Revision Overview
Monitor overall Knative revisions
Info
For accessing CoreWeave Grafana Dashboards instructions, see Access CoreWeave Grafana Dashboards.
The Knative Revision Overview dashboard provides a comprehensive look into the performance and health of a specific service revision. It visualizes key operational metrics like request volume, response times, and error rates. It also offers insights into the autoscaler's behavior.
| Panel | Description |
|---|---|
| Request Volume | Show the overall volume of requests. |
| Success Rate | Shows the success rate of requests. |
| Inference Time | Shows the average inference time. |
| Response Time | Shows the average response time. |
| Pods | Shows the total number of Pods. |
| Pods Actively Serving (2m) | Shows the number of Pods actively serving in the last 2 minutes. |
| Request Volume by Revision | Shows the volume of requests, broken down by service revision. |
| Request Volume by Pod | Shows the volume of requests, broken down by individual Pod. |
| 5xx Errors by Pod | Tracks the count of 5xx server errors, grouped by Pod. |
| Request Volume inter Pod StdDev | Shows the standard deviation of request volume between Pods. |
| Response Time by Revision (Successful with Queue) | Shows the response time (including time spent in queue) for successful requests, grouped by revision. |
| Response Time by Pod | Displays the response time, broken down by individual Pod. |
| Response Time by Response Code Class | Shows response times, categorized by the response code class (e.g., 2xx, 4xx, 5xx). |
| Response Time inter Pod StdDev | Shows the standard deviation of response times between Pods. |
| Observed Concurrency | Shows the observed level of concurrent requests per Pod. |
| Revision Pod Counts | Shows the number of Pods for each service revision over time. |
| Activator Request Concurrency | Shows the concurrent requests being handled by the activator. |
| Pod Queue Depth | Shows the number of requests waiting in the queue for each Pod. |
| Activator Request Count | Shows the total number of requests handled by the activator. |
| Total Pod Queue Depth | Shows the total number of requests waiting in queues across all Pods. |
| Error Code Breakdown | Shows requests by their resulting error codes. |
| Unique Pod count returning errors in same minute | Shows the count of unique Pods that have returned an error within the same minute. |
| Pods Actively Serving (2m) | Shows the number of Pods actively serving requests over a 2-minute window. |