Use the Optimize Capacity screen to assess workload status and how much capacity is remaining in data centers across your environment.

Where You Find Optimize Capacity

Click Capacity under Optimize in the left menu. From the Quick Start screen, select Assess Capacity in the second-from-left column.

Note: Double-click on a data center graphic to display the object details screen for the data center.

How the Optimize Capacity Works

The Capacity Optimization and Reclaim features are tightly integrated functions that allows you to assess workload status in data centers across your environment. You can determine time remaining until CPU, memory, or disk space resources run out and realize cost savings when underutilized VMs can be reclaimed and deployed where needed.

When you open the Optimize Capacity page, graphical representations of all the data centers and custom data centers in your environment appear. VMware Cloud on AWS data centers has a unique icon to differentiate it from the other data centers.

By default, they are shown in order of time remaining, beginning from the upper left, where the most constrained data centers appear. To review the status of a data center, click the graphic. The page refreshes to display the following data:
Time Remaining
Time Remaining specifies which clusters are most constrained and displays the criticality of the cluster.
Optimization Recommendations
vRealize Operations shows you the number of reclaimable VMs and the associated cost savings. Click View Reclaimable VMs to navigate to the Reclaim page.
Cluster Utilization

Cluster Utilization displays an interactive graph that shows time remaining by component. You can explore the demand percentage over time by CPU, memory, and disk space or by the most constrained component. By default, the data displayed is for the Demand model. If you have configured the Allocation model, then you can also see the CPU, memory, and disk space time remaining model based on the overcommit ratios that you have set in the policy.

Click the Edit icon to modify the criticality threshold, risk level, and allocation model. These changes affect the selected cluster's policy. Hence, any change that you make here, affects all the clusters under the same policy.

Set the Show History and Show Forecast variables to create the slice of time in which you want to see time-remaining data. The vertical axis of the graph shows the total capacity being used by the current amount of CPU, memory, or disk space respectively. The bold, black line across the top of the graph depicts the historical value of usable capacity. The horizontal axis is the timeline. Vertical lines in the graph are labeled at the bottom of each line. The first vertical dotted line on the left marks the projection calculation start point. The next line is the current date - now. The third vertical marks the date the resource runs out. If a resource has little time remaining, the current date and the date that time runs out may be the same.

vRealize Operations can make recommendations for increasing time remaining based on the data it receives and these recommendations appear at the bottom of the screen. You might see two options: Option 1 shows what you can achieve by reclaiming resources. Option 2 shows the results of adding capacity.

If you choose to reclaim resources, you can run that process immediately by clicking RECLAIM RESOURCES. To see the details or choose additional options before running a reclaim action, review the information provided in the Optimization Recommendations pane and then click VIEW RECLAIMABLE VMS to go to the Reclaim page.

Table 1. Capacity Optimization Options
Option Description

Select a datacenter

Select a data center from the carousel across the top of the page. Information about the datacenter is displayed below.

ALL DATACENTERS | X Toggle: click ALL DATACENTERS on the upper right when you want to switch the view to a filtered list of all data centers. Click X to return to a carousel view of data centers.
View: Filter results to include data centers, custom data centers, or both. This option appears if you select ALL DATACENTERS on the upper right.
Group BY: Filter results by criticality (least time remaining data centers/custom data centers listed first) or by the vCenter Server to which each data center belongs. This option appears if you select ALL DATACENTERS on the upper right.
Sort by:

Options (Options appear if you select ALL DATACENTERS on the upper right):

  • Alarm clock graphic - lists data centers/custom data centers by time remaining.
  • Dollar sign - lists data centers/custom data centers by potential cost savings.
  • Scales graphic - lists data centers/custom data centers by level of optimization.
Select datacenter or ADD NEW CUSTOM DATACENTER

Options (options appear if you select ALL DATACENTERS on the upper right):

  • Select a data center from the carousel across the top of the page. All data following refreshes with information for the selected object.
  • Select ADD NEW CUSTOM DATACENTER to display a dialog box that allows you to define a custom data center.
Time Remaining

Appears when you select a data center or custom data center from the top of the screen.

Gives overview of cluster status, including how many are at:
  • Critical
  • Medium
  • Normal
  • Unknown

"Critical" can indicate a resource contention, imbalance, or other stress condition. Thresholds you set in the policies define what is critical.

Optimization Recommendations

Lists potential cost savings by reclaiming unused resources.

Indicates if workloads can be optimized across clusters.

VIEW RECLAIMABLE VMS - displays the Reclaim screen, where you can research and run potential VM reclamation actions.

VIEW OPTIMIZATION - displays the Workload Optimization screen, where you can optimize workloads based on your policy settings.

Cluster Utilization and Time Remaining

Overall view of cluster health in the selected data center. You can select a cluster from the list to display information about that cluster, or use the options to sort and filter results. The options you select dictate the data displayed in the graph.

Sort by:

  • Most Constrained: most constrained element
  • CPU (allocation or demand)
  • Memory (allocation or demand)
  • Disk Space (allocation or demand)
Note: Demand model is always on and is the default.

Filter: search field.

Show History for: The period before forecasting begins (does not impact the forecast calculation).

Show Forecast For: The forecast period.

How is the criticality determined? Displays the criticality threshold you set for this type of object in the Policies Library.

Cluster Time Remaining Settings: Click the Edit icon to edit the default policy for the selected cluster. Change the criticality threshold, risk level, allocation model and capacity buffer. Applying these changes affects all objects in the policy. For more information, see Configuring Policies in the VMware vRealize Operations Configuration Guide

Time Remaining graph

Data shows current and trending resource usage and pinpoints when a given cluster is projected to run out of CPU, memory, or disk space based on the allocation or demand model (default). vRealize Operations takes into account the business hours which you set in the current policy.

Legend Displays a legend of the colours in the time remaining graph. Click on any one of the legends to toggle its display in the chart.
Recommendations Option 1: Reclaim Resources.

Shows resources that can be reclaimed to increase time remaining for the selected cluster.

RECLAIM RESOURCES - displays the Reclaim screen, where you can research and run potential VM reclamation actions.

Option 2: Add Capacity.

Shows resources that can be added to increase time remaining.

Note: You might see that a data center or cluster is labeled optimized when it has few or no days remaining before CPU, memory, or disk space is predicted to run out. The seemingly odd assessment is due to optimization and time remaining being two different measures of data center and cluster health. A data center can be running at optimum based on policy settings for balance and consolidation, yet be almost out of resources. It is important to consider both measures when managing your environment.