A tour of the console
What each screen in the VS-HCI console is for, and which one to reach for.
The console is the web interface to the whole deployment. It is served over HTTPS on port 443 at the cluster’s virtual IP, from whichever controller is currently active.
Everything in it is the REST API underneath — there is nothing the console can do that the API cannot, and nothing it hides from the CLI.
The layout
A fixed left-hand navigation, grouped by what each screen acts on. The groups follow the shape of the platform rather than the shape of the menu:
| Screen | Acts on | Page |
|---|---|---|
| Dashboard | The fleet, at a glance | Dashboard |
| Infrastructure | Servers, availability zones, datacenters | Infrastructure |
| Approvals | Nodes waiting to join | Approvals |
| VMs | Virtual machines | Virtual machines |
| Containers | Container workloads | Containers |
| Kubernetes | Managed clusters | Kubernetes |
| Networks | Uplinks, switches, fabrics, routers | Networking |
| Storage | Pools, volumes, replicated storage, exports | Storage |
| Tasks | Asynchronous work in flight | Tasks |
| Alerts | Rules and the events they raised | Alerts |
| Backups | Control-plane backup and restore | Backup and restore |
| Software | Versions, packages, upgrades | Updating |
| Settings | Everything administrative | Settings |
Things that behave the same everywhere
Work happens as tasks
Anything that takes more than an instant — creating a VM, migrating one, building a storage cluster, upgrading a node — becomes a task. The console returns immediately with the task created, and progress streams in live over a WebSocket.
This means a slow operation never blocks the screen, and closing the browser does not cancel anything. See Tasks.
The active controller serves writes
The console is served by whichever controller holds the virtual IP. If you reach a standby directly — by browsing to a node’s own address rather than the VIP — reads work, but a write returns HTTP 409 naming the active node instead of committing.
Use the VIP or the console name and this never arises.
Permissions are enforced by the API
What you can see and do is decided server-side, per domain. A user with read-only rights over networking sees networking and is refused when they try to change it — the button may be there, the answer is still no. See Users and roles.
Cluster-only features hide themselves
On a single-node deployment the screens that only mean something with peers — Approvals, controller HA, live migration — are not shown. They appear when the deployment becomes a cluster.
Your own account
The Profile screen holds your password, your two-factor enrolment and your API keys. Anything that affects other people is under Settings instead.
Keyboard and search
The sidebar has a filter box that narrows the navigation as you type — faster than hunting through groups once the deployment has a lot in it.