Deploying Temporal Workers on Google Cloud Run requires specific configuration to handle their long-polling nature. Key requirements include disabling CPU throttling, setting minimum instances, and using sidecar containers for metrics collection. The setup involves configuring OpenTelemetry collector to scrape Temporal metrics and export them to Google Cloud Managed Service for Prometheus. Scaling works best with instance-based billing using CPU utilization metrics, though custom scaling solutions may be needed for complex workloads.
Table of contents
Configuring Your Application #Sidecar Container #Viewing the Metrics #Scaling #Wrapping It All Up #Sort: