O RACLE E NTERPRISE B ENCHMARK R EV. 1.1 PEOPLESOFT HRMS 9.1 SELF-SERVICE USING ORACLE FOR SOLARIS (UNICODE) ON AN ORACLE S SPARC T4-4 and T4-2 s 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. Benchmark (English) Référence d'exécution (Français) SUMMARY OF RESULTS PeopleSoft HRMS 9.1 Self-Service Standard Data Model Average Response Search 1.01, Save 0.63 Concurrent 15,000 PeopleSoft Ressources Humaines 9.1 Norme modèle de données temps de réponse Search 1,01, Save 0,63 Concourants 15.000 Utilisateurs BENCHMARK PROFILE In August 2011, Oracle (PeopleSoft) conducted a benchmark in Burlington, MA to measure the online performance of Oracle s PeopleSoft Enterprise Human Resources Management System (HRMS) 9.1. The database server used Oracle Database 11gR2 running on Oracle s SPARC T4-4 server (4 cpus, 32 cores, 256 vcpus) with Oracle Solaris 10 8/11. The application server also used a SPARC T4-4 server and ran Oracle Solaris 10 8/11. The web server was run on Oracle s SPARC T4-2 server. Approximately 524 GB of storage from one 1,920 GB Oracle s Sun Storage F5100 Flash Array storage system was allocated to the database instance. Benchmark-Test (Deutsch) Patrón de rendimiento (Español) Benchmark (Português) PeopleSoft Personalmanagementsystem 9.1 Datenbankmodell Standard Antwortzeit Search 1,01, Save 0,63 Gleichzeitige 15.000 Benutzer PeopleSoft Recursos Humanos 9.1 Volumen Estándar de datos tiempo de reacción Search 1,01, Save 0,63 Simultáneos 15.000 Utilizadores Gerenciamento de Recursos Humanos, PeopleSoft 9.1 Volume Padrão dos dados tempo de resposta Search 1,01, Save 0,63 Simultâneos 15.000 Usuários The benchmark measured client response times for 5,000, 10,000 and 15,000 concurrent users. The standard database composition model represents a large-sized company profile. The testing was conducted in a controlled environment with no other applications running. The goal of this Benchmark was to obtain baseline results for PeopleSoft HRMS 9.1 self-service transactions with Oracle Database for Solaris on SPARC T4 s. Average Response Time in Seconds Oracle Enterprise HR Self-Service 9.1 using Oracle11g R2 for Solaris on SPARC T4-4 s 3.0 2.5 2.0 1.5 1.0 0.5 0.0 Retrieve Update 5,000 10,000 15,000 Figure 1: Average Response Times * This average is weighted based on the business mix as reflected in Table 1: Business Process Mix.
METHODOLOGY Oracle ATS was used as the load driver, simulating concurrent users. It submitted a business process at an average rate of one every five minutes for each concurrent user. Measurements were recorded when the user load was attained and the environment reached a steady state. Figure 2 shows a typical 4-tier benchmark configuration. 'Controller' Oracle ATS 'Agents' Oracle ATS 15,000 BUSINESS PROCESSES Oracle (PeopleSoft) defines a business transaction as a series of HTML pages that guide a user through a particular scenario, such as promoting an employee. The fourteen PeopleSoft Enterprise 9.1 HRMS business processes tested in this benchmark are as follows: EMPLOYEE SELF-SERVICE eprofile Update Home Address: Update address in Personal Data tion. Update Home Phone: Update phone number in Personal Data tion. ebenefits Web 13% 16-way View Benefits data. Summary: View overall benefits enrollment Benefits Change Life: View benefits and alter the beneficiaries allocations in the Basic Life Plan. App 40% 32-way epay View Paycheck: View current paycheck information. DB HCM 8% 32-way Update Direct Deposit Info: Add a direct deposit directive. Employee Adds Profile Items: Add competencies to personnel profile. Figure 2: 4-Tier Configuration MANAGER SELF-SERVICE edevelopment Load (search/retrieval) times were measured from the time the user clicks the <OK> button until all the data for the entire business transaction has been retrieved. View Employee Info: eprofile View job and personal information. Update (save) times were measured from the time the user clicks the <SAVE> button until the system has released the page. Initiate Termination: Initiate a termination by recording an effective date and reason for termination. Initiate Promotion: Initiate a promotion by entering a new job title and salary. ecompensation Initiate Employee Salary Change: Process a salary change for a single employee. HR ADMINISTRATION Add a Person: Add a person and their biographical details. Hire a Person: Enter the specified job data and work location, followed by the payroll and compensation details. Add a Job: Add job details to an existing employee. COPYRIGHT 2011 Oracle, Inc. All rights reserved. Page - 2
ONLINE PROCESS RESULTS The table below shows average retrieval (search) and update (save) times, in onds, for each business process. Process 5,000 10,000 15,000 HRMS Process % within Group Employee Self-Service (60%) % Overall Pacing in Min Update Home Address 3% 1.8% 5 Update Phone Numbers 3% 1.8% 5 View Benefits Summary 10% 6% 5 Update Beneficiary 2% 1.2% 5 View Paycheck 78% 46.8% 5 Update Direct Deposit 2% 1.2% 5 Employee Adds Profile Items 2% 1.2% 5 Manager Self-Service (20%) View Employee Info 50% 10% 5 Initiate Termination 20% 4% 5 Initiate Promotion 10% 2% 5 Initiate Employee Salary Change HR Administrator (20%) 20% 4% 5 Add a Person 20% 4% 5 Hire a Person 40% 8% 5 Add a Job Row 40% 8% 5 Total 100% 5 Table 1: Business Process Mix The table above shows the proportions of the business processes used in the measurements of this benchmark. The proportions are intended to simulate a typical user scenario. The database and application servers were processing a total of 3,000 business processes per minute at the peak load of 15,000 concurrent users. The estimated transaction rate is calculated by dividing the total number of concurrent users by the average pacing rate. Update Home Address Search 0.341 0.481 0.67 Save 0.44 0.625 0.859 Update Home Phone Search 0.283 0.395 0.535 Save 0.29 0.414 0.562 View Benefits Summary View 0.433 0.607 0.83 Update Beneficiary Search 0.442 0.624 0.848 Save 1 0.064 0.087 0.132 Save 2 0.094 0.135 0.197 Edit/ Calc 0.033 0.047 0.07 View Paycheck Search 0.758 1.031 1.346 View 0.373 0.515 0.688 Update Direct Deposit Info Employee Adds Profile Items View Employee Info Search 0.277 0.377 0.529 Save 0.058 0.082 0.117 Search 0.243 0.314 0.445 Save 0.881 1.231 1.656 Submit 1.435 1.94 2.615 Confirm 1.015 1.349 1.773 Search 0.153 0.211 0.285 Initiate Termination Search 0.363 0.49 0.649 Save 0.093 0.129 0.165 Confirm 0.093 0.125 0.159 Initiate Promotion Search 0.378 0.515 0.677 Save 0.377 0.472 0.64 Initiate Salary Change Search 0.415 0.573 0.763 Save 0.361 0.524 0.684 Calc 0.061 0.084 0.116 Add a Person Save 0.143 0.205 0.273 Hire a Person Save 1 0.145 0.205 0.277 Save 2 0.553 0.759 1.018 Add a Job Search 0.488 0.601 0.735 Save 0.508 0.734 1.024 Confirm 0.848 1.290 1.030 * Average Search 0.574 0.776 1.013 * Average Save 0.332 0.467 0.634 Trans/min Est. 1,000 2,000 3,000 Table 2: Employee/Manager Process Runtimes Performance may vary on other hardware and software platforms and with other data composition models. COPYRIGHT 2011 Oracle, Inc. All rights reserved. Page - 3
SERVER PERFORMANCE Figure 3 shows the average CPU utilization for each of the servers in this test. The CPU utilization is the average across all of the CPUs in each server. % Average CPU Utilization Oracle Enterprise HCM 9.1 Self-Service with Oracle11g R2 on SPARC T4 s DB 60 Web 50 40 30 20 10 0 5,000 10,000 15,000 Figure 3: Average CPU Utilization % CPU User System I/O Wait Idle DB 15,000 6 2 0 92 10,000 3 1 0 96 5,000 1 1 0 98 15,000 36 4 0 60 10,000 19 3 0 79 5,000 6 1 0 93 Web 15,000 10 3 0 87 10,000 6 2 0 92 5,000 3 1 0 96 Table 3: Summary of CPU Utilization 5,000 10,000 15,000 DB 45 GB 45 GB 45 GB 94 GB 106 GB 119 GB Web 63 GB 72 GB 80 GB Table 4: Average Memory Utilization I/O PERFORMANCE The Sun Storage F5100 Flash Array storage was used for storage of tables, indexes and for undo and redo logs. I/O performance is crucial to performance and is summarized as follows: DB 5,000 10,000 15,000 Avg. r/s 0 0 0.4 Avg. w/s 248 489 703 Avg. KB r/s 0 0 3.02 Avg. KB w/s 1,994 3,928 5,636 Avg. r/s 25.2 59.1 52 Avg. w/s 218 482 801 Avg. KB r/s 266 643 479 Avg. KB w/s 2,872 7,492 13,546 Table 5: I/O Metrics DATA COMPOSITION DESCRIPTION The standard database was comprised of: 500,000 Employees (5 per Department) 100,000 Managers 100,000 Department Table Entries BENCHMARK ENVIRONMENT HARDWARE CONFIGURATION Database : 1 Oracle s SPARC T4-4 server was used as the database server. It was equipped with the following: 4 3.0 GHz SPARC T4 Eight-Core processors (sockets), each with 16 Kilobytes of Instruction and 16 Kilobytes of Data Level-1 on core cache, 128 Kilobytes of shared Instruction and Data Level-2 cache per core, and 4 Megabytes of Level-3 on-chip cache (32 cores total 256 vcpus) 256 Gigabytes of Memory (~45 GB used at peak load) 3 300 GB SAS internal disks 1 Oracle s Sun Storage F5100 Flash Array storage systems was equipped with the following: 80 24 GB FMODs (FM once formatted ~22.89GB.) Approximately 524 GB of storage out of ~900 GB (aggregate) from one ~1,800 GB (mirrored) Oracle s Sun Storage F5100 Flash Array storage system was allocated to the database instance. COPYRIGHT 2011 Oracle, Inc. All rights reserved. Page - 4
SOFTWARE VERSIONS Application (s): Oracle s PeopleSoft HCM (HRMS) 9.1 1 Oracle s SPARC T4-4 server was used as the application Oracle s PeopleSoft Enterprise (PeopleTools) 8.51.02 server. It was equipped with the following: Oracle Database 11g 11.2.0.1.0 (64-bit) 4 3.0 GHz SPARC T4 Eight-Core processors (sockets), each with 16 Kilobytes of Instruction and 16 Kilobytes of Oracle Solaris 10 8/11 (on the Database, and Web Data Level-1 on core cache, 128 Kilobytes of shared ) Instruction and Data Level-2 cache per core, and 4 Megabytes of Level-3 on-chip cache (32 cores total 256 vcpus) Microsoft Windows 2003 Enterprise Edition w/sp 2 (on the Controller and Drivers) 512 Gigabytes of Memory (~120 GB used at peak load) 5 300 GB SAS internal disks 1 300 GB internal SSD 2 100 GB internal SSD Web (s): 1 Oracle s SPARC T4-2 server was used as the web server. It was equipped with the following: 2 2.85 GHz SPARC T4 Eight-Core processors (sockets), each with 16 Kilobytes of Instruction and 16 Kilobytes of Data Level-1 on core cache, 128 Kilobytes of shared Instruction and Data Level-2 cache per core, and 4 Megabytes of Level-3 on-chip cache (16 cores total 128 vcpus) 256 Gigabytes of Memory (~80 GB used at peak load) 1 300 GB SAS internal disk 1 300 GB internal SSD Oracle ATS Load Test software 9.20 Oracle (BEA) Tuxedo 10.3.0.0 Patch Level 031 (64-bit) with Java SE 1.6.0_26 Runtime Environment and Java HotSpot 64-bit VM 20.1 Oracle WebLogic 11g (10.3.3) Load Simulation Driver(s): 3 Sun Blade X6240 servers were used as the load driver controller and drivers. They were equipped with the following: 1 2.3 Gigahertz AMD Opteron 2356 Quad-Core Processors, each with 2 Megabytes of Level-2 Cache (4 cores total) 32 Gigabytes of Memory Oracle (PeopleSoft) Pleasanton 5815 Owens Drive P. O. Box 8018 Pleasanton, California 94588-8618 Tel 925/694-3000 Fax 925/694-3100 Email info@peoplesoft.com World Wide Web http://www.oracle.com Copyright 2011, Oracle and/or its affiliates. All rights reserved. Oracle and Java are registered trademarks of Oracle and/or its affiliates. Other names may be trademarks of their respective owners. Intel and Intel Xeon are trademarks or registered trademarks of Intel Corporation. All SPARC trademarks are used under license and are trademarks or registered trademarks of SPARC International, Inc. AMD, Opteron, the AMD logo, and the AMD Opteron logo are trademarks or registered trademarks of Advanced Micro Devices. UNIX is a registered trademark licensed through X/Open Company, Ltd. All rights reserved. C/N 06XX-1011 COPYRIGHT 2011 Oracle, Inc. All rights reserved. Page - 5