Cluster Summary
Cluster Summary is the health assessment for one cluster: utilization, build capacity, the limiting factor, and the FinOps cost view where costs are configured. It is the screen to open before anyone commits new workload to a cluster.
What this screen shows
- Capacity score gauge, VM categories, and overall utilization.
- CPU, memory, and disk breakdown, with build capacity calculated per resource.
- Weeks left to capacity, and DRS status where applicable.
- FinOps costs: operating, core, and memory, when cost data is configured.
Build capacity per resource is the point
Build capacity is calculated separately for CPU, memory, and disk, and the lowest of the three is the real answer. A cluster with room for two hundred more VMs on CPU and eleven on memory has room for eleven. Memory is the limiting factor in most environments, which is why the per-resource breakdown matters more than the headline number.
When build capacity is negative, a constraint has already been exceeded. Identify which resource, then model the mitigation rather than debating the number.
When to use it
- Assessing a cluster before onboarding workload. The placement decision in one screen.
- Identifying the true limiting factor. Not the resource that looks worst on a dashboard, the one that binds first.
- Reviewing VM distribution inside the cluster. The VM category breakdown shows whether consolidation or right-sizing would recover headroom without a purchase.
Tip: Before treating a constrained cluster as a hardware problem, read the VM category distribution. A cluster full of over-allocated VMs has headroom that right-sizing recovers for free, and that is a much easier conversation than a capital request.
Related screens
Cluster Trends shows when the constraint started, which usually explains why. Cluster Modeling turns the constraint into a costed plan. All Units: Hosts checks whether the cluster is genuinely full or simply unbalanced.