# Kubernetes Source: https://docs.ops0.com/docs/kubernetes/overview Monitor, manage, and troubleshoot your Kubernetes clusters with AI-powered insights. ![ops0 Kubernetes Resource Graph showing cluster topology](https://docs.ops0.com/images/resource-graph-k8s.png) ## What is Kubernetes in ops0? ops0 provides a unified view of all your Kubernetes clusters with real-time monitoring, incident detection, and AI-assisted troubleshooting. Connect clusters via the ops0 agent and get instant visibility into workloads, resources, and issues. - **Visibility**: Keep all clusters, workloads, and health signals in one operator view. - **Detection**: Surface incidents, rollouts, and resource pressure before they turn into outages. - **Resolution**: Move from symptom to root cause with logs, events, dependencies, and AI guidance. ## Key Features - **Cluster Dashboard**: Real-time view of cluster health, node status, resource usage, and running workloads. - **Incident Detection**: Automatic detection of pod crashes, resource pressure, failed deployments, and misconfigurations. - **Resource Graph**: Visual dependency graph with incident severity overlays (P1/P2/P3). - **AI Troubleshooting**: Root cause analysis and remediation suggestions with workload context attached. - **Deploy to Cluster**: Deploy Helm charts and manifests with per-file configuration and planning. - **Cost Analysis**: Per-namespace cost breakdown with CPU, memory, GPU, PV, and network costs. --- ## Resource Graph The Kubernetes resource graph provides a visual dependency map for a cluster: - Navigate to a cluster and click **Resource Graph** - See deployments, services, pods, configmaps, secrets, and their connections - Incidents are overlaid on the graph with severity indicators (P1 Critical, P2 High, P3 Medium) - Click any node to view resource details and drill into logs or events --- ## Deploy to Cluster Deploy Helm charts or Kubernetes manifests directly from ops0: | Feature | Description | |---------|-------------| | **Per-file configuration** | Configure each manifest or Helm values file individually | | **Helm support** | Deploy and manage Helm releases | | **kubectl apply** | Apply raw manifests to the cluster | | **Deployment planning** | Preview changes before applying | | **Outputs** | View deployment outputs and applied resource status | ## How It Works ### Connect Install ops0 agent in your cluster ### Monitor View real-time cluster status and metrics ### Detect Get alerted when incidents occur ### Resolve Use AI to diagnose and fix issues ## Cluster Status Clusters show real-time health status: | Status | Meaning | | --- | --- | | Healthy | All major cluster checks are passing and workloads are behaving normally | | Warning | Minor issues are present, such as resource pressure or degraded rollouts | | Critical | Immediate operator action is likely required | | Offline | The cluster is not connected or has stopped reporting | ## Incident Severities Incidents are categorized by severity: | Severity | Typical meaning | | --- | --- | | Critical | Service down, data-loss risk, security exposure, or widespread workload failure | | Warning | Degraded performance, rollout risk, or resource pressure that needs review | | Info | Changes, scaling events, or other signals that are useful context but not urgent | ## Supported Resources ops0 monitors all standard Kubernetes resources: | Category | Resources | |----------|-----------| | **Workloads** | Deployments, StatefulSets, DaemonSets, Jobs, CronJobs, ReplicaSets, Pods | | **Networking** | Services, Ingress, NetworkPolicies, Endpoints | | **Storage** | PersistentVolumes, PersistentVolumeClaims, StorageClasses, ResourceQuotas | | **Config** | ConfigMaps, Secrets, ServiceAccounts, Certificates | | **Scaling** | HorizontalPodAutoscalers, PodDisruptionBudgets, LimitRanges | | **RBAC** | Roles, ClusterRoles, RoleBindings, ClusterRoleBindings | | **Namespaces** | Create, view, and delete namespaces | | **Custom** | All CRDs installed in the cluster | ## Quick Start - [Connect a Cluster](https://docs.ops0.com/docs/kubernetes/clusters): Install ops0 agent and connect your first cluster. - [View Incidents](https://docs.ops0.com/docs/kubernetes/incidents): Monitor and respond to cluster issues with more context. - [Explore Resource Graph](https://docs.ops0.com/docs/resource-graph): Visualize resource relationships and dependencies. - [Deploy Manifests](https://docs.ops0.com/docs/configurations/deployments): Deploy Kubernetes configurations with ops0. --- ## Example: Investigating a Production Incident Here's how ops0 helps you troubleshoot a production issue: **1. Incident Detected** Critical Incident 2 minutes ago CrashLoopBackOff: api-gateway-7d9f8c6b4d-2xkjp Cluster: production-eks • Namespace: api-gateway **2. View in Resource Graph** The Resource Graph highlights the affected pod and its dependencies: - Pod `api-gateway-7d9f8c6b4d-2xkjp` shows red border - Connected Service, ConfigMap, and Secret are visible - Related Deployment shows warning status **3. AI Analysis** ``` Root Cause Analysis: ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ The pod is failing to start due to a missing environment variable DB_PASSWORD. The Secret 'api-gateway-secrets' exists but is missing the 'db-password' key. Last successful deployment: 3 hours ago Recent change: Secret 'api-gateway-secrets' was updated 45 minutes ago (removed db-password key) Recommended Actions: 1. Add 'db-password' key to Secret 'api-gateway-secrets' 2. Or update Deployment to reference correct Secret key ``` **4. Resolution** After fixing the Secret: Resolved Just now Pod successfully started 3/3 replicas running • All health checks passing --- ## Example: Cluster Dashboard View ``` ┌─────────────────────────────────────────────────────────┐ │ production-eks ● Healthy │ │ AWS EKS 1.28 • us-east-1 • 12 nodes │ ├─────────────────────────────────────────────────────────┤ │ │ │ Nodes Pods CPU Memory │ │ ━━━━━━━━━━━ ━━━━━━━━━━━ ━━━━━━━ ━━━━━━━ │ │ 12/12 Ready 156/200 42% 61% │ │ ● ● ● ● ● ████████░░ ████░░░ ██████░ │ │ ● ● ● ● ● │ │ ● ● │ │ │ │ Recent Incidents │ │ ───────────────────────────────────────────────────── │ │ ● Warning High memory usage on node-7 15m ago │ │ ● Info HPA scaled api-gateway 3→5 1h ago │ │ ● Resolved CrashLoop fixed 2h ago │ │ │ │ Top Namespaces by Pod Count │ │ ───────────────────────────────────────────────────── │ │ api-gateway ████████████████████ 45 │ │ web-frontend ████████████ 28 │ │ backend ████████ 19 │ │ monitoring ██████ 14 │ │ │ └─────────────────────────────────────────────────────────┘ ```