Run a few K3s apps without a platform team
Organize a K3s or Kubernetes cluster by environment and app. See nodes, artifacts, releases, Pods, metrics, logs, events and config in one place.
What it is
The Infra console is the second entry on the DevOpsPad rail. Where the Kubernetes explorer shows raw resources, Infra groups the cluster by environment (a connection plus a namespace) and application (a Pod label selector). Choose pro and svc-api, and every page — Deployments, Pods, logs, metrics, events, ConfigMaps, artifacts and releases — is scoped to that service.
It suits teams running a handful of services on one or a few K3s clusters, who want a clear view and a safe release button without running Rancher, Argo CD or a CI server. The interface labels in the Infra console are currently in Chinese.
How it works
- Connect K3s. The 接入 K3s (Connect K3s) wizard reads workloads from a namespace through a saved connection and records their selectors in a local
project.yaml. Nothing is written to the cluster. - Nodes. Ready status, roles, IPs, capacity, OS and Kubernetes version for every node of the environment’s cluster.
- Deployments and Pods. Ready counts per workload, with Pods matched by the full label selector. Lists refresh on demand or every 10 seconds.
- Observe. CPU and memory snapshots from metrics-server, bounded log snapshots including the previous container instance, and UID-scoped events.
- Configure. See the environment variables, mounts and ConfigMaps each container really uses, and edit ConfigMap text with conflict detection. Secrets are shown only as references. Optional host Nginx editing over SSH with
nginx -tand automatic restore on failure. - Release. Register images by digest, then release, scale, restart and roll back Deployments through reviewed plans.
Local files, shared with Git
Environments, applications, artifact indexes, release records and AI run records are plain YAML and JSON in a project directory on your machine. Point it at a Git checkout to share with teammates: the app can show status, pull --ff-only, and commit and push only project files. Kubeconfig files and keychain secrets never leave each person’s machine.
Limits
- Changes are limited to existing Deployments: image release, scale, restart and image rollback.
- No metrics history, live log streaming or
exec. - Lists are capped at 500 resources.
- Git sync is manual. It is not real-time collaboration.
- Fleet-wide releases across many machines are planned, not implemented.
Help articles
What is the Infra console?
The Infra console organizes K3s and Kubernetes apps by environment, with nodes, artifacts, releases, Pods, logs, metrics, events and config on one screen.
Connect K3s workloads to the Infra console
Use the Connect K3s wizard to read workloads from a namespace and map them to an environment and application. Only a local mapping is saved.
Map environments and apps in project.yaml
project.yaml maps each environment to a saved cluster connection and namespace and each application to a Pod label selector. Format, rules and examples.
See every node in a K3s cluster
The Nodes page lists every node of the selected environment's cluster with Ready status, roles, internal and external IPs, CPU, memory, OS and version.
View and edit an app's ConfigMaps
See the environment variables, mounts and ConfigMaps an app actually uses, edit ConfigMap text with conflict checks, and edit host Nginx config over SSH.
Check Pod CPU and memory usage
The Monitoring page shows current CPU (mCPU) and memory (MiB) for an app's Pods from metrics-server, with the sample time. There is no history.
Frequently asked questions
Do I need to install anything in the cluster to use the Infra console?
No controller or agent is required. DevOpsPad uses the Kubernetes API through your kubeconfig. CPU and memory figures need metrics-server, which K3s includes by default.
How does a small team share the Infra project?
Put the project directory in a Git repository. Each person clones it, and the app can view status, pull with fast-forward only, and commit and push project files. Credentials are never shared.
Which workloads can the Infra console change?
Existing Deployments only. StatefulSets and DaemonSets can be viewed but not modified, and workloads cannot be created or deleted.