Temporal Cloud metrics reference
A metric is a measurement or data point that provides insights into the performance and health of a system. This document describes the metrics available on the Temporal Cloud platform. Temporal Cloud metrics help you monitor performance and troubleshoot errors. They provide insights into different aspects of the Service.
This guide covers these topics:
- Gather metrics: Capture Temporal Cloud metrics using a Prometheus-style endpoint so they can be visualized using Prometheus and Grafana or exported to observability platforms like Datadog.
- Available Temporal Cloud metrics: The metrics emitted by Temporal Cloud include counts of gRPC errors, requests, successful task matches to a poller, and more.
- Metrics labels: Temporal Cloud metrics labels can filter metrics and help categorize and differentiate results.
- Operations: An operation is a special type of label that categorizes the type of operation being performed when the metric was collected.
This document discusses metrics emitted by Temporal Cloud. Temporal SDKs also emit metrics, sourced from Temporal Clients and Worker processes. You can find information about Temporal SDK metrics on its dedicated page.
Please note:
- SDK metrics start with the phrase
temporal_. - Temporal Cloud metrics start with
temporal_cloud_.
Gather metrics
How do you capture and use Temporal Cloud metrics?
Temporal Cloud emits metrics in a Prometheus-supported format. Prometheus is an open-source toolkit for alerting and monitoring. The Temporal Service exposes Cloud metrics with a Prometheus HTTP API endpoint. Temporal Cloud metrics provide a compatible data source for visualizing, monitoring, and observability platforms like Grafana and Data Dog.
You can use functions like rate or increase to calculate the rate of increase for a Temporal Cloud metric:
rate(temporal_cloud_v0_frontend_service_request_count[$__rate_interval])
Or you might use Prometheus to calculate average latencies or histogram quartiles:
# Average latency
rate(temporal_cloud_v0_service_latency_sum[$__rate_interval])
/ rate(temporal_cloud_v0_service_latency_count[$__rate_interval])
# Approximate 99th percentile latency broken down by operation
histogram_quantile(0.99, sum(rate(temporal_cloud_v0_service_latency_bucket[$__rate_interval])) by (le, operation))
Set up Grafana with Temporal Cloud observability to view metrics by creating or getting your Prometheus endpoint for Temporal Cloud metrics and enabling SDK metrics.
Available Temporal Cloud metrics
What metrics are emitted from Temporal Cloud?
The following metrics are emitted for your Namespaces:
Frontend Service metrics
temporal_cloud_v0_frontend_service_error_count
This is a count of gRPC errors returned aggregated by operation.
temporal_cloud_v0_frontend_service_request_count
This is a count of gRPC requests received aggregated by operation.
temporal_cloud_v0_resource_exhausted_error_count
gRPC requests received that were rate-limited by Temporal Cloud, aggregated by cause.
temporal_cloud_v0_state_transition_count
Count of state transitions for each Namespace.
temporal_cloud_v0_total_action_count
Approximate count of Temporal Cloud Actions.
Poll metrics
temporal_cloud_v0_poll_success_count
Tasks that are successfully matched to a poller.
temporal_cloud_v0_poll_success_sync_count
Tasks that are successfully sync matched to a poller.
temporal_cloud_v0_poll_timeout_count
When no tasks are available for a poller before timing out.
Replication lag metrics
temporal_cloud_v0_replication_lag_bucket
A histogram of replication lag during a specific time interval for a multi-region Namespace.
temporal_cloud_v0_replication_lag_count
The replication lag count during a specific time interval for a multi-region Namespace.
temporal_cloud_v0_replication_lag_sum
The sum of replication lag during a specific time interval for a multi-region Namespace.
Schedule metrics
temporal_cloud_v0_schedule_action_success_count
Successful execution of a Scheduled Workflow.
temporal_cloud_v0_schedule_buffer_overruns_count
When average schedule run length is greater than average schedule interval while a buffer_all overlap policy is configured.
temporal_cloud_v0_schedule_missed_catchup_window_count
Skipped Scheduled executions when Workflows were delayed longer than the catchup window.
temporal_cloud_v0_schedule_rate_limited_count
Workflows that were delayed due to exceeding a rate limit.
Service latency metrics
temporal_cloud_v0_service_latency_bucket
Latency for SignalWithStartWorkflowExecution, SignalWorkflowExecution, StartWorkflowExecution operations.
temporal_cloud_v0_service_latency_count
Count of latency observations for SignalWithStartWorkflowExecution, SignalWorkflowExecution, StartWorkflowExecution operations.
temporal_cloud_v0_service_latency_sum
Sum of latency observation time for SignalWithStartWorkflowExecution, SignalWorkflowExecution, StartWorkflowExecution operations.
Workflow metrics
temporal_cloud_v0_workflow_cancel_count
Workflows canceled before completing execution.
temporal_cloud_v0_workflow_continued_as_new_count
Workflow Executions that were Continued-As-New from a past execution.
temporal_cloud_v0_workflow_failed_count
Workflows that failed before completion.
temporal_cloud_v0_workflow_success_count
Workflows that successfully completed.
temporal_cloud_v0_workflow_terminate_count
Workflows terminated before completing execution.
temporal_cloud_v0_workflow_timeout_count
Workflows that timed out before completing execution.
Metrics labels
What labels can you use to filter metrics?
Temporal Cloud metrics include key-value pairs called labels in their associated metadata. Labels help you categorize and differentiate metrics for precise filtering, querying, and aggregation. Use labels to specific attributes or compare values, such as numeric buckets in histograms. This added context enhances the monitoring and analysis capabilities, providing deeper insights into your data.
Metrics for all Namespaces in your account are available from the metrics endpoint. Use the following labels to filter metrics:
| Label | Explanation |
|---|---|
le | Less than or equal to (le) is used in histograms to categorize observations into buckets based on their value being less than or equal to a predefined upper limit. |
operation | This includes operations such as:
|
resource_exhausted_cause | Cause for resource exhaustion. |
task_type | Activity, Workflow, or Nexus. |
temporal_account | Temporal Account. |
temporal_namespace | Temporal Namespace. |
temporal_service_type | Frontend or Matching or History or Worker. |
is_background | This label on temporal_cloud_v0_total_action_count indicates when actions are produced by a Temporal background job, for example: hourly Workflow Export. |
namespace_mode | This label on temporal_cloud_v0_total_action_count indicates if actions are produced by an active vs a standby Namespace. For a regular Namespace, namespace_mode will always be “active”. |
The following is an example of how you can filter metrics using labels:
temporal_cloud_v0_poll_success_count{__rollup__="true", operation="TaskQueueMgr", task_type="Activity", temporal_account="12345", temporal_namespace="your_namespace.12345", temporal_service_type="matching"}
Operations
What operation labels are captured by Temporal Cloud?
Operations are a special class of metrics label. They describe the context during which a metric was captured. Temporal Cloud includes the following operations labels:
- AdminDescribeMutableState
- AdminGetWorkflowExecutionRawHistory
- AdminGetWorkflowExecutionRawHistoryV2
- AdminReapplyEvents
- CountWorkflowExecutions
- CreateSchedule
- DeleteSchedule
- DeleteWorkflowExecution
- DescribeBatchOperation
- DescribeNamespace
- DescribeSchedule
- DescribeTaskQueue
- DescribeWorkflowExecution
- GetWorkerBuildIdCompatibility
- GetWorkerTaskReachability
- GetWorkflowExecutionHistory
- GetWorkflowExecutionHistoryReverse
- ListBatchOperations
- ListClosedWorkflowExecutions
- OperatorDeleteNamespace
- PatchSchedule
- PollActivityTaskQueue
- PollNexusTaskQueue
- PollWorkflowExecutionHistory
- PollWorkflowExecutionUpdate
- PollWorkflowTaskQueue
- QueryWorkflow
- RecordActivityTaskHeartbeat
- RecordActivityTaskHeartbeatById
- RegisterNamespace
- RequestCancelWorkflowExecution
- ResetStickyTaskQueue
- ResetWorkflowExecution
- RespondActivityTaskCanceled
- RespondActivityTaskCompleted
- RespondActivityTaskCompletedById
- RespondActivityTaskFailed
- RespondActivityTaskFailedById
- RespondNexusTaskCompleted
- RespondNexusTaskFailed
- RespondQueryTaskCompleted
- RespondWorkflowTaskCompleted
- RespondWorkflowTaskFailed
- SignalWithStartWorkflowExecution
- SignalWorkflowExecution
- StartBatchOperation
- StartWorkflowExecution
- StopBatchOperation
- TerminateWorkflowExecution
- UpdateNamespace
- UpdateSchedule
- UpdateWorkerBuildIdCompatibility
- UpdateWorkflowExecution
As the following table shows, certain metrics groups support operations for aggregation and filtering:
| Metrics Group / Operations | All Operations | SignalWithStartWorkflowExecution / SignalWorkflowExecution / StartWorkflowExecution | TaskQueueMgr | CompletionStats |
|---|---|---|---|---|
| Frontend Service Metrics | X | |||
| Service Latency Metrics | X | |||
| Poll Metrics | X | |||
| Workflow Metrics | X |