Cluster Trends

Cluster Trends is the history for one cluster, and the metric it exists for is VM build capacity over time. A cluster that is constrained today became constrained on a particular date, for a particular reason, and this is the screen that finds it.

What this screen shows

• Trend lines for VM count, active VMs, VM build capacity, and total capacity. Which series appear is configurable.
• Trend table holding CPU, memory, and disk used against capacity across the same period.

Find the zero crossing

The single most useful thing on this screen is the date where the build capacity line crossed zero. That is the moment the cluster stopped having room, and it almost always corresponds to a specific event: a batch of VMs added, a host removed for maintenance and never returned, a reservation change, a failed consolidation.

Finding the date turns an open-ended capacity argument into a specific question with a specific answer, and sometimes the answer is that the headroom can be given back rather than bought.

When to use it

• Understanding consumption rate. How fast this cluster is eating its headroom, which sets how urgent the response is.
• Locating the inflection point. The date build capacity turned negative.
• Supplying forecast inputs. The observed rate is what modeling extends forward.

Tip: Watch VM count and active VMs together. Count rising while active VMs stays flat means the cluster is filling with workloads nobody is using, which is a cleanup job rather than a capacity job.

Related screens

Cluster Summary is the current-state view this adds history to. Cluster Modeling projects the observed rate forward against a proposed purchase. Delta Report quantifies precisely what changed around the inflection point.