Render dashboards
Tip: Make sure you have completed the Render setup and have a working Metrics Stream destination before continuing with this section.
Once a Metrics Stream is set up, these dashboards are generated automatically as Automated dashboards. No further action is required. Depending on the Render services your workspace uses, you may see:
HTTP dashboard
The HTTP dashboard shows request volume and latency for Render web services. Render aggregates these metrics across all instances of a service, so each line corresponds to a host and status code combination across the whole service rather than a specific replica.
HTTP requests: Number of HTTP requests received, broken down by host and status code. It uses the following metric: render.service.http.requests.total.
HTTP latency: Response latency at the p50, p95 and p99 percentiles, broken down by host and status code. It uses the following metric: render.service.http.requests.latency.
Postgres dashboard
The Postgres dashboard shows information about Render Postgres instances connected to your workspace.
Connections: Active connections per database, compared against the connection limit. It uses the following metrics: render.postgres.connections and render.postgres.connection.limit.
Storage size: Database size and indexes size per database. It uses the following metrics: render.postgres.database.size and render.postgres.indexes.size.
Transaction volume: Number of transactions over time. It uses the following metric: render.postgres.transaction.volume.
Sequential table scans: Sequential scans per database. A high number of sequential scans can indicate missing indexes. It uses the following metric: render.postgres.table.scans.
Transaction exhaustion: Per-database transaction ID exhaustion, useful for monitoring vacuuming health. It uses the following metric: render.postgres.transaction.exhaustion.
Slow locks: Number of locks that exceeded the slow threshold. It uses the following metric: render.postgres.slow.lock.count.
Slow lock wait time: Time spent waiting on slow locks. It uses the following metric: render.postgres.slow.lock.time.
Replication lag: Replication delay per replica host. It uses the following metrics: render.postgres.replication.lag and render.postgres.replication.apply.lag.
Key Value dashboard
The Key Value dashboard shows connection metrics for Render Key Value (Valkey/Redis) services.
Connections: Active connections, compared against the connection limit. It uses the following metrics: render.keyvalue.connections and render.keyvalue.connection.limit.
CPU, memory, disk and network usage for Key Value services are reported as host metrics and appear in the "host metrics" section instead of on this dashboard.
