The Capacity Remaining badge shows the percentage of usable capacity that has not yet been consumed, therefore an object with a score of 100 has all of its Capacity Remaining, and vice versa; an object having a score of 0 would indicate that there is no capacity available at that moment. For every cluster, virtual machine, or Datastore, Capacity Remaining allows you to determine which resource (CPU, memory, disk, network) might run out first. The Capacity Remaining can be calculated based on average consumption, or consumption at peak times.
For example, average consumption might be 48% of total capacity. The capacity used for vSphere HA and buffers might be 10% of total capacity. So, the Capacity Remaining is 42% (100 - (10 + 48)). You can replace average consumption with peak consumption in the equation. For example, if peak consumption reached 60% of total capacity, then the Capacity Remaining is 30% (100 - (10 + 60)).
In vCenter Operations, Manager 5.x, Capacity Remaining was based on the number of Virtual Machines as the unit of capacity. In vRealize Operations 6.0, the unit of capacity varies depending on the object type. This is primarily due to the fact that capacity management has been opened up to non-vSphere objects.
It is important to note that Capacity Remaining is based on the resource that is the most constrained. For example, in the following screenshot, we see a vSphere Cluster with CPU, Memory, Disk Space, and Configuration Limits selected as resources to include in Capacity Management as the currently—applied policy.
The badge changes its color based on the badge score thresholds:
- Green (>= 10%): The Capacity Remaining for the object is at a high level
- Yellow (< 10%): The Capacity Remaining for the object is at a medium level
- Orange (< 5%): The Capacity Remaining for the object is at a seriously low level
- Red (= 0%): The object is expected to run out of capacity soon or has already run out of capacity
Looking at the four selected resources, memory is the most constrained at 64.53% remaining after buffers are taken into account, therefore this is where the Capacity Remaining score is derived from for this object:
For each resource, the Capacity Breakdown panel provides the following details:
- Total capacity configured for the resource
- Percentage of resource capacity used for buffers
- Amount of usable capacity consumed by the object
- Percentage of usable capacity consumed during peak times
- Percentage of usable capacity demanded during non-stress times (stress accumulates when a workload exceeds the stress line)
- Percentage of usable capacity remaining for the resource
The Capacity Breakdown panel also indicates the number of average, small, medium, and large virtual machine profiles that can fit into the capacity remaining for this object. In addition to the default profiles, you can create custom profiles based on specific sizing requirements in your environment. Custom profiles show how many more of a specified object can fit in your environment given the available capacity and object configuration.
When creating capacity projects, you can use custom profiles in addition to the default profile. We will cover capacity projects later in this book.
The vSphere Configuration Limit row shows the remaining virtual machine count (at the bottom–right of the panel). This count represents the number of virtual machines that can be deployed on the selected object. This value is based on the current amount of unused resources in a cluster and the average virtual machine profile.