Platform Health Dashboard
Health
Overall Platform Health is derived from all of the health metrics in the Platform Health Dashboard.
Default thresholds:
| Name | Healthy Threshold | Degraded Threshold |
|---|---|---|
| Node Status | 80% | 60% |
| Alerts | 80% | 60% |
| Resource Capacity | 80% | 60% |
| Control Plane API | 80% | 60% |
Node Status
"Nodes Online" over "Nodes Available".
Examples:
| Metric | Value | Status |
|---|---|---|
| 5/6 Nodes Operational | 83% | Healthy |
| 3/6 Nodes Operational | 50% | Degraded |
Alerts
Critical and Warning alerts on the cluster.
| Name | Healthy Threshold | Degraded Threshold | Weight |
|---|---|---|---|
| Critical Alerts | 0 | >=1 | 50% |
| Active Warnings | <=5 | >5 | 50% |
Critical Alerts
Alerts that qualify as system breaking or system failure.
For example Istio returning a 500 status which would break user authentication and authorization for some services.
| Metric | Value | Status |
|---|---|---|
| Critical Alerts: 0 | 0 | Healthy |
| Critical Alerts: 1 | 1 | Degraded |
Active Warnings
Warnings that may require attention when seen in large numbers.
For example, a Kubernetes deployment failure.
| Metric | Value | Status |
|---|---|---|
| Active Warnings: 2 | 2 | Healthy |
| Active Warnings: 60 | 60 | Degraded |
Resource Capacity
Overall health base on resource usage across the cluster.
| Name | Healthy Threshold | Degraded Threshold | Weight |
|---|---|---|---|
| Memory | <=80% | >80% | 33% |
| CPU | <=70% | >70% | 34% |
| Storage | <=60% | >60% | 33% |
Memory
Aggregated memory usage on the cluster.
Example:
| Metric | Value | Status |
|---|---|---|
| Memory: 46% | 46% | Healthy |
| Memory: 81% | 81% | Degraded |
CPU
Aggregated CPU usage on the cluster.
Example:
| Metric | Value | Status |
|---|---|---|
| CPU: 17% | 17% | Healthy |
| CPU: 71% | 71% | Degraded |
Storage
Aggregated disk storage usage on the cluster.
Example:
| Metric | Value | Status |
|---|---|---|
| Storage: 54% | 54% | Healthy |
| Storage: 61% | 61% | Degraded |
Control Plane API
Current operational effectiveness of the Kubernetes Control Plane.
| Name | Healthy Threshold | Degraded Threshold | Weight |
|---|---|---|---|
| Latency (ms) | <=100 | >100 | 50% |
| Success Rate | >=95% | <95% | 50% |
Latency
The time it takes for the Kubernetes Control Plane API to receive a request.
| Metric | Value | Status | | Latency: 51 milliseconds | 51 | Healthy | | Latency: 120 milliseconds | 120 | Degraded |
Success Rate
The percentage of successful responses returned by the Kubernetes Control Plane API.
| Metric | Value | Status |
|---|---|---|
| Success Rate: 98% | 98% | Healthy |
| Success Rate: 90% | 90% | Degraded |
Refresh Data
Refreshing the data will send a request to retrieve current data metrics from Prometheus.