HP 742n - Pavilion - 512 MB RAM Installation and operating instructions

application note
application noteapplication note
application note
pc resource
pc resourcepc resource
pc resource
monitoring and
monitoring and monitoring and
monitoring and
performance
performance performance
performance
advisor
advisoradvisor
advisor
www.hp.com/go/desktops

1
Overview
HP Toptools is a modular web-based device
management tool that provides dynamic
information about HP hardware on your LAN.
Among the many features of HP Toptools
(version .5 onward) is the resource monitoring
facility, accessible using the Toptools
Management Console. This feature permits you
to monitor the hardware resources of remote
PCs on the network.
Why Monitor Resources?
Resource monitoring can help you to:
•Detect and prevent system resource
shortages and bottlenecks
•Identify potential areas for reallocating
resources
•Determine system upgrade needs and plan for
them
•Determine how to maintain or improve
system performance
•Troubleshoot problems on remote PCs.
How Resource Monitoring Works
For each hardware resource on a remote PC,
you can:
•Enable or disable monitoring of the individual
resource
•Enable or disable event alerts.
Using the Toptools Device Manager, you can
perform these actions on a single remote PC or
on a selection of remote PCs. Also, when you
install the Toptools for Desktops DMI Agent, you
can select the option to enable the resource
monitoring functions.
Note that the Resource Monitoring module is
always installed automatically together with the
Toptools for Desktops Agent, but is not
necessarily enabled: When installing the DMI
agent, you must select the option to enable
Resource Monitoring; when installing the WMI
Agent, Resource Monitoring is enabled by
default without providing a user-option.
Once you have chosen the resources, Toptools
monitors these resources continuously. If a
resource exceeds a threshold, an alert is
generated on the Toptools Management
Console.
You can also view the monitoring of a remote
PC’s resources in real time from the Toptools
Management Console.
Architecture
To set up network PC resource monitoring from
Toptools, you require:
•Either Toptools Device Manager or the
Toptools Value Pack installed on the Device
Manager PC
•Supported HP PCs on your network must
each have a copy of Toptools Agent installed
and Resource Monitoring enabled
•All managed PCs other than supported HP
models must each have installed a
Performance Management Agent.
Managed PCs
HP Toptools Device Manager
or
HP Toptools Value Pack
Toptools Agent
installed
on supported HP PCs
Performance
Management Agent
installed
on other PCs
Device Manager

2
To enable management of non-HP devices on
your LAN using Toptools Device Manager, the
non-HP PCs must run one or more of the
management protocols SNMP, DMI 2.x, WMI or
HTTP. The corresponding Agents must be
obtained from the respective manufacturer.
For more effective management of non-HP
devices, the HP Toptools Value Pack adds
features for all devices including a System
Performance Advisor, and SNMP trap reception
for non-HP systems with loaded MIBs.
For further details, plus version information and
availability of these software elements, see the
‘More Information’ section at the end of this
Application Note.
Event Alerts
Event alerts are generated when certain pre-
defined thresholds are exceeded, eliminating
the need for constant visual monitoring.
Two levels of alert can be generated:
•Warning – this means that the situation has
not yet reached critical level but the first
threshold has been exceeded, so a warning is
issued.
•Critical – this is the most serious alarm and it
indicates that the level has become critical.
You can change these thresholds directly on the
resource monitoring screen in the Toptools Device
Manager.
Sampling Intervals
When monitoring resources, Toptools takes
measurements in two different ways. The first, the
simpler of the two, is a single measurement at an
instant in time. For example, the amount of free
disk space at a particular time.
The second type of measurement is taken over a
fixed period of time, and reflects the average value
during that time. For example, CPU usage and
LAN bandwidth. This type of measurement is
useful when a resource, during normal use, is
subject to dramatic peaks and troughs over a short
period of time. In this case, an average value is a
better reflection of resource utilization, and it
avoids the false alerts that would be produced
when resource utilization suddenly peaks for a
short period.
To get the average value, Toptools takes regular
measurements in very short time intervals, for
example, every 5 to 30 seconds.
This shows the monitoring of hard
disk drives on a single remote
Vectra PC.
The PC has two logical hard disk
drives, C and D. or each, disk
usage (over a time period) and the
current amount of free disk space
is shown.
You can disable or enable both
monitoring and event alerts on
each disk drive.
To change threshold values, simply
drag the red or yellow arrow up or
down, as desired.
Resource Monitoring
Example 1

3
These sample values are used to provide an
average value over a much longer window of
time, for example, over a period of an hour. It is
this average value which is used to trigger alerts.
Which Resources can be Monitore ?
CPU Usage
The remote PC’s processor usage can be
monitored. The value shows the percentage of
time that the processor is busy (not idle). This
measurement is taken over a fixed period of
time, and reflects the processor’s average usage
during that time.
Time
Value
Sample Values
Average
Value
If the remote PC has two processors, each is
monitored separately.
This measurement can be useful for detecting
processor malfunctions, identifying processor
undercapacity, or processor overcapacity for
reallocating resources.
There are cases where the processor usage value
is artificially high. For example, some screen
savers use 100% of processor time when
running, even though the PC is not being used.
Another example is when the PC has been
started but the user has not yet logged on. In
some Windows operating systems, the login
window that is displayed uses 100% of processor
time until the user logs in.
Memory Usage
There are three useful measurements provided:
swap activity, swap file usage, and physical
memory.
Two of these measurements concern the
Windows swap file. The swap file is a large file
that the operating system uses to switch
memory chunks between physical memory and
the hard disk. In this way, virtual memory is
created, extending the PC’s available memory
significantly. The PC can run much larger
applications and process larger data chunks than
it could without a swap file.
This shows memory usage on a
single remote Vectra PC
It shows physical memory usage,
swap file usage, and swap
activity.
You can disable or enable
monitoring and event alerts for
these measures (except on
physical memory usage).
To change thresholds for the
swap activity, simply enter new
values in the space provided and
click Save Settings.
Resource Monitoring
Example 2

4
Swap Activity
This measurement shows the average number
of pages per second that are read from or
written to the Windows swap file. This
measurement is taken over a fixed period of
time, and reflects the average rate during that
time.
The swap activity measurement is particularly
useful. A high value means that the remote PC
does not have enough physical memory and it
compensates by continually swapping data
between system memory and the swap file. This
slows down the PC considerably.
You can monitor the swap activity to easily
identify remote PCs that require memory
upgrades.
How high can this value go before it indicates a
memory shortage? This depends on the PC. A
high speed system can support a higher swap
rate (without slowing down the PC) than a slow
PC. With experience, you will learn which
values can be regarded as too high for different
types of PCs.
Swap ile Usage
This shows the amount of swap file space
currently being used.
The total amount of space that the swap file can
potentially use is also shown. The total amount
of space available depends on the operating
system settings used, and, in some cases,
available disk space.
If there is not enough free space for the swap
file, memory allocation can fail, disrupting
applications and even causing their operations
to fail.
You can monitor this measurement to identify
remote PCs that require a different swap file
setting in the operating system, or more disk
space (by installing a second hard disk, for
example).
Physical Memory
This shows the amount of memory, in bytes,
currently used on the remote PC. The total
amount of memory is also shown.
This measurement is not used to detect
memory shortages, since Windows memory
management uses swap files. Therefore, if all
physical memory is being used, it does not
mean that there is a memory bottleneck.
Rather, it is provided for information only.
Consequently, no alerts are generated for this
item.
Performance
Advisor
Example1
This window shows
the current settings
for performance
monitoring. You can
select which remote
systems to monitor.
The information on
each remote system
includes System
Identity, Type,
Status (alerts), CPU
Use, Memory Use,
Disk Input/Output,
Storage (capacity)
and Condition
(performance data
collection).

5
Disk Usage
Two measurements are provided: disk activity
and disk space.
Disk Activity
This shows the percentage of time that the PC’s
hard disk drives are busy. A hard disk is busy if it
is performing an input or output operation. This
measurement is taken over a fixed period of
time, and reflects each disk drive’s average
activity during that time.
This measurement is only used in Toptools
Value Pack and is only available for PCs running
Windows NT or Windows 2000.
Disk overactivity could indicate, for example,
memory shortages through swap file activity
(described earlier) lack of free disk space, or an
inadequate mass storage system. For example,
this PC could benefit from replacing the IDE
system with a faster SCSI system.
Disk Space
The current free disk space and total disk space
is shown for each logical disk (that is, for each
hard disk partition on the PC).
This is useful for viewing the PC’s current hard
disk configuration and usage. For example, if
there is not enough free disk space on one of the
logical disks, the PC may benefit from a different
disk partition setup or even a disk drive upgrade.
Network Usage
Three measurements are provided: LAN I/O
packets, LAN I/O error rate, and LAN bandwidth
utilization. These measurements are provided for
each LAN card installed in the PC.
These measurements are all taken over a fixed
period of time, and reflect the average value
during that time.
LAN I/O Packets
This measurement shows the number of network
packets received and sent by the LAN card.
It can be useful for determining that the PC’s
network connection is working correctly, as well
as for showing network traffic patterns.
This measurement is only used in Toptools Value
Pack and is only available for PCs running
Windows NT or Windows 2000.
Performance Advisor
Example 2
This shows a remote PC’s
memory usage over the last
hour.
You can also view the memory
usage over the last day, the last
week, and the last month.
Similar trend analysis
information is available for CPU,
disk input/output, and disk
capacity.

6
LAN I/O Error Rate
This measurement shows the percentage of
network packets sent and received that
contained errors.
This is useful for spotting network errors and
for isolating problem areas. For example, a very
high error rate between a specific PC and a hub
could point to a faulty network cable between
these two devices.
This measurement is available for PCs running
either Windows NT or Windows 200 which have
the Microsoft SNMP Service installed. However,
in some cases, for Windows PCs without
Microsoft SNMP, the Toptools Device Manager
may be able to calculate an approximate value
using the LAN card’s DMI instrumentation.
LAN Bandwidth Utilization
This measurement shows the percentage of the
LAN card’s transfer capacity that is actually
being used.
This information can help to optimize a network
configuration, or can indicate network
undercapacity. If network bandwidth utilization
is nearing saturation, this is a useful warning
that network facilities must be upgraded or
reconfigured to provided more capacity in the
areas where it is needed.
The calculation of LAN bandwidth use is
available for PCs running either Windows NT or
Windows 2000 and with the Microsoft SNMP
Service installed. In some cases, for Windows
PCs without Microsoft SNMP, the Toptools
Device Manager may be able to calculate an
approximate value using the LAN card’s DMI
instrumentation.
Typical Uses of Resource Monitoring
•You can set alerts to permanently monitor
the hard disk drive space on all systems. In
this way, you can ensure that there is
sufficient disk space on all devices for future
growth. A critical alert on a single PC may
prompt you to tell the user that they must
remove all unneeded files from their system.
Alerts are also useful for receiving early
warning of undercapacity, enabling you to
plan for major disk upgrades.
•Before the installation of a new software
application, you can install it in several trial
PCs and monitor the PCs’ page swapping
rates to determine if more physical memory
is needed to successfully run that
application.
•If a PC user complains about a slow PC or
disruptive system errors, you can monitor
resources (such as memory and network
usage) on that remote PC to help track down
and resolve the problem.
Overhea Use by Resource
Monitoring
Since resource monitoring is done by the
remote PCs themselves, no network bandwidth
is used for this task. Furthermore, less than 1%
of the processor time is used on the remote
PCs.
Alert thresholds are also stored and managed
by each remote PC. If alerts are generated, a
small amount of network bandwidth is used to
send the alert to the Device Manager.
If you are using the Device Manager to display a
remote PC’s resource monitoring, requests are
sent between these two devices on the network
to keep the display up to date. However, even
this task requires very little network bandwidth.
There are two cases in which no monitoring is
done by the remote PCs, and therefore zero
system capacity is used. Firstly, if resource
monitoring is disabled, and secondly, if the
remote PC goes into standby or sleep mode.
Toptools allows the PC to go into standby or
sleep mode, rather than waking it up to
continue monitoring its resources. If the remote
PC goes into standby or sleep mode, it implies
that the PC is not currently being used and
there is no need to monitor resources.

7
System Performance A visor
The Toptools Value Pack includes a system
performance module. This module is a powerful
tool for centrally monitoring and managing
system performance of groups of PCs, servers,
and the network.
It enables system administrators, help desk
personnel, and VAR service providers to
proactively identify, diagnose, and alleviate
system performance bottlenecks on PC and PC
server systems. The System Performance
Advisor continuously monitors key performance
parameters against configurable thresholds,
providing configuration advice and upgrade
recommendations based on trend analysis in
the sampled data.
The System Performance Adviser contributes
to:
•Increased network and system uptime
Increased network and system uptimeIncreased network and system uptime
Increased network and system uptime, by
providing proactive bottleneck detection
with continuous, at-a-glance, performance
status of all monitored systems from
anywhere on the network.
•Optimized configuration, upgrade, and
Optimized configuration, upgrade, and Optimized configuration, upgrade, and
Optimized configuration, upgrade, and
replacement decisions
replacement decisionsreplacement decisions
replacement decisions.
The System Performance Advisor includes
one-click access to system performance
histories, facilitating trend analysis and
highlighting areas of sustained performance
degradation that require action.
•Improved system capacity planning
Improved system capacity planningImproved system capacity planning
Improved system capacity planning, with
recording of critical performance statistics
for all monitored PC and server resources,
allowing a thorough understanding of actual
system performance baselines.
There is no need to manually deploy
performance agents on every monitored
system. Toptools System Performance Advisor
lets you select the systems to monitor and then
automatically deploys the necessary client
code. (Remote systems running the Toptools
Agent already have the necessary code.) You
can even configure which subsystems (CPU,
Memory, Disk I/O) to monitor on each system.
Toptools System Performance Advisor collects
performance data and builds a centralized
database for trend analysis. From the main
screen, the status of all monitored systems can
be viewed as red, yellow, or green based on
configurable thresholds. An at-a-glance
summary is displayed with dynamically updated
pie charts organized by user-configurable
custom groups of devices. Access to
performance trend data on any of the
monitored systems is a simple mouse click
away.
Information on the Toptools Value Pack can be
viewed when you install the standard Toptools
Device Manager.
For More Information
For product information, ordering details, and
FAQs on Toptools Value Pack and other
Toptools software, connect to:
www.hp.com/toptools