Dashboard Cards

KubeStellar Console provides a variety of cards to monitor and manage your clusters.

Available Card Types

Card Type	Description	Data Source
Cluster Health	Availability graph per cluster	`get_cluster_health`
App Status	Multi-cluster application health	`get_app_status`
Event Stream	Live event feed with filtering	`get_events`
Deployment Progress	Rollout status visualization	`get_app_status`
Pod Issues	CrashLoopBackOff, OOMKilled pods	`find_pod_issues`
Deployment Issues	Stuck or failing rollouts	`find_deployment_issues`
Resource Capacity	CPU/memory/GPU utilization	`list_cluster_capabilities`
GPU Inventory	GPU nodes and counts across clusters	`get_gpu_nodes`
GPU Status	Real-time GPU allocation and usage	`get_gpu_nodes`
GPU Overview	Summary of GPU resources	`get_gpu_nodes`
Security Issues	Privileged, root, host access	`check_security_issues`
Upgrade Status	Cluster version and upgrade state	`get_upgrade_status`

GPU Cards

GPU Inventory

Shows all GPU nodes across clusters with:

Node name and cluster
GPU type (e.g., NVIDIA A100)
Total GPU count
Allocated vs available

GPU Status

Real-time GPU utilization:

Allocation percentage per cluster
Memory usage
Temperature (if available)

GPU Overview

Summary card showing:

Total GPUs across all clusters
Overall utilization
Top consumers

Card Configuration

Each card can be configured with:

interface CardConfig {
  id: string;
  type: string;
  title: string;
  position: { x: number; y: number };
  size: { width: number; height: number };
  config: {
    clusters?: string[];      // Filter to specific clusters
    namespaces?: string[];    // Filter to specific namespaces
    refreshInterval?: number; // Update frequency in seconds
    warningsOnly?: boolean;   // For event stream
  };
}

AI Recommendations

In High AI mode, the console analyzes your cluster state and suggests relevant cards:

Pod Issues - Suggested when >5 pods have issues
GPU Status - Suggested when GPU utilization >90%
Event Stream - Suggested when >10 warning events
Cluster Health - Suggested when clusters are unhealthy

You can:

Accept - Add the recommended card
Snooze - Hide suggestion for 1 hour
Dismiss - Don’t suggest this card again