Temporal Integration
Monitor workflow execution metrics, activity task latency, and worker throughput across your Temporal clusters. AI-powered anomaly detection for every workflow type.
How It Works
Enable Prometheus Metrics
Configure Temporal Server to expose Prometheus metrics by setting the prometheusEndpoint in your temporal.yaml. All server, frontend, and worker metrics are available immediately.
Instrument Your Workers
Use the TigerOps SDK wrapper or the Temporal SDK's built-in Prometheus reporter to emit worker-side metrics — activity task latency, workflow execution counts, and error rates.
Configure TigerOps Scrape
Add Temporal's Prometheus endpoint to TigerOps's scrape config. Server, frontend, history, and matching service metrics all flow into unified dashboards.
Set SLOs per Workflow Type
Define schedule-to-start and activity execution latency SLOs per workflow type. TigerOps alerts when a specific workflow type begins experiencing elevated latency or failure rates.
What You Get Out of the Box
Workflow Execution Metrics
Per-workflow-type execution counts, success rates, failure rates, cancel rates, and timeout counts. Drill into which workflow types are experiencing elevated errors.
Activity Task Latency
Schedule-to-start and start-to-close latency histograms per activity type and task queue. Identify slow activities and correlate with worker resource exhaustion.
Worker Throughput
Workflow and activity task processing rates per worker, sticky cache hit ratio, and goroutine/thread count. Detect worker saturation before it causes schedule lag.
Schedule Lag Monitoring
Track schedule-to-start lag per task queue. When lag increases, TigerOps identifies whether the bottleneck is worker capacity, activity failure rate, or server-side queuing.
History Service Health
Temporal History service metrics: persistence latency, workflow state machine transitions, timer fire lag, and replication lag for multi-cluster deployments.
AI Workflow Anomaly Detection
TigerOps AI learns normal execution patterns per workflow type. Unusual retry storms, unexpected completion rate drops, or duration regressions trigger automated alerts.
temporal.yaml Metrics Configuration
Enable Prometheus metrics on Temporal Server and configure TigerOps scraping.
# temporal.yaml — Temporal Server configuration with Prometheus metrics
global:
metrics:
prometheus:
framework: tally
listenAddress: "0.0.0.0:9090"
handlerPath: /metrics
timerType: histogram
services:
frontend:
rpc:
grpcPort: 7233
membershipPort: 6933
history:
rpc:
grpcPort: 7234
matching:
rpc:
grpcPort: 7235
worker:
rpc:
grpcPort: 7239
# TigerOps scrape config (add to tigerops-agent ConfigMap):
# scrape_configs:
# - job_name: temporal-server
# static_configs:
# - targets:
# - temporal-frontend.temporal.svc:9090
# - temporal-history.temporal.svc:9090
# - temporal-matching.temporal.svc:9090
# - temporal-worker.temporal.svc:9090
# labels:
# cluster: prod
# - job_name: temporal-workers
# kubernetes_sd_configs:
# - role: pod
# relabel_configs:
# - source_labels: [__meta_kubernetes_pod_label_temporal_worker]
# action: keep
# regex: "true"
# remote_write:
# - url: https://ingest.atatus.net/api/v1/write
# bearer_token: "${TIGEROPS_API_KEY}"Common Questions
Which Temporal versions does TigerOps support?
TigerOps supports Temporal Server 1.17 and later, including Temporal Cloud. The Prometheus metrics endpoint is available in all versions. Temporal Cloud customers can use the metrics export API to stream metrics to TigerOps.
Can TigerOps monitor Temporal Cloud?
Yes. Temporal Cloud supports metric export via the gRPC API or Prometheus scrape endpoint (in supported plans). Configure TigerOps to poll the Temporal Cloud metrics endpoint with your Cloud API key.
How do I get worker-side metrics into TigerOps?
In the Temporal Go and Java SDKs, configure the metrics handler to use the Prometheus reporter and expose a /metrics endpoint on your worker process. TigerOps scrapes this endpoint alongside the server-side metrics for complete visibility.
Can TigerOps alert when a specific workflow type has elevated failures?
Yes. Create a TigerOps alert rule targeting the temporal_workflow_failed_total metric with a workflow_type label filter. The alert fires when the failure rate for that specific type crosses your threshold.
Does TigerOps support distributed Temporal clusters?
Yes. Multi-cluster Temporal deployments expose metrics from each cluster's services. TigerOps aggregates these into a unified view while preserving cluster labels for per-cluster drill-down.
Full Visibility into Every Temporal Workflow
Activity latency SLOs, worker saturation alerts, and AI anomaly detection. Connect in minutes.