K8s Cluster Management

Console: Admin Console
Route: /camp/k8s-dashboard
Primary Users: Platform operations, SRE, cluster administrators

K8s Cluster Management is designed for platform administrators and operations engineers, providing native Kubernetes resource management capabilities for the current cluster. Through this module, you can directly view and operate nodes, GPUs, Pods, Deployments, Services, Ingresses, ConfigMaps, PVCs, Namespaces, and other K8s resources without logging in to the cluster command line.

What Can I Do

Scenario	Capability
View cluster dashboard	View cluster basics, resource utilization, node and GPU status in Resource Overview
Manage GPU scheduling	Configure GPU scheduling policies, overcommit strategies, resource pools, and licenses in Compute Configuration
Troubleshoot workloads	View running status of Pods, Deployments, StatefulSets, Jobs, and more in Workloads
Manage networking and storage	Manage Ingresses/Services in Traffic; manage PVCs/PVs/StorageClasses in Storage
Manage configuration and security	Manage ConfigMaps/Secrets/HPAs in Container Configuration; manage RBAC in Security
Manage namespaces and events	View Namespaces, Events, CRDs, and system components in Others

Sub-modules

Sub-module	Doc	Description
Resource Overview	Resource Overview	Cluster basics, resource utilization, node and GPU management
Compute Configuration	Compute Configuration	GPU scheduling policies, overcommit strategies, virtual devices, resource pools, licenses
Workloads	Workloads	Pod, Deployment, StatefulSet, DaemonSet, Job, CronJob
Traffic	Traffic	Ingress, Service
Storage	Storage	PVC, PV, StorageClass
Container Configuration	Container Configuration	ConfigMap, Secret, HPA
Security	Security	ServiceAccount, Role, RoleBinding, ClusterRole, ClusterRoleBinding
Others	Others	Namespace, Event, CRD, system components

General Operations

Most K8s resource management pages support the following common operations:

1. Search and Filter: Filter by name, namespace, status, labels, and other criteria.
2. Column Visibility/Sorting/Pagination: Customize table display with cluster-local persistent configuration.
3. Export CSV: Export the current list to a local file.
4. YAML Editing: Edit YAML online in the resource detail page and save.
5. Describe Modal: View kubectl describe output.
6. Change History and Rollback: View resource modification history, compare YAML differences, and roll back to a previous version with one click.
7. Event Viewing: View resource-related events.
8. Batch Delete: Select multiple rows for batch deletion; important resources require name confirmation.

Typical Workflow

1. Navigate to Resource Overview to view overall cluster health and node/GPU status.
2. In Compute Configuration, configure GPU scheduling policies, overcommit strategies, and resource pools.
3. In Workloads, troubleshoot running status of Pods, Deployments, and more.
4. In Traffic, view Ingress and Service configurations.
5. In Storage, manage PVCs/PVs/StorageClasses.
6. In Container Configuration, Security, and Others, manage ConfigMaps, Secrets, RBAC, Namespaces, and more.