Service View

The Service View in Kerno provides detailed insights into a specific service, allowing you to monitor performance, investigate issues, and understand dependencies for efficient troubleshooting.

Service View Use-Cases

Use the Service View to:

• Monitor Service Performance:

  • Use the Performance Graph to observe trends in traffic, latency, and errors.

  • Correlate performance dips with events displayed on the graph timeline.

• Analyze Dependencies:

  • Explore the Endpoint Table to identify problematic upstream or downstream paths.

  • Switch to the Map View for a visual representation of service dependencies.

• Inspect Infrastructure Resources:

  • Drill into individual pods in the Infra Resources panel to analyze their resource usage or investigate issues.

• Troubleshoot Events and Issues:

  • Review the Event Timeline to understand how Kubernetes or configuration changes impacted the service.

  • Use the Issues Panel to identify and resolve application-level errors.

Key Components

Performance Graph

The graph offers a comprehensive overview of the service's performance over time, including:

• Requests/Second: Tracks the number of requests handled by the service.

• % of Failed Requests: Shows the percentage of requests that resulted in errors or were rejected.

• Latency: Displays the service's average response time (P90, P50, P95, P99).

• Events: Highlights correlated Kubernetes or change events (e.g., pod restarts, configuration changes) that impacted the service.

Endpoint Table

A detailed table listing all upstream and downstream paths (endpoints) connected to the service, with key metrics:

• Requests/s: The traffic rate for each path.

• Latency: Average response time for requests on the path.

• Error %: The percentage of requests that failed.

• Rejected %: The percentage of requests rejected by the service.

Map View

The Map View visualizes the service's interactions with other components in the cluster. This service map visually represents upstream and downstream dependencies, helping you understand the service's role within the architecture.

Infrastructure Resources The Infra Resources panel lists the pods running the service and provides their current status. Clicking on a pod reveals detailed information, including:

• Resource Usage: CPU and memory consumption over time.

• Pod Metadata: Name, namespace, IP, creation timestamp, and conditions.

Event Timeline

Displays a chronological list of Kubernetes and change events associated with the service. For example, pod terminations or restarts are shown here, allowing you to correlate events with performance changes.

Issues

Highlights detected issues within the service, such as application errors or exceptions, making it easy to identify potential root causes.

If you encounter issues or have questions, message us on Slack, and we’ll gladly help.