Last week VMware released version 5.7 of the vCenter Operations Manager (vCops), part of the vCenter Operations Management Suite. This new version vCops contains some compelling improvements which are all mentioned in the release notes. In this article I will discuss the improvements on the Capacity Remaining badge.
Upgrade to vCops 5.7
Before you can leverage all the improvements, you have to upgrade vCops to version 5.7 which is a pretty straight forward process:
- Save the .pak file for the vCenter Operations Manager 5.7 build to your local storage.
- Log in to the Administration portal using the admin user credentials.
- On the Update tab, click Browse and select the .pak file that you saved to local disk.
- Click Update.
- Accept the EULA by selecting the check box and click OK.
- Accept and confirm the update.
- The update process will start, and the Administration portal page might refresh after some time and ask you to login again.
- Log in to the Administration portal, wait for the update to complete, and verify that the Upgrade tab indicates a successful upgrade.
Note: You might need to refresh the browser to view the updated product version and build number.
- Log in to the vCenter Server that manages the license for vCenter Operations Manager, and apply the new vCenter Operations Manager 5.7 license key.
Note: You must re-apply the new license key in the vCenter Server after the upgrade.
Before upgrading it is always a good idea to first backup the vCops appliance. One option is to use VMware snapshots for this; in this case I would advise to shutdown vCenter Operations Manager to ensure you have a consistent snapshot. After taking the snapshot, start the vCops vApp and resume the upgrade.
The improved capacity remaining badge
The capacity remaining badge gives you an indication of the remaining available virtual machines as a percentage of the total virtual machine capacity. The number of remaining virtual machines depends on the remaining resources of the selected object: a datacenter, cluster or ESXi server. In this article I’m focussing on cluster resource usage.
When the minor badge Capacity Remaining is clicked, you will be directed to the Planning –> Virtual Machine Capacity view. This view shows the current VM capacity of the selected object, the number of deployed VMs and the remaining VM capacity. The capacity is shown for i.a. CPU, Memory, Disk Space & IOs and network. The most limiting resource determines the remaining capacity for the object, which will be shown in the dashboard.
In the example below you will see memory is the most limiting resource, which is displayed in bold (click to enlarge):
In this case the cluster capacity is 107 VMs, we have 21 VMs powered on, which leaves us with a remaining capacity of about 87 virtual machines (forget the rounding here).
My personal experience is that vCops is sometimes a bit too positive about the remaining capacity. This has to do with the fact that vCops is not evaluating the allocated resources but the demanded resources by default (this is the only option in vCops pre-5.7). Demanded resources are made up of both Effective Demand and Reserved Capacity to determine the remaining capacity for the cluster.
Change the policy configuration
vCops 5.7 lets you choose which measure is used for the remaining capacity:
- By demand (what we used to do) – Demand looks at what resources are requested to determine what is remaining. This method is more aggressive and efficient.
- By allocation (new option in 5.7) – Allocation looks at resources that are allocated as used and unallocated resources as capacity remaining. This method is more conservative.
Select the option “Change Default Policy Configuration” and select option 3a “Capacity and time remaining”:
This will give you the option to enable Demand and/or Allocation for remaining resource calculation. You can enable it per resource type.
Enabling both “by Demand” and “by Allocation” will result in the use of the most constraining measure. In the example above I’ve selected by Allocation (Containers): This means that vCops will use the configured/used allocation for container objects like world, datacenter, cluster or host for the calculation. Selecting this configuration displays the following results:
As you can see total capacity has decreased tremendously to 36 virtual machines and remaining capacity is now about 15 virtual machines. In this example the memory allocation is used, which is about 57%.
Calculating remaining recourses by allocation is definitely a more safe option, but also a lot more conservative. Depending on the capacity calculation requirements you’ve set for your clusters, you now have the option to configure vCops 5.7 accordingly with this new option.