Observability

Slot provides built-in observability features for your deployments, enabling you to monitor performance metrics and visualize system behavior through integrated Prometheus and Grafana instances.

Overview

The --observability flag enables monitoring capabilities for both Katana and Torii deployments. When enabled, Slot provisions:

• A dedicated Prometheus instance to scrape and store metrics
• A Grafana dashboard for visualization and analysis
• Secure access via username/password authentication

Cost: $10/month flat rate per deployment with observability enabled.

Enabling Observability

On Deployment Creation

Enable observability when creating a new deployment:

# For Katana
slot deployments create <Project Name> katana --observability
 
# For Torii
slot deployments create <Project Name> torii --observability --world 0x3fa481f41522b90b3684ecfab7650c259a76387fab9c380b7a959e3d4ac69f

On Deployment Update

Add observability to an existing deployment:

# For Katana
slot deployments update <Project Name> katana --observability
 
# For Torii
slot deployments update <Project Name> torii --observability

Accessing Dashboards

Once observability is enabled, you can access the monitoring interfaces at the following paths on your deployment URL:

Prometheus

https://<your-deployment-url>/prometheus

Access raw metrics, query using PromQL, and explore time-series data.

Grafana

https://<your-deployment-url>/grafana

Features

• Performance Monitoring: Track request rates, response times, and throughput
• Resource Utilization: Monitor CPU, memory, and network usage
• Debugging: Correlate metrics with application behavior during issues
• Capacity Planning: Analyze trends to plan for scaling needs
• Alerting: Set up alerts based on metric thresholds (via Grafana)

Examples

Enable observability for a production Katana instance:

slot deployments create my-game katana --observability

Add observability to an existing Torii deployment:

slot deployments update my-game torii --observability