infraweaver/console
The console
Once ArgoCD settles, the platform is driven from a web console rather than a pile of kubectl aliases. Six views cover almost everything you do in a week.
SYNTHETIC DATA The live console sits behind SSO, so nothing here is a capture of a running system. These views are rebuilt at production fidelity from the real routes and the real data shapes: node names, sync states, policy names and secret paths are all things the console genuinely renders.
What each view is for
OVERVIEW/home
The morning screen. Node readiness, how many Applications are synced, volume count, open advisories and endpoint health, then the Application list itself and a capacity trace. If something went wrong overnight it is visible here without opening anything else.
CLUSTER/cluster
Nodes, etcd membership, per-node pressure and where workloads actually landed. This is the view you use before draining a node, and the one that tells you whether the cluster can survive losing one.
CATALOG/apps
What is installed, what has an update waiting, and what is available. Installing writes the Application manifest and pushes it to Onedev — the console does not apply anything behind GitOps' back.
GITOPS/gitops-diff
The difference between what the cluster is running and what git says it should run, as a real diff, before you approve the sync. Plus the commit history that produced it.
SECRETS/secrets
Paths, sync state and age — never values. Age matters: the view flags anything past ninety days so rotation is a decision rather than an oversight.
SECURITY/security
Which admission policies are enforcing and what they have blocked, image advisories with the version that fixes them, and the network posture per namespace.
And the routes that did not fit in six tabs
The console is wider than this tour. A partial list of what else is in there: