Dell NativeEdge: Endpoint Metrics are Unavailable Due to High Endpoints Count

On the Endpoints > Endpoint detail > Metrics tab, metrics fail to display.

The issue happens only if there are many endpoints connecting to a NativeEdge Orchestrator (NEO).

The UI shows no error, but it fails to load or display the metrics graphs.

NativeEdge Orchestrator uses Victoria Metrics for its gathering of metrics data. In NEO 3.1, the storage PersistentVolumeClaim (PVC) for where this data is stored (vmstorage) is sized for 50 GB.

Workaround

Users planning to onboard over 50 endpoints in a single NativeEdge Orchestrator (NEO) should update the configuration to ensure metrics display correctly for all endpoints.

Recommended value

• For 50-100 endpoints, change the retention period to 30 days.
• For 100-200 endpoints, change the retention period to 15 days.
• For 200-400 endpoints, change the retention period to 7 days.
• For 400-800 endpoints, change the retention period to 3 days.

On the Native Edge Orchestrator, use kubectl to edit the mist-ee-vmstorage file to change the value of the retention period for Victoria Metrics. The default value is retentionPeriod=12 (12 months):

1. Edit the mist-ee-vmstorage stateful set:
   kubectl edit statefulsets mist-ee-vmstorage -n hzp
2. Using the editor, find the retentionPeriod parameter which is set to retentionPeriod=12, and change its value accordingly such as retentionPeriod=30d for a retention period of 30 days.
3. Save and exit the editor.

This issue is planned to be addressed in a future release of NativeEdge.