Examine the status of your objects in the views and heat maps so that you can identify the trends and spikes that are occurring with the resources on your cluster and objects. To determine whether any deviations have occurred, you can display overall summaries for an object, such as for the cluster disk space usage breakdown.

To examine the problems with your USA-Cluster further, use the Details views to display the metrics and collected capacity data for your cluster. Each view includes specific metrics data collected from your objects. For example, trend views use data collected from objects over time to generate trends and forecasts for resources such as memory, CPU, disk space.

Use the heat maps to examine the capacity levels on the cluster, host systems, and virtual machines. The block sizes and colors are based on the metrics selected in the heat map configuration.

Prerequisites

Use the Troubleshooting tabs to look for root causes. See Troubleshoot Problems with a Host System.

Procedure

  1. Click Environment > vSphere Hosts and Clusters > USA-Cluster.
  2. Examine the detailed information about the USA-Cluster in the views.
    1. Click the Details tab and click Views.

      The views provide multiple ways to look at different types of collected data by using trends, lists, distributions, and summaries.

    2. In the search text box, enter capacity.

      The list filters and displays the capacity views for clusters and other objects.

    3. Click the view named Cluster Capacity Overview, and examine the number of virtual machines listed for the USA-Cluster in the lower pane.

      Even though the USA-Cluster has two host systems and 30 virtual machines, no capacity exists.

  3. Examine the host systems in the cluster, and reclaim capacity from the descendant virtual machines.
    1. Click the Capacity tab.
    2. In the inventory tree, expand USA-Cluster, and click each of the host systems in turn.
    3. The host system w2-vcopsqe2-009 is in a critical state, with no capacity remaining.
    4. Click the Details tab, then click Views, and click Cluster Configuration View.
    5. To reclaim capacity from several virtual machines, select the cluster name
    6. Click the gear icon, and select Set CPU Count and Memory for VM.
    7. In the workspace that appears, click the Current CPU column title to sort the list according to the highest number of CPUs.

      Based on the actual use of the virtual machines listed, the New CPU column suggests fewer CPUs for each virtual machine.

    8. Click the check box next to each virtual machine that has a suggested lower CPU count, and click Begin Action. A confirmation message indicates that the action is underway and provides the task ID that you use to track the action in the Recent Tasks section under Administration. Click OK.

      By reducing the number of CPUs for each virtual machine, you free up capacity on your host system, and improve the USA-Cluster capacity and workload.

  4. Examine the heat maps for the host system and virtual machine objects in the USA-Cluster.
    1. In the inventory tree, click the USA-Cluster.
    2. Click Details, click Heatmaps, and click through the list of heat map views.
    3. Click Which VMs currently have the highest CPU demand and contention?

      The heat map displays blocks that represent the objects in the USA-Cluster. The block for a virtual machine appears in red, which indicates that it has a critical problem.

    4. Point to the red block and examine the details.

      The cluster, host system, and virtual machine names appear, with links to more information about the object.

    5. Click Show Sparkline to display the activity trend on the virtual machine.
    6. Click each of the Details links to display more information.

Results

To verify that freeing up memory on the virtual machines has improved the workload of the host system and the cluster, you can now examine the status of the host system and cluster.

You used views and heat maps to evaluate the status of your objects and identify trends and spikes, and free up capacity for your host system and the USA-Cluster. To further narrow in on problems, you can examine the other views and heat maps. You can also create your own views and heat maps.

What to do next

Examine the status for the objects in your environment hierarchy to determine which objects are in a critical state. Then examine the object relationships to determine whether a problem on one object is affecting one or more other objects. See Examine the Environment Relationships.