Docs / Infrastructure
Kubernetes clusters
Browse and operate your managed and on-prem Kubernetes clusters: workloads, networking, config, storage and RBAC.
Once you connect a cloud account, Pangea finds the Kubernetes clusters in it. You can browse and operate them yourself, or hand the work to an agent.
Supported clusters
| Provider | Service |
|---|---|
| AWS | EKS |
| Google Cloud | GKE |
| Azure | AKS |
| Tencent Cloud | TKE |
| Alibaba Cloud | ACK |
| Volcengine | VKE |
| Linode | LKE |
| Your own data centre | On-prem clusters, reached through the Pangea relay |
Every cluster opens in the same browser with the same pages. The provider only decides how Pangea gets the credential.
On-prem clusters
Some clusters have an API server that Pangea cannot reach from the internet. Pangea reaches these through a relay. You install a small agent inside the cluster, and it dials out to Pangea. You open nothing inbound.
The setup is tunnel-first. You install the agent and wait for it to connect before you hand over any credential. See On-prem Kubernetes.
Connecting to a cluster
Open a cluster from its account and Pangea provisions access for you. You don't need to manage a kubeconfig by hand.
Cloud access follows the connected identity. Cluster tokens and kubeconfigs have their own lifetimes and RBAC permissions. A cloud IAM grant is not always an in-cluster grant. See IAM trust model.
Pangea issues kubeconfigs and cluster tokens only to Admins and Members. Viewers can't get them. See Access control.
What you can browse
The cluster sidebar groups resources like this:
- Cluster: Events, Overview and Nodes, with per-node status and utilization.
- Workloads: Helm Releases, Deployments, Pods, ReplicaSets, StatefulSets, DaemonSets, Jobs and CronJobs.
- Networking: Services, Ingresses, Endpoint Slices and Network Policies.
- Access Control: Service Accounts, Roles, Role Bindings, Cluster Roles and Cluster Role Bindings.
- Configuration: ConfigMaps and Secrets.
- Storage: Storage Classes, Persistent Volumes and PVCs.
- Custom Resources: CRDs and their instances.
Working inside a cluster
- Pod exec: open an interactive terminal inside a pod to debug.
- Logs: stream and search container logs.
- Bulk actions: select several resources and act on them at once.
You can also ask the agent: "Why is the checkout deployment crash-looping?" It reads events, logs and config across the cluster and walks you to the cause.
Private clusters and clusters you run yourself
Use On-prem Kubernetes for the relay setup, the cluster credential and the RBAC checks. The same flow works for clusters you run yourself on cloud VMs.
- Check read access to the namespace you need before you change a workload.
- When you retire access, remove the relay and any RBAC you no longer need.