Compaq 179740-001 - ProLiant - 3000 User manual


September 1998
ECG096/0998
Prepared by Internet Solutions
Business Unit
Compaq Computer Corporation
Contents
The Compaq ProLiant 3000........ 3
Key Features............................. 3
Performance Results................. 3
What the Benchmarks Do
Not Reveal................................. 5
Load Simulator .......................... 5
Balancing Scalability and
Availability.................................. 6
APPENDIX A: TEST
DISCLOSURE........................... 7
APPENDIX B: RELATED
DOCUMENTS ......................... 10


Abstract: The Compaq ProLiant 3000 offers the best performance
and leading price/performance of any Intel-based or RISC-based
dual processor systems. To demonstrate Exchange Server 5.5
scalability on the Compaq ProLiant 3000 server, tests of user loads
from 1000 to 6500 simultaneous MAPI e-mail users were conducted
at Compaq’s Microsoft Competency Center located in Redmond,
Washington.
Microsoft Exchange Server has been the focal point for extensive
development and testing by both Microsoft and Compaq. Throughout
this activity, Compaq and Microsoft have worked to optimize
Microsoft Exchange Server performance on Compaq Server
Products to provide an optimal balance between performance,
availability, manageability, and cost. This performance brief
provides a basis for the number of users that can be supported when
deploying Microsoft Exchange Server on the ProLiant 3000 server.

 
ECG096/0998
Notice
The information in this publication is subject to change without notice and is provided “AS IS” WITHOUT
WARRANTY OF ANY KIND. THE ENTIRE RISK ARISING OUT OF THE USE OF THIS
INFORMATION REMAINS WITH RECIPIENT. IN NO EVENT SHALL COMPAQ BE LIABLE FOR
ANY DIRECT, CONSEQUENTIAL, INCIDENTAL, SPECIAL, PUNITIVE OR OTHER DAMAGES
WHATSOEVER (INCLUDING WITHOUT LIMITATION, DAMAGES FOR LOSS OF BUSINESS
PROFITS, BUSINESS INTERRUPTION OR LOSS OF BUSINESS INFORMATION), EVEN IF
COMPAQ HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH DAMAGES.
The limited warranties for Compaq products are exclusively set forth in the documentation accompanying
such products. Nothing herein should be construed as constituting a further or additional warranty.
This publication does not constitute an endorsement of the product or products that were tested. The
configuration or configurations tested or described may or may not be the only available solution. This test
is not a determination or product quality or correctness, nor does it ensure compliance with any federal
state or local requirements.
Product names mentioned herein may be trademarks and/or registered trademarks of their respective
companies.
Compaq, Contura, Deskpro, Fastart, Compaq Insight Manager, LTE, PageMarq, Systempro, Systempro/LT,
ProLiant, TwinTray, ROMPaq, LicensePaq, QVision, SLT, ProLinea, SmartStart, NetFlex, DirectPlus,
QuickFind, RemotePaq, BackPaq, TechPaq, SpeedPaq, QuickBack, PaqFax, Presario, SilentCool,
CompaqCare (design), Aero, SmartStation, MiniStation, and PaqRap, registered United States Patent and
Trademark Office.
Netelligent, Armada, Cruiser, Concerto, QuickChoice, ProLiant, ProSignia, Systempro/XL, Net1, LTE
Elite, Vocalyst, PageMate, SoftPaq, FirstPaq, SolutionPaq, EasyPoint, EZ Help, MaxLight, MultiLock,
QuickBlank, QuickLock, UltraView, Innovate logo, Wonder Tools logo in black/white and color, and
Compaq PC Card Solution logo are trademarks and/or service marks of Compaq Computer Corporation.
Microsoft, Windows, Windows NT, Windows NT Server and Workstation, Microsoft SQL Server for
Windows NT are trademarks and/or registered trademarks of Microsoft Corporation.
NetWare and Novell are registered trademarks and intraNetWare, NDS, and Novell Directory Services are
trademarks of Novell, Inc.
Pentium is a registered trademark of Intel Corporation.
Copyright ©1998 Compaq Computer Corporation. All rights reserved. Printed in the U.S.A.
Microsoft Exchange Server 5.5 on the Compaq ProLiant 3000
Performance Brief prepared by Internet Solutions Business Unit
First Edition (September 1998)
Document Number ECG096/0998

 
ECG096/0998
The Compaq ProLiant 3000
The Compaq ProLiant 3000 delivers performance and expandability levels that allows customers
expansive growth potential. This high performance server uses the latest in processor and system
architecture technology to deliver best-in-class performance while providing increased expansion
capabilities to meet the ever-increasing requirements of high-volume file services or entry-level
applications. Additionally, the Compaq ProLiant 3000 includes advanced fault-tolerant
capabilities and rapid recovery features providing maximum uptime and reliable server operation
while lowering total cost of ownership.
Key Features
•300/333-MHz Pentium II Processor (dual processor capability) and an upgrade path to next
generation Intel Pentium II processor architectures that support SDRAM and 100 MHz GTL
bus technology next generation Intel Pentium II processor architectures that support SDRAM
and 100 MHz GTL bus technology.
•512-KB level 2 writeback cache per processor
•I2O Connector
•Highly Parallel System Architecture for improved system bandwidth
•ECC-protected memory bus and cache
•64 MB of ECC, EDO memory expandable to 3 GB
•5 PCI and 3 shared PCI/EISA slots (8 total slots)
•Internal mass storage capacity up to 109.2 GB
•External mass storage capacity up to 1.5 TB (12 ProLiant Storage Systems Model F1)
•Netelligent 10/100 TX PCI UTP Controller
•Dual Channel Integrated Wide-Ultra SCSI-3 Controller
•Integrated Remote Console, and Automatic Server Recovery-2 (ASR-2)
•750 Watt hot pluggable power supply with optional hot-pluggable redundant power supply
(RPS)
•Compaq Insight Manager and SmartStart
Protected by a three-year on-site limited warranty and extended Pre-Failure Warranty which
covers Pentium II processors, memory and disk drives, as well as a wide range of Compaq
Services.
Performance Results
The tests were conducted using Microsoft’s Load Simulator tool. For each test, the Load
Simulator Medium MAPI canonical profile was chosen. The Medium MAPI profile reflects the
task workload of a typical corporate e-mail user - including daily common mail tasks such as
send, browse, read, forward as well as adding calendar entries for tasks and distribution list usage.
The Response Time score represents a 95th percentile score of the measured test run. The score is

 
ECG096/0998
reflected in milliseconds (ms). A Response Time score of 1000ms or less is considered to be an
acceptable response time for e-mail users utilizing Exchange Server’s MAPI protocol.
Performance Highlights (ProLiant 3000, 2 CPU, 512MB RAM)
User Load 1000 2000 3000 4,000 5000 6500
Response Time
(milliseconds) 92ms 102ms 130ms 181ms 241ms 312ms
Average CPU Utilization 6.40% 14.53 23.55% 35.86% 54.89% 77.54%
Messages Delivered (8
hour) 19119 38853 57089 77572 94552 114558
Note: This document uses accepted industry-standard benchmarks to help illustrate the performance
capabilities of the Compaq ProLiant servers. This document provides test results on server products as well
as an overview of the test, results summary, and server configurations used to generate performance results.
Similar results can be expected using a Compaq ProLiant 3000R. Tested configurations may support higher
user loads. This summary indicates performance results for tested configurations only. Test result
disclosure can be found in Appendix A.
Compaq ProLiant 3000 (Configuration a)
(2) Pentium II/ 300MHz – 512KB Level 2 writeback cache per processor
512MB RAM
(2) SMART-2DH Array Controller
OS/Pagefile/ Exchange DS/MTA Files: (2) 4.3-GB Drives – RAID1
Exchange Log Files: (2) 4.3-GB Drives – RAID1
Exchange Information Store Files: (10) 4.3-GB Drives – RAID5
Compaq Netelligent (100BaseTX) network interface card (NIC)
Windows NT Server v4.0 + Service Pack 3
Exchange Server version 5.5 – Standard Edition (Tuning: Perfwiz defaults)
Compaq ProLiant 3000 (Configuration B)
(2) Pentium II/ 333MHz – 512KB Level 2 writeback cache per processor
512 MB RAM
(2) SMART-2DH Array Controller
OS/Pagefile/ Exchange DS/MTA Files: (2) 4.3-GB Drives – RAID1
Exchange Log Files: (2) 4.3-GB Drives – RAID1
Exchange Information Store Files: (8) 9.1-GB Drives – RAID5
Compaq Netelligent (100BaseTX) network interface card (NIC)
Windows NT Server v4.0 + Service Pack 3
Exchange Server version 5.5 – Standard Edition (Tuning: Perfwiz defaults)

 
ECG096/0998
Compaq ProLiant 3000 (Configuration C)
(2) Pentium II/ 333MHz – 512KB Level 2 writeback cache per processor
2147 MB RAM
(2) SMART-2/DH Array Controller
OS/Pagefile/ Exchange DS/MTA Files: (2) 4.3-GB Drives – RAID1
Exchange Log Files: (2) 4.3-GB Drives – RAID1
Exchange Information Store Files: (8) 9.1-GB Drives – RAID0
Compaq Netelligent (100BaseTX) network interface card (NIC)
Windows NT Server v4.0 + Service Pack 3
Exchange Server version 5.5 – Standard Edition (Tuning: Perfwiz defaults)
What the Benchmarks Do Not Reveal
It is important to understand that benchmarks such as these are designed to give Exchange Server
implementation planners baseline references for understanding the capabilities of hardware
platforms from a single vendor such as Compaq or other competing hardware vendors. When
interpreting these benchmarks, two things should be kept in mind. First, consider whether
benchmarks are performed on what can be referred to as customer deployable configurations. A
hardware vendor may publish a result that is based on a platform or configuration that one would
never use in a real world Exchange Server deployment. For example, many vendors have
published results using disk subsystems configured with RAID0 disk arrays. While RAID0 does
provide the highest levels of disk subsystem performance, it fails to provide any protection
against data loss. One must consider the implications of the same benchmark performed on a
system using RAID5. This would provide the necessary fault tolerance that typical deployments
require but has different performance implications. In addition, most vendors, including Compaq,
conduct benchmarks for Exchange Server that are single-server in nature. Benchmarks also do
not account for issues such as backup and disaster recovery as well as information store
maintenance sizing concerns. Whatever the issue, care must be taken when interpreting
benchmarks from any vendor to ensure that results represent useful information to your Exchange
Server deployment and are based on valid simulation methodologies
Load Simulator
The main tool used in generating the workload used in this scalability demonstration was the
Microsoft Exchange Server Load Simulation utility called Load Simulator. As its name implies,
Load Simulator is a tool for simulating a client user load on an Exchange Server. Its purpose is to
enable a single Windows NT machine—called a LoadSim client—to simulate multiple Microsoft
Exchange client users.
The operation of Load Simulator users is governed by a Load Simulator profile. This profile
controls factors such as how long a Load Simulator ‘day’ is, how many e-mail messages to send
in a day’s time, how many times to open and read existing e-mail, whether to use distribution lists
and public folders.
Load Simulator creates a highly accurate simulation of reality. It mimics the full Microsoft
Exchange Client in many respects. First, it uses .MSG files, the same format used by the
Exchange Client. This guarantees that messages generated by Load Simulator have the same
properties as those sent by real users of the Exchange Client. Second, Load Simulator uses the

 
ECG096/0998
same MAPI remote procedure call (RPC) semantics as those used by the Client. Third, Load
Simulator registers MAPI change notifications in the same manner as they are registered by the
Client. Finally, Load Simulator even emulates the Microsoft Exchange Client list-box cache,
which the Client uses for folder and message panes in the viewer when a user browses and selects
messages on the server.
Several steps are necessary to performing a successful simulation. The Load Simulator setup and
initialization process comes first. Load Simulator creates the test topology by first generating the
user directory entries. Next, the test store is initialized and populated with the test messages and
folder items. The tests are typically run for up to 8 hours depending upon the user load simulated
and amount of time required to reach a steady-state for measurement purposes. During a test run,
users logon to the Exchange Server and begin processing various messaging tasks. Task response
time data is logged to the LSPERF.LOG file and client messages and error logging is stored in the
LOADSIM.OUT file. To produce test score the LSLOG utility is used to parse out the
LSPERF.LOG file and calculate the response time score. By default, 95th and 50th percentile
response time scores are calculated. 95th percentile response time scores for the MAPI/RPC
protocol should be less than 1000 milliseconds (ms), according to Microsoft. Also, the Exchange
Server IS Send Queue and the MTA Work Queue (other message and delivery queues should also
be considered depending on the protocol) must consistently return to near-zero during the steady-
state period for which test measurements are taken. Queues that continue to grow and fail to
return to near zero indicate that the server is not sustaining the required workload. There should
be no errors logged by the LoadSim clients during the test. When these conditions are met, a
successful test run has been completed. For more information on LoadSim Medium canonical
profiles, please refer to the Load Simulator documentation at:
(http://www.microsoft.com/exchange/library/loadsim55x86.exe).
Balancing Scalability and Availability
While server performance and capacity are key criteria in selecting a messaging deployment
platform, one must also consider price/performance. Several competing hardware vendors offer
platforms capable of supporting heavy user loads. They also provide these systems at
significantly higher cost. Compaq delivers leading performance on industry-standard platforms
with the lowest total cost of ownership.
For many corporations, messaging and collaboration have quickly become mission critical.
Unscheduled downtime for any server can result in a significant loss of productivity. To limit
exposure to downtime, Exchange Server 5.5 and Compaq ProLiant Clusters provide high
availability through Microsoft Cluster Server (MSCS). When MSCS is deployed on Compaq
ProLiant Clusters, enterprise-messaging customers can achieve scalability without sacrificing the
reliability that is required in an enterprise environment.
Another critical concern is backup and disaster recovery. Compaq provides industry leading tape
array and library hardware solutions partnered with applications like Computer Associates’
Cheyenne ArcServe. These solutions will help meet the requirements of enterprise customers
deploying messaging and collaboration applications.


ECG096/0998
APPENDIX A: TEST DISCLOSURE
LoadSim Clients 
Network type (10Base T, Token Ring, etc.) 100 Base-TX
Number and type of clients (≤40) 2x5/133, 128MB RAM (≤250 users each) or
better (indicates minimum configuration)
Number and type of hubs/concentrators (full
duplex, switching, etc.) Compaq Netelligent 5708 Switch and Netelligent 2624 Hub
Number of clients/segment 20
Client CPU type and speed in percentages 2P/133 MHz Pentiums or better
Client network controller broken down by
percentages Compaq Netelligent 10/100
Client network software name and version
(drivers, protocols, redirector) Microsoft Windows NT Workstation 4.0 + SP3
TCP/IP
Size of any client network cache None
Network controller software Compaq Netelligent 10/100 driver
LoadSim version 5.5 (Build 2187)
Note: Response time measurements were taken from a LoadSim Control Client simulating 100 users
configured with 96MB RAM and a Pentium/166 CPU. The client is located on an isolated network segment
connected to a 100Mb/s switch.

 !
ECG096/0998
Performance Data Disclosure (Configuration A) (Measured during test run at steady state)
Indicator 1000
Users 2000 Users 3000 Users 4000 Users 5000 users
Response Time (ms) 92 102 130 181 249
IS Buffer Cache Hit % 99.53% 99.29% 99.19% 99.11% 98.72%
Disk Queue Length – IS
Volume 0.225 0.68 1.29 2.442 4.828
Disk Queue Length – Log
Volume 0.008 0.015 0.02 0.026 0.029
Average Read I/Os – IS 12.552 39.884 72.177 118.071 196.096
Average Write I/Os – IS 8.119 24.126 39.247 56.217 76.195
Average Write I/Os – Log 20.951 37.80 51.083 63.611 71.824
Average Pages/Sec 0.022 0.036 0.046 0.092 0.131
Average Available Bytes 4.85MB 4.94MB 4.89MB 4.88MB 4.86MB
IS Send Queue Average 0.01 0.172 0.293 0.596 2.273
IS Receive Queue Average 0.707 0.545 0.515 0.626 0.818
MTA Work Queue Average 0.051 0.111 0.232 0.303 0.556
Messages Open/sec 4.069 8.076 11.905 16.334 19.784
Average RPC Operations/sec 49.941 99.412 147.227 200.172 245.75
Messages Submitted (8-
hours) 14892 30284 44815 60718 74516
Calculated Messages/User 14.892 15.142 14.938 15.179 14.90
Average CPU Utilization 6.40% 14.53% 23.55% 35.86% 55.08%
Average Context
Switches/Sec 498 855 1210 1582 1954
Average CPU Queue Length 0.152 0.303 0.687 1.141 1.818
Working Set – STORE 420MB 467MB 464MB 463MB 462MB
Virtual Bytes – STORE 724MB 773MB 793MB 805MB 835MB
NOTE: Performance results measured using Microsoft NT Performance Monitor. Measurements
were obtained by measuring averages for the period of steady-state activity (i.e. after 5,000 users
were successfully logged on). Tests measure the messaging throughput of a single server, single
site topology. For deployment specific information contact a Microsoft or Compaq representative.
More information can be found at:
http://www.microsoft.com/exchange/support/deployment/planning/deploy.asp?A=5&B=1

 "
ECG096/0998
Performance Data Disclosure (Configuration B, C)(Measured during test run at steady state)
Indicator 5000 Users (Configuration B) 6500 Users (Configuration C)
Response Time (ms) 241 312
IS Buffer Cache Hit % 98.21% 98.91%
Disk Queue Length – IS
Volume 5.144 3.269
Disk Queue Length – Log
Volume 0.032 0.034
Average Read I/Os – IS 190.686 154.486
Average Write I/Os – IS 77.33 99.064
Average Write I/Os – Log 77.325 78.33
Average Pages/Sec 0.046 0
Average Available Bytes 4.85MB 1.1GB
IS Send Queue Average 1.889 3.222
IS Receive Queue Average 0.889 0.919
MTA Work Queue Average 0.566 0.636
Messages Open/sec 19.695 23.119
Average RPC Operations/sec 246.148 300.112
Messages Submitted (8-
hours) 74256.272 90866.928
Calculated Messages/User 14.851 13.979
Average CPU Utilization 54.89% 77.54%
Average Context
Switches/Sec 1952 1944
Average CPU Queue Length 1.939 4.101
Working Set – STORE 463MB 1.14GB
Virtual Bytes – STORE 851MB 1.52GB
Note: Configuration B, C (333 MHz CPU) Performance results measured using Microsoft NT
Performance Monitor. Measurements were obtained by measuring averages for the period of steady-state
activity (i.e. after 6,500 users were successfully logged on). Tests measure the messaging throughput of a
single server, single site topology. For deployment specific information, contact a Microsoft or Compaq
representative. More information can be found at:
http://www.microsoft.com/exchange/support/deployment/planning/deploy.asp?A=5&B=1

 #
ECG096/0998
APPENDIX B: RELATED DOCUMENTS
These documents are available on the Compaq web site.
Compaq and Microsoft Demonstrate Enterprise Scalability with Exchange Server 5.5,
http://www.compaq.com/support/techpubs/whitepapers/ECG00961197.html
Performance of Exchange Server 5.0 on Compaq ProLiant 6000-class Servers,
http://www.compaq.com/support/techpubs/whitepapers/ECG0520897.html
Performance of Exchange Server 4.0 on Compaq ProLiant Servers,
http://www.compaq.com/support/techpubs/whitepapers/444A0696.html
Disk Subsystem Performance and Scalability,
http://www.compaq.com/support/techpubs/whitepapers/ECG0250997.html
Configuring Compaq RAID Technology for Database Servers,
http://www.compaq.com/support/techpubs/technotes/184206-1html
Compaq SMART-2 Array Controller Technology,
http://www.compaq.com/support/techpubs/whitepapers/667A0697.html
Hardware vs. Software Fault Tolerance,
http://www.compaq.com/support/techpubs/whitepapers/ECG066/0298.html
Pentium II Processor Technology,
http://www.compaq.com/support/techpubs/whitepapers/046_0897.html
Configuring the Compaq ProLiant 5000 Server for Peak Performance,
http://www.compaq.com/support/techpubs/whitepapers/679A0697.html
Compaq White Paper Index,
http://www.compaq.com/support/techpubs/whitepapers
Compaq ProLiant 3000,
http://www.compaq.com/products/servers/proliant3000/index.html
Compaq ProLiant 3000R,
http://www.compaq.com/products/servers/proliant3000r/index.html