5.6

Alerts and Faults
vCenter Operations Manager generates alerts when events occur on the monitored objects, when data analysis
indicates deviations from normal metric values, or when a problem occurs with one of the
vCenter Operations Manager components.
Events that the vCenter Serverpublishes are the main source for faults. These events might originate in the
vCenter Server itself, or ESX servers might generate them and the vCenter Server publishes them externally.
Only a subset of vCenter events are considered as important for fault generation.
Metric Concepts for vCenter Operations Manager
Preparing to monitor your environment with vCenter Operations Manager involves some familiarity with
metrics that help you to identify a problem.
vCenter Server presents a use-based model of metrics. vCenter Operations Manager presents a demand-based
model of metrics. Some knowledge of the metrics that affect the data and graphs is useful to determine what
to do next in a workflow.
Table 1-1. Major Metric Concepts
Metric Description
Provisioned Amount of a resource that the user configures.
The provisioned metric might apply to the amount of
physical memory for a host or the number of vCPUs for a
virtual machine.
Usable Actual amount of a resource that the object can use.
The usable amount is less than or equal to the provisioned
amount. The difference between the provisioned amount
and usable amount stems from virtualization overhead. This
overhead might include the memory that an ESX host uses
to run the host, to support reservations for virtual machines,
and to add a buffer for high availability.
The usable metric does not apply to virtual machines.
Usage Amount of a resource that an object consumes.
The usage amount is less than or equal to the usable amount.
Demand Amount of a physical resource that the object might consume
without any existing constraints.
An object becomes constrained because of under-
provisioning or contention with other consumers of the
resource. A virtual machine might require 10GB of memory
but can only get 5GB because the virtual machine must share
resources with other virtual machines on the host.
When the demand amount is less than the usage amount, the
environment might have wasted resources. When the
demand amount is greater than the usage amount, the
environment might incur latency and exhibit decreased
performance.
Contention Difference between the amount of the resource that the object
requires and the amount of the resource that the object gets.
This metric measures the effect of conflict for a resource
between consumers. Contention measures latency or the
amount of time it takes to gain access to a resource. This
measurement accounts for dropped packets for networking.
Chapter 1 Introducing Key Features and Concepts
VMware, Inc. 9