Performance & Scalability of mySAP.com Bernd F. Lober SAP AG

Characteristics of scalability Standard Application Benchmarks Agenda Characteristics of scalability Performance indicators Standard Application Benchmarks Characteristics Published benchmark results

Multi-tier Internet Architecture Load Layer 2-tier 3-tier Multi-tier Internet Architecture 10-20% Presentation Services Presentation Services Web Browser User Dialog: Graphical information processing Presentation Web Server Web Server Handling Internet access 5-10% Internet Server Internet Server Processing Internet Transactions Internet Application Services Processing application logic: System Management Transaction Monitoring Production Create Orders Release Schedule Customer Accept Order Delivery Confirm Products Build Material Bill-of- Explode Reserve Service Rep Personnel Plant Part Task Application Services 60-70% Application Database Services 10-20% Database Services Information Storage Database Backup Database

Performance Indicator Response Time of Screen Change Response time of screen change depends Response time of the application server Can be Performance optimized Network runtime Latency Cannot be tuned On satellite connection at least 0.5 seconds Data transmission time Can be tuned Display of images on frontend

Associations With Scalability No system wait times, no bottlenecks Technical: Linear resource consumption Double CPU, double throughput No chaotic behavior Linear intrapolation and extrapolation possible Application is predictable Independent from hardware Same software on notebook and mainframe No add-ons for loading required

Vertical Scalability Multi-tier Internet architecture One-tier: Laptop Demo System Two-tier: Central Installation (database and application part on one physical server), multiple presentation servers Three-tier: One database server, multiple application servers and multiple presentation servers Multi-tier: One database server, multiple application servers multiple presentation servers, mutiple Internet Transaction Server and multiple Web servers

Horizontal Scalability Presentation Layer More than 47,000 very active users connected to one database have been tested Internet Transaction Server and Web Server More than 100 hits /sec Application Layer Up to 161 application servers have been tested successfully The highest number of physical application servers at customer installations is less than 32 Database Layer Scalability through SMP architecture of the database server More than 120 CPUs and more than 10 TB Database size Scalability through parallel databases Scalability through Components

Testing Performance with Standard Application Benchmarks Goals Represent real world applications as effectively as possible Compare the performance of computer systems from different vendors. Business scenarios SCM, CFM, for example: SD, MM, FI, APO, ATO Simulate user behavior or batch load Method is monitored and approved by SAP Benchmark Council Can be used in collaborative framework For example: SD (FI, MM)

Benchmark Characteristics Generate reproducible and representative load Generate scalable and configurable load Available for all platforms but platform independent Defined and controlled by SAP on behalf of the SAP Benchmark Council Represent typical processes in a customer environment Measure throughput (in SAPS) and dialog response times Analyze system configurations and parameter settings Cannot reflect all customer environments

SAP Standard Application Benchmarks mySAP Supply Chain Management mySAP Human Resources Materials Management - MM Cross Application Time Sheets - CATS Sales & Distribution - SD Payroll Production Planning - PP Warehouse Management - WM mySAP Financials Assemble - To - Order - ATO Financial Accounting - FI Advanced Planning and Optimizing - APO Industry Solutions mySAP Product Lifecycle Management Retail Project System - PS Banking - Bank Customer Accounts - BCA Utilities - Customer Care and Service - CCS mySAP Business Intelligence Business Information Warehouse - BW mySAP Customer Relationship Mgmt Internet Sales Customer Interaction Center E-commerce Online Store 39

2,000 fully processed order line items / hour* Definition: SAPS SAPS S AP A pplication Benchmark P erformance S tandard SD SD Bench- mark 2,000 fully processed order line items / hour* 100 SAPS * 6,000 dialog steps and 2,000 postings or 2,400 SAP transactions = ^

Dialog Steps SD Benchmark 0 Logon 1 Main screen 2 Call /nva01 (Create customer order) 3 1st screen 4 2nd screen (with 5 items) 5 [F11 - Save] 6 Call /nvl01 (Create a delivery) 7 1st screen 8 [F11 - Save] 9 Call /nva03 (Display customer order) 10 [Enter] 11 Call /nvl02 (Change delivery) 12 [F9] (Posts goods issue) 13 Call /nva05 (List orders) 14 [Enter] 15 Call /nvf01 (Create invoice) 16 [F11 - Save] 17 Call /nend 18 Confirm logoff Dialog steps 2 to 16 are repeated n times (15 dialog steps -> min. 150 sec duration). Business aspect: One run (dialog steps 2 to 16) corresponds to the selling of 5 items.

Dialog Steps ATO Benchmark 0 Logon 1 Main screen 2 Call /nva01 (Create customer order) 3 Enter order & organizational data 4 Enter customer and material 5 1st level characteristic value assignment 6 2nd level characteristic value assignment 7 2nd level characteristic value assignment 8 Control of resulting price 9 Create assembly order 10 Call /nmf44 (Make-to-Order Backflush) 11 Enter sales order data 12 Save 13 Call /nvl01 (Create a delivery) 14 Select sales order 15 Select items 16 Call /nlt03 (Create transfer order) 17 Save transfer order 18 Call /nlt12 (Confirm transfer order) 19 Confirm transfer 20 Call /nso01 (SAP Office - Inbox) 21 Select Workflow-Inbox 22 Start goods issue via workflow. 23 Call /nvf01 (Create invoice) 24 Save delivery note 25 Logoff or start process again 26 Logoff Dialog steps 2 to 24 are repeated n times (23 dialog steps -> min. 230 sec duration). Business aspect: One run corresponds to the full assemble-to-order scenario for one item.

CPU Utilization of Different Servers

Published Results for SD Benchmarks (Two-tier Client/Server Configuration) Two-tier Internet Architecture 300 600 1410 1708 3000 4100 7800 1000 2000 4000 5000 6000 7000 8000 9000 Rel. 2.2 C 8-way Server Rel. 3.0 E 64-way Server Rel. 4.0 B 24-way Server Rel. 2.2 G Rel. 4.6 C 76-way Server 128-way Server 1993 1994 1995 1996 1997 1998 1999 2001 2002 2003 Number of SD Benchmark users 42 25

Published Results for SD Benchmarks (Highest Number of Users) 1993 1994 1995 1996 1997 1998 25560 1999 2000 2001 2002 50000 47528 45000 47008 Number of SD Benchmark users 40000 35000 30000 26000 25560 25000 23000 24000 20000 19360 16440 15000 14400 10000 5320 5000 3700 2000 1700 300 710 1400 120 Three-tier Internet Architecture

Key Figures for the 47,528 SD User Benchmark 3,660 DB transactions per second (commits) 8,300 MB average disk write per second 75 MB average disk read per second 256 GB data blocks written in one hour 54 GB data blocks read in one hour 314,223 network packets per second with an average packet size of 859 bytes 151,250 statements and 51,620 rows inserted per second! 6,228 GB disk space 1,600 SAP transactions / second 3,999 Screen changes / second Database used a 32-processors SMP server R/3 was running on 424 processors 4,799,330 fully business processed order line items per hour ( 1,333 per second ) 31

Published Results for SD Benchmarks (Highest Number of Line Items) 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 5000000 4799330 4500000 4713000 Fully business-processed order line items per hour 4000000 3500000 3000000 2573330 2606000 2500000 2353670 2411330 2000000 1954670 1669000 1575000 1500000 1000000 534000 380000 500000 230000 33000 81000 12000 175000 146000 Three-tier Internet Architecture

Published Results for SD Benchmarks (Highest Number of SAPS) 1999 1998 2000 1993 1994 1995 1996 1997 2001 2002 250000 239970 Highest Number of SAPS 235650 200000 150000 130300 128670 120570 117680 100000 97370 83450 78750 50000 26700 11517 19117 8766 8820 600 3110 4500 Three-tier Internet Architecture

Published Results for the ATO Benchmark (3-Tier) 2001 2000 1999 1998 9130 22610 54220 130570 144090 20000 40000 60000 80000 100000 120000 140000 160000 2002 Highest Number of Assembly Orders

Published Result for the APO Benchmark (Demand Planning) 2000 2001 2002 2003 500000 474162 450000 400000 350000 336510 300000 250000 200000 150000 129871 100000 79611 50000 53199 Number of Planned Characteristic Combinations at Aggregated Level Per Hour

www.sap.com/benchmark

Comments From an Analyst AberdeenGroup in August 2002 on the SAP Benchmark "Over the past three to four years, the three-tier SAP SD benchmark has replaced TPC-C as the most credible and influential high-end OLTP benchmark in the industry." "It is today’s most visible benchmark for “size matters” zealots who seek the largest, most scalable platforms for real-world applications. "Not only is SAP the market leader in ERP software, but its benchmarking and certification program […] is acknowledged as the most effectively managed and “well-policed” platform benchmarking activity in the industry."

Summary Proven Scalability - with enough potential for future growth mySAP.com exploits the latest available components Useful addresses: http://www.sap.com/benchmark http://service.sap.com/performance http://service.sap.com/sizing

Questions And Answers

Copyright 2002 SAP AG. All rights reserved No part of this publication may be reproduced or transmitted in any form or for any purpose without the express permission of SAP AG. The information contained herein may be changed without prior notice. Some software products marketed by SAP AG and its distributors contain proprietary software components of other software vendors. Microsoft®, WINDOWS®, NT®, EXCEL®, Word®, PowerPoint® and SQL Server® are registered trademarks of Microsoft Corporation. IBM®, DB2®, OS/2®, DB2/6000®, Parallel Sysplex®, MVS/ESA®, RS/6000®, AIX®, S/390®, AS/400®, OS/390®, and OS/400® are registered trademarks of IBM Corporation. ORACLE® is a registered trademark of ORACLE Corporation. INFORMIX®-OnLine for SAP and Informix® Dynamic ServerTM are registered trademarks of Informix Software Incorporated. UNIX®, X/Open®, OSF/1®, and Motif® are registered trademarks of the Open Group. Citrix®, the Citrix logo, ICA®, Program Neighborhood®, MetaFrame®, WinFrame®, VideoFrame®, MultiWin® and other Citrix product names referenced herein are trademarks of Citrix Systems, Inc. HTML, DHTML, XML, XHTML are trademarks or registered trademarks of W3C®, World Wide Web Consortium, Massachusetts Institute of Technology. JAVA® is a registered trademark of Sun Microsystems, Inc. JAVASCRIPT® is a registered trademark of Sun Microsystems, Inc., used under license for technology invented and implemented by Netscape. SAP, SAP Logo, R/2, RIVA, R/3, SAP ArchiveLink, SAP Business Workflow, WebFlow, SAP EarlyWatch, BAPI, SAPPHIRE, Management Cockpit, mySAP.com Logo and mySAP.com are trademarks or registered trademarks of SAP AG in Germany and in several other countries all over the world. All other products mentioned are trademarks or registered trademarks of their respective companies.