Dashboard
Fleet health at a glance: capacity, hosts, storage, networking, alerts and controller state.
The Dashboard is the landing screen. It answers one question — is anything wrong — and gives you the path to whatever is.
The top row
Six counters across the top:
| Counter | Meaning |
|---|---|
| Online Hosts | Servers online, out of the total registered |
| Running VMs | Virtual machines currently running |
| Containers | Container workloads running |
| Storage | Capacity across the fleet |
| Networks | Virtual networks defined |
| Active Alerts | Alert events not yet acknowledged |
A non-zero Active Alerts is the one to look at first.
Capacity
Three meters showing what is committed against what exists:
- vCPU allocation — virtual CPUs assigned to VMs against physical cores.
- Memory allocation — memory assigned against installed memory.
- Live CPU load — what the fleet is actually using right now.
Allocation above 100% is normal for vCPU — virtual machines rarely all run flat out, and oversubscribing CPU is the point of virtualization. Memory is different: allocating more memory than the fleet has means a host that loses a peer may not have anywhere to restart its VMs.
Note — Read allocation alongside live load. High allocation with low load is healthy consolidation. High allocation with high load means you are out of headroom.
Host health
One row per server, each with its status and three utilization bars — CPU, RAM and DISK — plus a temperature reading where the hardware reports one.
The bars change colour as they climb: amber from 75%, red from 90%. Temperature follows the same idea, turning amber at 68 °C and red at 78 °C.
Click a row to open that server — see Infrastructure.
Block storage
A summary of iSCSI exports: targets, storage nodes, LUNs, how much is provisioned, how much is attached, and whether multipath is in play. See Exports.
Networking
Counts of fabrics, switches and routers in the virtual network layer. See Networking.
Running VMs
The virtual machines currently running, with the host each is on. Empty on a new deployment.
Alerts
The most recent alert events, coloured by severity — emergency, critical and error in red, warning in amber, informational in blue. See Alerts.
Recent tasks
The last operations the platform ran, with status and progress. A failed task here is usually the explanation for whatever else looks wrong. See Tasks.
Controller HA
Which controller is ACTIVE, which are STANDBY, and whether each one’s database is in sync. On a single-node deployment this panel is not shown.
This is the panel that matters during an incident: a standby that is alive but not in sync cannot take over. See High availability.
Licensing
The current licence state and what it entitles. See Licensing.
Refresh
The Dashboard polls on its own; you do not need to reload it. Live metrics arrive over a WebSocket, so the numbers move without the page flickering.