HP Capacity Advisor 7.2 User Guide

Power Tab
Table 19 Attributes available for use in analysis queries: Power Tab
DescriptionAttribute NameSection
The amount of power used by the systems. This
value is computed from the power calibration
and may not exactly match the exact power
consumption.
Peak Power (Watts)Power Utilization (Watts)
15–Min Power (Watts)
90%ile Power (Watts)
Average Power (Watts)
The Optimal Performance/Watt. When fully
utilized, this indicates the system's relative
performance per watt of energy used.
Optimal Perf/WattSystem Performance per Watt
The utilized performance of the system over
the average power consumed. The larger the
number, the more work is done per watt of
energy used.
Observed Performance/Watt
The system's estimated peak power
consumption when all resources are fully
consumed.
Max Power (W)Power Capacity
The system's estimated power consumption
while idle.
Min Power (W)
This value indicates whether the system's
minimum and maximum power consumption
values are known. The minimum and maximum
power consumption values have been tested
via Integrated Lights-Out (iLO) or manually
entered for the system.
See “Calibrating power on actual systems”
(page 85) for information on setting the power
calibration on systems where this value is
false.
Power Calibrated
NOTE: Peak, 15-minute sustained, 90th percentile, and average values are available as separate
attributes. See Table 13 (page 115).
Custom query examples
Example: Finding large underused VMs that could be replaced with medium or small
VMs
Background
In this example, the data center has three sizes of Hyper-V virtual machines — small, medium, and
large. A medium VM has two vCPUs and 4 GB of RAM. A large VM has four vCPUs and 8 GB
of RAM. Most applications in the data center start to have performance issues if the CPU utilization
is greater than 80%, or if the memory use is within 1 GB of the capacity.
As administrator, you want to locate large, underused VMs that could be replaced by smaller VMs.
Requirements
To find the oversized VMs, you need to develop a query to find systems that:
• Are Hyper-V VMs
• Have four cores
122 Using Capacity Analysis