Capacity Advisor is a utility with which the user monitors and evaluates
system and workload utilization to help make the most of the available systems.
The systems under consideration can be a single system or multiple systems
that are connected in a cluster configuration
or to a network. A single system can include multicore or hyperthreaded processors.
Capacity Advisor helps the user evaluate the effect of varying workloads
and decide how to move workloads
to improve utilization. The quantitative results from Capacity Advisor can
aid the planner in estimating future system workloads and in planning for
changes to system configurations.
With HP Integrity Essentials Capacity Advisor, the user can perform the following
tasks:
Collect utilization data on CPUs and memory
View historical
utilization for customized workloads
View historical workload utilization and aggregate utilization across the
partitioning continuum (nPars, HP-UX vPars, HP-UX Virtual Machines)
Generate utilization reports
Plan workload or system changes and assess impact on utilization
Assess utilization impact for proposed changes in workload
location or size
Capacity Advisor can be used to pose questions
about changes in system configuration. Changes, such as the following, can
be tested:
Consolidating several systems into one system
Resizing a system for an upgrade
Resizing the demands on a system to reflect a forecast
Replacing older, small to mid-sized systems with virtual machines
Capacity Advisor can use data collected over time to show the
results of these configuration changes in many ways. Graphical views are
available so you can see what the effects of the changes are over time.
Tables are available that give the percentage of time and the degree to which
the system is busy; this information is valuable in comparing utilization
and quality of service before and after a change. Other tables show how many
minutes per month the system is unacceptably busy–a measure valuable
for both quality of service and for estimating TiCAP bills. Because Capacity
Advisor works from data traces collected over time, it is much more accurate
than using only peak data or average data
in understanding your systems and the workloads they support.
The significant advantage in using Capacity Advisor, rather than ad
hoc processes or guesswork, is that it provides a quantitative basis for examining
the activity of current resources. Additionally, it provides the capability
to try out what-if scenarios for moving workloads or other resources before
implementing a move.
Capacity Advisor incorporates numerical values of several components
in its analysis and modeling, including: