and durations for Postgres wait events.
plan_history Captured query executions and their plans.
pg_stat_statements_counters Per-window execution statistics for normalized statements.
ddl_history Captured top-level DDL and schema-change events.
database_counters Per-window Postgres database counter deltas.
database_gauges Point-in-time Postgres database values.
compute_counters Per-window compute and host counter deltas.
compute_gauges Point-in-time compute capacity and health values.
postgres_logs Structured Postgres log events.

Table path: These tables are located in the system.lakebase schema, for example system.lakebase.active_session_history.

For the full column reference for each table, see Lakebase telemetry in system tables.

Retention

The system.lakebase tables retain 7 days of data.

Access

Access to system tables is governed by Unity Catalog. Users with both the account admin and metastore admin roles can read the system.lakebase tables and manage access to them. See Manage privileges in Unity Catalog.

Access applies at the schema level, so a user who can read system.lakebase sees telemetry for every project in the account. A user with access queries the telemetry the way they would any Unity Catalog table.