Skip to main content

Monitor Temporal Cloud

View Markdown

Monitor Temporal Cloud and the Workers connected to it to detect service issues, Task Queue backlogs, and Worker capacity problems.

Temporal provides Cloud metrics, SDK metrics, health-monitoring guidance, and notifications. Use the following guides based on what you need to detect.

Choose what to monitor

GoalGuideWhat it helps you detect
Collect metrics from Temporal Cloud and your WorkersSet up metricsService, Task Queue, Worker, and Client behavior
Monitor Temporal Cloud service healthMonitor service healthService latency and errors, Workflow and Activity failure trends, throttling, and capacity limits
Monitor Workers connected to Temporal CloudMonitor Worker healthTask backlogs, insufficient or excess Worker capacity, configuration problems, and Worker availability
Receive operational updatesReceive notificationsTemporal Cloud incidents, expiring credentials, billing events, and completed or failed failovers

Understand metric sources

Temporal provides two complementary metric sources:

  • Cloud metrics show what the Temporal Service observes, including request latency, service errors, Task Queue behavior, throttling, and Namespace-level activity.
  • SDK metrics show what your Workers and Clients observe, including Schedule-To-Start latency, available Worker slots, cache behavior, and request failures.

Use both sources to distinguish a Temporal Cloud service issue from a Worker or application issue.

To begin collecting metrics:

  1. Set up Cloud metrics with OpenMetrics.
  2. Connect Temporal Cloud to your observability tool.
  3. Set up SDK metrics.
  4. Use the service health and Worker health guides to create monitors and alerts.

Monitor Nexus

To monitor Nexus metrics and debug Nexus Operations, see Temporal Nexus observability.

For records of Control Plane changes, see Audit Logs. Audit Logs answer who changed a Temporal Cloud resource, what they changed, and when. They are separate from runtime health monitoring.