-
Story
-
Resolution: Unresolved
-
Undefined
-
None
-
None
-
None
-
Product / Portfolio Work
-
False
-
-
False
-
Not Selected
-
NEW
-
NEW
-
-
We would like to add telemetry metrics for the kubernetes-mcp-server component
to track MCP tool usage across OpenShift clusters.
-
- Metrics to Add
| Metric Name | Labels | Purpose | Cardinality |
| ------------- | -------- | --------- | ------------- |
| cluster:mcp_tool_calls:sum | (none) | Track total MCP tool invocations | 1 |
| cluster:mcp_tool_errors:sum | (none) | Track tool execution errors | 1 |
| cluster:mcp_tool_error_rate:ratio | (none) | Monitor error rate for reliability | 1 |
| cluster:mcp_server_info:count | (none) | Track deployment count | 1 |
-
- Total Timeseries: 4
-
- Use Cases
1. *Adoption tracking*: Understand how many clusters use MCP tools
2. *Reliability monitoring*: Track error rates across the fleet
3. *Capacity planning*: Understand tool usage patterns
-
- Recording Rules
Recording rules are defined in the kubernetes-mcp-server Helm chart:
https://github.com/containers/kubernetes-mcp-server/blob/main/charts/kubernetes-mcp-server/templates/prometheusrule.yaml
-
- Privacy
These metrics contain NO personally identifiable information (PII).
- No user identifiers
- No cluster names
- No tool arguments or parameters
- Only aggregate counts
-
- Component Details
- Repository: https://github.com/containers/kubernetes-mcp-server
- Namespace: User-defined (requires openshift.io/cluster-monitoring=true label)
- ServiceMonitor: Included in Helm chart