ORACLE REVISION 1.

ORACLE’S PEOPLESOFT ENTERPRISE CAMPUS

SOLUTIONS 9 USING ORACLE 10g ON HEWLETT-
PACKARD INTEGRITY AND PROLIANT SERVER
BLADES
As a global leader in e-business applications, Oracle is committed to delivering high performance solutions that meet our customers’
expectations. Business software must deliver rich functionality with robust performance. This performance must be maintained at volumes
that are representative of customer environments.
Oracle benchmarks demonstrate our software’s performance characteristics for a range of processing volumes in a specific configuration.
Customers and prospects can use this information to determine the software, hardware, and network configurations necessary to support
their processing volumes.
The primary objective of our benchmarking effort is to provide as many data points as possible to support this important decision.

SUMMARY OF RESULTS
This benchmark measured the online performance for a large
PeopleSoft QuickTest Client Timers
database model.
Enterprise Campus
Solutions 9 Single 5,000 6,000 Testing was conducted in a controlled environment with no
Benchmark User Users Users other applications running. The tuning changes, if any, were
approved by Oracle Enterprise Development and will be
Average Logon 0.58 sec 0.58 sec 0.71 sec generally available in a future update. The goal of this
Avg. Load Student 0.61 sec 0.65 sec 0.91 sec
Benchmark was to obtain reference response times for
Center Campus Solutions Release 9 using Oracle 10g on HP
Integrity servers and Server Blades.
Avg. Page Load 0.69 sec 0.72 sec 0.83 sec
Avg. Page Save 1.12 sec 1.44 sec 1.63 sec

Batch Throughput PeopleSoft Enterprise Campus Solutions

9.0 using Oracle 10g on an HP Integrity
10,000 Invoices 24.83 26.37 and ProLiant Server Blades
min min
3.0
Invoices Printed per 24,164 22,753 Page Load
Average Response Time in

Hour 2.5 Save

10,000 Transcripts 17.07 19.83
min min 2.0
1.63
Seconds

Transcripts/Hour 35,149 30,257 1.44

1.5
1.12
1.0 0.83
0.69 0.72
BENCHMARK PROFILE 0.5
In May 2008, Oracle conducted a benchmark in Pleasanton,
CA to measure the comparative online and batch 0.0
performance of selected processes in Oracle’s PeopleSoft Single User 5,000 Users 6,000 Users
Enterprise Campus Solutions 9 with Oracle 10g™ 10.2.0.3.
A single 4-processor (8 cores total) Hewlett-Packard®
Figure 1: Oracle Enterprise Campus Solutions 9
Integrity™ rx6600 server was used for the database server,
Response Times
running Hewlett-Packard HP-UX 11iv3 (11.31). Two 2-way
(8 cores total) HP ProLiant® BL460c Server Blades were
The weighted averages in Figure 1 correspond to the
used as Application Servers running Red Hat® Linux 2.6.9-
transaction proportions described in Table 1.
42.Elsmp #1 SMP. A single 2-way (4 cores total) BL460c
was used as the Web Server. An HP Storage Works
EVA8100 disk array was used for storage.
 BUSINESS PROCESSES
ONLINE METHODOLOGY Oracle (PeopleSoft) defines a business transaction as a series
Hewlett-Packard LoadRunner® was used as the load driver, of HTML pages that guide a user through a business process,
simulating concurrent users. It submitted a business such as creating a new business case.
transaction at an average rate of one every 10 minutes for
each concurrent user to the application servers via the web The eight PeopleSoft Enterprise Campus Solutions 9.0 self-
servers. service processes tested in this benchmark are as follows:

Hewlett-Packard QuickTest® Professional was used to CAMPUS SOLUTIONS

automatically submit transactions and to record the
benchmark measurements on the client PC.
Measurements were recorded when the user load was
attained and the environment reached a steady state.
Accept/Decline Awards: - The student logs on and
navigates to the Student Center page. They click on the
“Accept/Decline Awards” link and accept, decline and
reduce separate awards. The student completes the
Accept/Decline Awards transaction and hits “Submit.”

Figure 2 shows a typical 4-tier benchmark configuration. Calculate Tuition: - The student logs on and navigates to
This benchmark was run using a physical 4–tier the Student Center page. They click on the “Finances
configuration; with the database server, the application Section” link and navigate to the ‘Account Inquiry’ option.
server and the web server all being hosted on separate boxes. After viewing their balance, sign out.
The Process Scheduler was also hosted on a separate server
Drop Class: - The student logs on and navigates to the
Load Drivers LoadRunner Student Center page. They click on the “Drop a Class” link
LoadRunner Controller
6,000 Users and enter the term and class id. The student completes the
extensive enrollment options and hits “Save.”

(Browser) Enrollment Backpack: - The student logs on and navigates

'Client' 1 × 2-way to the Student Center page. They click on the “My
QuickTest Web (4 cores)
44% Academic Requirements” link. The Academic Advisement
Server
Utilized engine runs to reassess degree progress data. The student
reviews their degree progress data then sign out.
2 × 2-ways
(8 cores) Add Class: - The student logs on and navigates to the
App.
Servers 65%
Utilized Student Center page. They click on the “Add a Class” link
and enter the term and class id for two separate classes. The
Process student completes the extensive enrollment options and hits
Scheduler
1 × 4-way “Save.”
(8 cores)
DB HCM 61%
Server
Utilized Load Student Center: - The student logs on and navigates
to the Student Center page. They view some information
then sign out.
Figure 2: 4-Tier Configuration
Make A Payment: - The student logs on and navigates to
Load times were measured from the time the user clicks a the Student Center page. They click on the “Finances
hyperlink or push button until the new HTML page has been Section” link and navigate to the ‘Make a Payment’ option.
rendered. Update times were measured from the time the Select ‘Pay by Credit Card’ and enter the credit card info
user clicks the <SAVE> button until the new HTML page and payment info. Hit “Save” and sign out.
has been rendered.
Schedule Planner Validation: - The student logs on and
Measurements were recorded on all of the servers when the navigates to the Student Center page. Navigate to the
user load was attained and the environment reached a steady “Schedule Planner” page, enter the term and course
state. The batch jobs were initiated about 25 minutes after numbers. Click to “Validate” the selected courses. Hit
LoadRunner started. QuickTest started about 35 minutes “Save” and sign out.
after LoadRunner.

COPYRIGHT© 2008 PeopleSoft, Inc. – Oracle, Inc. All rights reserved. Page - 2
 Process Single 5,000 6,000
User Users Users
Process % of Users Avg. Pacing
Add Class
Accept/Decline Awards 10% 10 min Logon 0.58 0.58 0.91
Calculate Tuition 10% 10 min Load Student Center 0.62 0.71 0.96
Drop Class 10% 10 min Load Shopping Cart 0.57 0.57 0.79

Enrollment Backpack 10% 10 min Load Preferences 01 0.61 0.57 0.62

Add Course 01 to Cart 0.56 0.56 0.62
Add Class 20% 10 min
Load Preferences 02 0.55 0.56 0.61
Load Student Center 20% 10 min
Add Course 02 to Cart 0.59 0.60 0.64
Make a Payment 10% 10 min
Load Related Class 01 0.56 0.61 0.56
Schedule Planner Validation 10% 10 min
Load Preferences 03 0.63 0.62 0.65
Total 100% Add Course 03 to Cart 0.60 0.60 0.80
Load Preferences 04 0.57 0.57 0.66
Table 1: Business Process Mix
Add Course 04 to Cart 0.57 0.59 0.68
Table 3 shows the proportions of the business processes used
in the measurements of this benchmark. The proportions are Load Confirmation Page 0.59 0.70 0.95
intended to simulate a typical user scenario. Page (Classes) Save 1.15 1.56 1.91
Load Student Center
ONLINE PROCESS RESULTS Logon 0.56 0.57 0.61
Table 2 shows the corresponding average QuickTest Load Student Center 0.59 0.61 0.82
response times, in seconds, for each business process along Make Payment
with the overall averages.
Logon 0.58 0.57 0.63
Load Student Center 0.60 0.63 0.92
Process Single 5,000 6,000
User Users Users Page (Pay Details) Load 0.58 0.60 0.64

Accept/Decline Awards Page (Amount) Load 0.59 0.59 0.58

Logon 0.59 0.60 0.67 Load Confirmation Page 0.56 0.56 0.55
Load Student Center 0.57 0.64 0.88 Page Save 1.14 1.41 1.50
Click Awards Link 0.58 0.57 0.59 Schedule Planner Valid
(Awards) Page Load 0.56 0.55 0.62 Logon 0.59 0.59 0.68
Select Award 01 0.56 0.57 0.57 Load Shopping Cart 0.56 0.57 0.68
Submit Selection 0.56 0.57 0.57 Select Term 0.09 0.09 0.12
Page (Selection) Save 1.04 1.18 1.33 Load Class Sections 92 0.57 0.59 0.57
Calculate Tuition Load Pref. Page 92 0.59 0.60 0.59
Logon 0.59 0.59 0.63 Add Course 92 to Cart 0.60 0.60 0.60
Load Student Center 0.66 0.68 1.06 Load Pref. Page 56 0.56 0.57 0.57
(Accounts) Page Load 2.08 2.24 2.77 Add Course 56 to Cart 0.59 0.60 0.59
Drop Class Load Pref. Page 12 0.67 0.58 0.58
Logon 0.58 0.59 0.82 Add Course 12 to Cart 0.57 0.57 0.62
Load Student Center 0.61 0.61 0.85 Validate Shopping Cart 1.06 1.20 1.30
Shopping Cart Page Ld. 0.58 0.64 0.91
Weighted Logon 0.58 0.58 0.71
Load Class Schedule 0.56 0.59 0.67
Weighted Load SC 0.61 0.65 0.91
(Confirm Drop) Page Ld. 0.60 0.63 0.60
Weighted Page Load 0.69 0.72 0.83
Page (Classes) Save 1.11 1.47 1.52
Weighted Page Save 1.12 1.44 1.63
Enrollment Backpack
Transactions / minute n/a 500 600
Logon 0.56 0.62 0.62
(Require.) Page Load 1.64 2.24 2.65 Table 2b: Business Process Response Times
The database and application servers were processing a total
Table 2a: Business Process Response Times of ~600 business processes per minute at the peak load of
6,000 concurrent users. The transaction rate is calculated by
dividing the number of users by the corresponding pacing.

COPYRIGHT© 2008 PeopleSoft, Inc. – Oracle, Inc. All rights reserved. Page - 3
 DATA COMPOSITION DESCRIPTION
BATCH PROCESS RESULTS Table 6 summarizes the structure of the target organization.
10,000 Invoices and 10,000 Transcripts were processed.
Data Model Large
Print Invoice Transcripts
Single Stream 4 Job Streams Person records 600,000
Enrolled students 100,000
5,000 Users 24.83 17.07
Employees 10,000
6,000 Users 26.37 19.83
Student Career records 500,000
Table 3: Batch Run Times in Minutes Courses in Course Catalog 5,000
Classes in Schedule of Classes 150,000
Student Enrollment records 10,000,000
SERVER PERFORMANCE
SF Accounts 2,000,000
PeopleSoft Enterprise Campus Solutions Item Types 2500
9.0 using Oracle 10g on an HP Integrity ITEM_SF lines 15,000,000
and ProLiant Server Blades FA Student Award Records 2,000,000
70
Table 6: Selected Data Composition
% Average CPU Utilization

60

50 Note that the Large Data Model represents five year’s worth
of history data.
40
DB Server
30 BENCHMARK ENVIRONMENT
A pp Servers
20 HARDWARE CONFIGURATION
Web Server
10 A Hewlett-Packard® Integrity™ rx6600 was used as the
P ro cess Sch. database server. It was equipped with the following:
0
5,000 Users 6,000 Users • 4 × 1.6 GHz Intel® Itanium™ 9050 Dual-Core
processors, each with 24 Megabytes of Level-3 write-
back cache per CPU
Figure 3: Average CPU Utilization • 32 Gigabytes of Memory (~30 GB used at peak load)
• 1 SAN-Connected HP StorageWorks EVA8100 disk
5,000 Users 6,000 Users array with 4 Emulex 2GB PCI HBA fibre-channel
connections
DB Server 49% 61%
52.91% 64.81%
• 2 × Hewlett-Packard 4GB Dual-Port Fibre Channel
App Servers
Host Bus Adapters
Web Server 35.81% 44.36%
18.71% 21.90%
• ~8 Terabytes of total Disk Space available (58 × 146
Process Sched.
GB + 4 × 72 GB internal disk drives), approximately
DB Server 29.53 GB 30.00 GB 512 GB of RAID 1 storage used for this benchmark.
App Servers 2.83 GB 3.60 GB The database size was ~70 GB.
Web Server 3.58 GB 3.77 GB
1.15 GB 1.48 GB
Process Scheduler:
Process Sched.
1 × HP ProLiant® BL460c server blade was used as the
Table 4: Average CPU and Memory Utilization process scheduler server. It was equipped with the following:
• 2 × 3.0 GHz Intel® Xeon™ 5160 Dual-Core processors,
each with64 Kilobytes of Level-1 Cache, 4 Megabytes
5,000 Users 6,000 Users
of Level 2 Cache and 1 Megabyte of Level-3 write-back
69 95 cache per CPU
DB Reads/Sec
DB Writes/Sec 171 215 • 8 Gigabytes of Memory
Table 5: Average I/O Activity • ~72 GB of storage with an integrated SmartArray 5i
Plus Controller

COPYRIGHT© 2008 PeopleSoft, Inc. – Oracle, Inc. All rights reserved. Page - 4
 Application Server(s):
2 × HP ProLiant® BL460c server blades were used as the
application servers. They were equipped with the following:
• 2 × 3.16 GHz Intel® Xeon™ 5460 Quad-Core
processors, each with 12 Megabytes of Level 2 write-
back cache per CPU
• 16 Gigabytes of Memory
• ~146 GB of storage with an integrated SmartArray 5i
Plus Controller
Web Server(s):
1 × HP ProLiant® BL460c server blade was used as the web
server. It was equipped with the following:
• 2 × 3.0 GHz Intel® Xeon™ 5160 Dual-Core processors,
each with64 Kilobytes of Level-1 Cache, 4 Megabytes
of Level 2 Cache and 1 Megabyte of Level-3 write-back
cache per CPU
• 8 Gigabytes of Memory
• ~72 GB of storage with an integrated SmartArray 5i
Plus Controller
SOFTWARE VERSIONS
Oracle’s PeopleSoft Enterprise Campus Solutions 9
Oracle’s PeopleSoft Enterprise (PeopleTools) 8.48.13
Oracle 10g 10.2.0.3
Hewlett-Packard HP-UX 11iv3 (11.31) (64 bit) (on the
database server)
Red Hat Linux 2.6.9-42Elsmp #1 SMP (32 bit)(on the
Application, Web and Process Scheduler servers)
Microsoft® Windows 2003 Server Enterprise Edition 5.2
Build 3790 (on the LoadRunner Drivers)
Microsoft® Windows® XP Professional Version 5.1 w/SP 2
(on the QTP client)
Microfocus COBOL Server Express 4.0 w/SP2
HP’s LoadRunner® 8.0 (Controller Build 1546)
HP’s QuickTest® Professional 9.1 Version 9.1.0.0 Build:
2565
BEA Tuxedo® 8.1 RP175 with Jolt 8.1
Microsoft Internet Explorer® 6.0 on the QTP client

QuickTest Client PC: BEA WebLogic Server™ 8.10 w/SP 6

1 × HP ProLiant® was used as the client
Java version "1.4.2_08"
• 1 × 2.66 GHz Intel® Pentium® IV Processor Java(TM) 2 Runtime Environment, Standard Edition
• 1 Gigabyte of Memory (build 1.4.2_08-b03)
Java HotSpot(TM) Client VM (build 1.4.2_08-b03, mixed
mode)
Load Simulation Driver(s):
1 × Hewlett-Packard® ProLiant®DL320 G2 was used as the
ICE Tracking:
load driver controller. It was equipped with the following:
1692373001
• 1 × 2.66 Gigahertz Intel® Xeon™ Processor,
with 2 Megabytes of Level-2 Cache
• 4 Gigabytes of Memory
2 × Hewlett-Packard® ProLiant®DL320 G2s were used as
the load drivers. They were equipped with the following:
• 1 × 2.66 Gigahertz Intel® Xeon™ Processors,
with 2 Megabytes of Level-2 Cache Oracle (PeopleSoft) Pleasanton

• 4 Gigabytes of Memory 4500 Oracle Lane

P. O. Box 8018
1 × Hewlett-Packard® ProLiant®DL380 G3 was used as the Pleasanton, California 94588-8618
load driver. It was equipped with the following:
Tel 925/694-3000
• 2 × 3.2 Gigahertz Intel® Xeon™ Dual-Core Processors, Fax 925/694-3100
each with 4 Megabytes of Level-2 Cache Email info@peoplesoft.com
• 4 Gigabytes of Memory WorldWideWebhttp://www.oracle.com

PeopleSoft, PeopleTools, PS/nVision, PeopleCode, PeopleBooks, PeopleTalk, and Vantive are registered trademarks, and Pure Internet Architecture, Intelligent Context
Manager, and The Real-Time Enterprise are trademarks of PeopleSoft, Inc. – Oracle, Inc. All other company and product names may be trademarks of their respective
owners. The information contained herein is subject to change without notice. Copyright © 2008 PeopleSoft, Inc. – Oracle, Inc. All rights reserved. C/N 0635-0608
©2008 Hewlett-Packard, Inc. All rights reserved. HP, Hewlett-Packard, the HP logo, Integrity and HP-UX are trademarks or registered trademarks of Hewlett-
Packard, Inc. in the United States and other countries.

COPYRIGHT© 2008 PeopleSoft, Inc. – Oracle, Inc. All rights reserved. Page - 5