Skip to content

temporal-workflow-metrics.md

Latest commit

 

History

History
32 lines (21 loc) · 1.73 KB

temporal-workflow-metrics.md

File metadata and controls

32 lines (21 loc) · 1.73 KB
Raw

Temporal Workflow Metrics

Thermos uses Temporal Cloud to orchestrate background workflows — auth refresh, trigger processing, cleanup, and MCP toolkit sync. Each workflow emits execution metrics to Datadog (via StatsD) and optionally to OpenTelemetry.

Metric Overview

All workflows emit three core metrics:

Metric Type Description
thermos.workflow.execution.total Counter Incremented on every workflow run. Tagged with workflow:<name> and status:success|failure.
thermos.workflow.execution.duration Distribution Execution duration in milliseconds. Tagged with workflow:<name> and status:success|failure.
thermos.workflow.last_success_epoch Gauge Unix epoch of the last successful run. Used for staleness alerting.

Tags

All workflow metrics include:

  • workflow:<name> — workflow identifier (see table below)
  • status:success or status:failure — execution outcome

Some workflows add extra tags (e.g., toolkit:<slug>, trigger_name:<name>).

Additional Domain-Specific Metrics

Beyond the three core workflow metrics, each domain emits its own counters:

  • Auth Refresh: auth_refresh.workflow.{success,failed,skipped} (tagged by app, group)
  • Webhook Triggers: webhook.trigger.{matched,non_matched,tool_error,apollo_error,e2e_latency}
  • Polling Triggers: polling.trigger.runs.{total,chunked,unchunked,duplicate}, polling.trigger.{tool_error,chunk_updates_error,apollo_updates_error,auth_refresh_error}
  • Webhook Trigger Refresh: webhook.trigger_refresh.{total,success,not_required,failed}
  • MCP Toolkit Sync: mcp_toolkit_sync.workflow.{completed,failed,skipped,no_change,no_connected_account}, mcp_toolkit_sync.tools.synced