All Units: Hosts

All Units with the unit selector set to hosts is the ESXi host portfolio: every host, categorized by CPU and memory utilization, with the hardware profile and build capacity behind each one. It is where overloaded hosts and idle hosts show up next to each other, which is usually the whole point.

What this screen shows

  • Utilization donuts. Hosts distributed into CPU and memory utilization ranges, so the shape of the estate is visible before any sorting.
  • Trend charts. Movement for hosts above the utilization thresholds, which distinguishes a host that has always run hot from one that just started.
  • Table. Host model, serial, and OS, VM counts, CPU and memory totals with average utilization, and VM build capacity.

When to use it

  • Detecting overload. Find hosts that are constrained now, before DRS runs out of room to work with.
  • Detecting the reverse. Underutilized hosts are consolidation candidates and, in a licensing conversation, they are the expensive kind of idle.
  • Balancing and consolidation planning. Both answers come off the same table.

How to work it

  • Filter to hosts above 90% memory. Memory is the usual first constraint, so this filter finds the immediate problems faster than sorting on CPU.
  • Group by cluster. Uneven distribution inside a cluster looks very different from a cluster that is genuinely full, and the fix is completely different too.

Tip: One hot host in an otherwise healthy cluster is a placement problem, not a capacity problem. Check the cluster before approving hardware, because a rebalance is free and a host is not.

ESX Host Summary is the per-host deep dive, including which VMs are resident and could be moved. All Units: Clusters establishes whether the cluster around a hot host has headroom. Cluster Summary confirms the limiting factor.

Last updated: August 24, 2026