5.6

n
Build a Super Metric Formula on page 49
A super metric formula can include one or more metric specifications, super metric functions, arithmetic
operators (such as the plus or minus sign), and constants. You can enter any number of constants as part
of the formula.
n
Add a Super Metric on page 50
You create a super metric when only a combination of metrics can let you know if your systems are
behaving normally.
n
Add a Super Metric Package on page 51
When you create a super metric package, you specify the metrics that it contains, configure threshold
characteristics for each super metric, and indicate which threshold violations should be considered KPIs.
If you include the same super metric in more than one package, you can set different characteristics for
it in each package.
n
Assigning Super Metric Packages on page 52
You can assign a super metric package to any resource, regardless of whether any metrics from the
resource are used in the super metric package. In most cases, you assign each super metric package to a
related resource.
n
Export a Super Metric on page 52
You can export a super metric from one vCenter Operations Manager instance and import it to another
vCenter Operations Manager instance.
n
Import a Super Metric on page 53
You can import a super metric that was exported from another instance of
vCenter Operations Manager.
Design a Super Metric
Because super metric formulas can be complex, design a super metric before you use the
vCenter Operations Manager user interface to create it. The key to creating a super metric that alerts you to
the right situations is knowing your own enterprise and your data.
Procedure
1 Determine the resources that are involved in the behavior to track.
When you define the metrics to use, you can select either specific resources or resource kinds. For example,
you can select the specific resources Database Server 2 and Database Server 4, or you can select the resource
kind Database Servers. When you select a resource kind, the super metric uses all of the resources of that
kind that are children of the resource to which you assign the super metric.
2 Determine the metrics to include in the super metric.
If you are tracking the transfer of packets along a network, the metrics are packets in and packets out
because you are interested in the ratio of those metrics. In another common use of super metrics, the metrics
might be the average CPU use or average memory use of the resource kind that you select
3 Decide how to combine or compare the metrics.
For example, to find the ratio of packets in to packets out, you must divide the two metrics. If you are
tracking CPU use for a resource kind, you might want to determine the average use, or you might want
to determine what the highest or lowest use is for any any resource of that kind. In more complex scenarios,
you might need a formula that uses constants or trigonometric functions.
VMware vCenter Operations Manager Administration Guide
46 VMware, Inc.