WebBench(TM) 1.1 Frequently Asked Questions

This file contains answers to some commonly asked questions about WebBench(TM). For more information about WebBench, see the online manual "The WebBench(TM) Tester's Handbook."

General questions about WebBench

Getting a copy of WebBench and installing it

If you have a problem

Deciding when to use WebBench

Running WebBench's test suites

WebBench's results

About ZDBOp

General questions about WebBench

Q. What is WebBench?
A.
WebBench is a Ziff-Davis benchmark that lets you measure the performance of Web servers. WebBench's primary results are two overall server scores -- requests per second and throughput as measured in bytes per second.

WebBench runs on a server networked to a PC running the controller program and one or more PCs each running the client program. The controller runs on Windows NT® 3.51/4.0 and the clients run on both Windows® 95 and Windows NT.

To help you test Web servers, WebBench provides two types of standard test suites: static suites that involve only HTML pages, and dynamic suites, which use dynamic executables and static requests. Different dynamic test suites access Common Gateway Interface (CGI) executables, Internet Server API (ISAPI) and Netscape Server API (NSAPI) dynamic link libraries, NSAPI shared objects, and an IntranetWare local-CGI NetWare loadable Module (NLM). You can use WebBench's static test suites to test any Web server. The dynamic test suites, due to the nature of dynamic executables, are platform dependent. WebBench's dynamic test suites using a CGI application run on the following server platforms: Apple’s Mac OS 7.53, Digital UNIX with Alpha processors, IBM’s OS/2 Warp Server, Linux with x86 processors, Microsoft’s Windows NT 3.51/4.0 with x86 processors, Silicon Graphics’ IRIX, and Sun’s Solaris 2.5.1 on SPARC. WebBench provides additional dynamic test suites where the CGI application has been rewritten as a Windows NT Microsoft Internet Server API dynamic link library (ISAPI.DLL), a Windows NT Netscape Server API dynamically loadable library (NSAPI.DLL), an IntranetWare Local-CGI Netware Loadable Module (L-CGI NLM), and a Solaris on SPARC, Digital UNIX on Alpha, and IRIX NSAPI shared object (.so).

Q: Are Ziff-Davis benchmarks "freeware" or "shareware"?
A:
Our benchmarks are licensed software. While the software is free, you'll need to pay your normal connection time and downloading charges (if any) if you download a benchmark from one of ZD's online services. (If you order a CD-ROM, you'll need to pay a shipping and handling fee.) You must license and register your copy of the software before you can use it. WebBench displays its license agreement the first time you run it.

Go back to the top of the page.

Getting a copy of WebBench and installing it

Q. How can I get a copy of WebBench and other Ziff-Davis benchmark programs?
A.
You can download WebBench (and most of the other ZD benchmarks) from the following Ziff-Davis online ZDNet(TM) services:

You'll need to download two WebBench self-extracting files: one for the controller program (WBCONTLR.EXE) and one for the client program (WBCLIENT.EXE). You'll have to pay your normal connection time or downloading charges when you download these files. You can also get a copy of WebBench by requesting the Ziff-Davis Server Benchmarks CD-ROM, which contains each ServerBench® 4.0 port and NetBench® 5.01. (WebBench is the Ziff-Davis industry-standard Web server benchmark program, and ServerBench is the Ziff-Davis industry-standard client/server benchmark program.) In addition, Ziff-Davis provides PC benchmark programs on the Ziff-Davis PC Benchmarks CD-ROM and a Mac OS system benchmark program on the Ziff-Davis Macintosh® CD-ROM.

The PC Benchmarks CD-ROM contains:

The 3D WinBench CD-ROM contains 3D WinBench® 97, a subsystem-level benchmark that measures the performance of a PC's 3D subsystem, which includes the Direct3D software, the monitor, the graphics adapter, the graphics driver, and the bus used to carry information from the graphics adapter to and from the processor subsystem.

The Macintosh Benchmarks CD-ROM contains MacBench® 4.0, a subsystem-level benchmark for Mac OS systems.

You can get copies of these CD-ROMs from the Ziff-Davis Benchmark Operation (ZDBOp). You'll need to pay a shipping and handling fee of $5 (US), $6 (Canada), or $7 (International) for your first CD-ROM. Add $1 for each additional CD-ROM you request. Send credit card information (a VISA, MasterCard, or American Express credit card number, the expiration date, and the name as it appears on the card) via:

Or send a check or money order to:

Ziff-Davis Benchmark Operation
1001 Aviation Parkway, Suite 400
Morrisville, N.C. 27560
ATTN: Distribution

NOTE: We don't accept cash nor do we accept purchase orders for amounts under $200.

A CD-ROM takes four to six weeks to arrive. To receive it sooner, include your Federal Express account number and shipping instructions with your request. ZDBOp will then send the benchmark to you via Federal Express and charge the Fedex shipping cost to your account.

Go back to the top of the page.

Q. Do I have to buy anything special to run WebBench?
A.
You must have the Web server software you want to test and Excel to view your results. You don't need to buy anything else. WebBench's client and controller programs are self-contained and communicate via the TCP/IP network protocol that comes with Windows 95 and Windows NT. (This is the protocol that computers on the Web use to communicate.)

Q. Do I load special WebBench code onto my server?
A.
WebBench doesn't use a server program, but it does require you to install its test files in the HTML document root on the Web server. In addition, if you're running WebBench on one of the server platforms that it supports for its dynamic tests, you'll need to move its dynamic executable to the correct directory on the server. (If this directory isn't called CGI-BIN, you'll need to create an alias to CGI-BIN. If your system won't allow you to create an alias, you'll need to modify the workload file. See Chapter 6 in "The WebBench(TM) Tester's Handbook" for information on how to modify workload files.)

Q. How do I install WebBench?
A.
The basic WebBench installation is easy. You run the controller's Windows-based SETUP.EXE program to install WebBench on the controller and the client's Windows-based SETUP.EXE program to install WebBench on each client. If you downloaded WebBench, you'll need to execute each self-extracting file in a separate directory to get the files you need. If you're installing WebBench from the CD-ROM, you'll find these files in WB11\CONTROLR and WB11\CLIENT.

As part of the client installation, WebBench asks you for a unique ID number for each client. We recommend you number the clients sequentially, starting at 1. (You may also want to spread the clients out across different network segments.) The controller must be running Windows NT 3.51 or 4.0, and the clients can be running either Windows 95 or Windows NT.

Detailed installation instructions are in "The WebBench(TM) Tester's Handbook." We recommend you use this manual when you install WebBench. WebBench automatically places this manual in Word 7.0 for Windows format in the DOCS directory in the WebBench installation directory on the controller. This manual is also the basis for WebBench's online help file.

Once you've run SETUP.EXE on the clients, you'll need to edit each client's HOSTS file to include the IP address for each server you want to test against and the IP address for the controller. While you can use any name you choose for the server (the standard WebBench test suites use the default name "server"), you must use the name "controller" for the Windows NT workstation you're using as the controller.

NOTE: By default, the clients try to connect to the server at port 80 and the controller at port 5090. If you want to change these ports, you'll need to edit the SERVICES file of each client's TCP/IP stack. If you change the port number that the clients use to connect to the controller, you'll also need to edit the SERVICES file on the controller to change the port number that the controller listens for the clients on.

After you install the controller and client programs, you'll need to place the WebBench test files on the server. These files are in compressed format in the WORKLOAD directory, which is in the WebBench installation directory on the controller. Choose the version of the compressed file that you can expand on your server:

If you’re running WebBench on: You’ll need the file:
A Digital UNIX server with an Alpha processor DEC_Alpha_tar.Z
An IntranetWare server running NetWare Web Server netware.exe
An IRIX server IRIX_SGI_tar.Z
A Linux server with an x86 processor Linux_x86_tar.Z
A Macintosh server MAC.HQX
An OS/2 Warp Server os2.EXE
A Solaris for SPARC server Solaris_SPARC_tar.Z
A Windows NT server with an x86 processor NT_x86.EXE

Move that file to your server and expand it. You'll then have the directory WB10TREE and the WebBench dynamic executable(s). Move WB10TREE to the HTML document root on the Web server (this is where the Web server looks for its HTML files). If you're running WebBench on a server platform that it supports for its dynamic tests, you'll need to move the dynamic executable to your Web server's CGI-BIN directory. This is the directory where the Web server looks for its dynamic scripts. If this directory has a different name on your server, you'll need to create the alias CGI-BIN for that directory. Make sure CGI-BIN and the directory it is in have execute permission. You may also need to do some additional setup before you can use the dynamic executable. See "The WebBench(TM) Tester's Handbook" for more information.

NOTE: If your Web server can’t expand one of the above files, you can expand the NT_x86.EXE file on your controller and then place the directory containing the test files, WB10TREE, in the correct directory on your server. The test files take up about 3 MB, so you should expand NT_x86.EXE on a network directory from which you can simply move the WB10TREE directory tree to your server. If you need to put WB10TREE on a 1.44 MB diskette, you’ll have to compress WB10TREE using the compression tool of your choice. This way you'll be able to run WebBench's static test suites against any Web server platform.

You can get copies of the WebBench handbook by:

Q. Is there a short test that I can run just to make sure I've set everything up correctly?
A.
Sure. Run VERIFY.TST. This test suite verifies that the controller, clients, and server are all communicating properly. If VERIFY.TST finishes without any errors, you should be able to run WebBench's static test suites without any problem. (To determine whether any errors occurred, look in the results workbook and see if WebBench created Table 9: Error Information. WebBench only creates this table when errors occur during a test run.)

NOTE: VERIFY.TST doesn't perform dynamic requests so it does not verify that the dynamic executable is installed correctly. WebBench's dynamic test suites are platform dependent, so we would have had to create a separate VERIFY.TST for each dynamic platform.

Q. How can I make sure that I have installed the dynamic executable correctly?
A.
There are two ways to do this. Pick the method that is easiest for you. If you learn you haven't installed the dynamic executable correctly, look in Chapter 2 of "The WebBench(TM) Tester's Handbook" for detailed instructions on installing the dynamic executable. However, before you check the dynamic executable, make sure you've installed the static HTML test files correctly. See the previous question for information on how to check your test files.

Method 1 for checking the dynamic executable:
This is a quick and simple method that assumes you have a basic understanding of the Internet, servers, and browsers.

1. Install a browser such as Netscape Navigator(TM) or Microsoft Internet Explorer on one of the clients. 2. From the browser, try to open the correct dynamic executable for your server platform:

Digital UNIX server, an IRIX server, or a Solaris server http://test_server/cgi-bin/simcgi
http://test_server/cgi-bin/simnsapi.so
IntranetWare running NetWare Web Server 2.51 http://test_server/cgi-bin/simlcgi.nlm
Linux server http://test_server/cgi-bin/simcgi
Mac OS http://test_server/cgi-bin/simcgi.cgi
OS/2 Warp Server http://test_server/cgi-bin/simcgi.exe
Windows NT http://test_server/cgi-bin/simcgi.exe
http://test_server/wb10tree/simnsapi.zdb
http://test_server/cgi-bin/simisapi.dll

where test_server is the name of your server as it appears in the client's HOSTS file. (See Chapter 2 of "The WebBench(TM) Tester's Handbook" for information on setting up the HOSTS file.)

3. If the browser displays text similar to the following, you've installed the dynamic executable correctly.

SERVER_SOFTWARE = Microsoft-IIS/3.0
SERVER_NAME = 127.0.0.1a
GATEWAY_INTERFACE = CGI/1.1
SERVER_PROTOCOL = HTTP/1.0
SERVER_PORT = 80
REQUEST_METHOD = GET
HTTP_ACCEPT = image/gif, image/x-xbitmap, image/jpeg, */*
HTTP_USER_AGENT = Mozilla/1.22 (compatible; MSIE 2.0d; Windows NT)
HTTP_REFERER =
PATH_INFO =
PATH_TRANSLATED = C:\InetPub\wwwroot
SCRIPT_NAME = /cgi-bin/simcgi.exe
QUERY_STRING =
REMOTE_HOST = 127.0.0.1
REMOTE_ADDR = 127.0.0.1
REMOTE_USER =
AUTH_TYPE =
CONTENT_TYPE =
CONTENT_LENGTH = 0
ANNOTATION_SERVER =

Method 2 for checking the dynamic executable:
Instead of using a browser, just try to run a test suite that accesses the dynamic executable. If the test suite doesn't cause any errors, you've installed the executable correctly. To see the steps for running a test suite, look in Chapter 3 of "The WebBench(TM) Tester's Handbook." When WebBench asks you which test suite to run, choose the appropriate test suite for your server platform. See Chapter 5 of "The WebBench(TM) Tester's Handbook for a list of test suites.

You may immediately get the following error on the client:

Fatal Error from Server: 404 File Not Found

This error means you haven't installed the dynamic executable correctly.

If you don't get any errors on the client, wait for the first mix to end (about five minutes). You'll know when the first mix has finished because four clients will be at the Run stage (R) and the rest will be at the excluded stage (X). The Test Suite Progress bar will read "1/16 mixes."

Click on the Test Suite Progress bar (the completed portion of the bar). In the Results for Suite window that pops up, check to see how many errors are in the Total Errors column. As a rule of thumb, if there are more than 10 errors, you probably haven't installed the dynamic executable correctly.

Go back to the top of the page.

If you have a problem

Q. What do I do if I have a problem with WebBench?
A.
Check "The WebBench(TM) Tester's Handbook" to make sure you have your system set up correctly. If you don't find the solution in the manual, check WebBench's README.WRI for a list of known problems. (See the question on installing WebBench for information on getting a copy of the manual and README.WRI file.) You can also look though the rest of this file to see if someone else has had the same problem.

If you don't find the answer to your problem in one of these places, report the problem using the Problem Report Form at the back of the WebBench manual. To submit a report:

Mail the form to ZDBOp:

Ziff-Davis Benchmark Operation
1001 Aviation Parkway, Suite 400
Morrisville, N.C. 27560
Attention: WebBench Technical Support

When you contact us, include very specific information on how you've set up your system (server, controller, and clients) and what you were doing when WebBench encountered the problem.

If you have an e-mail address, be sure to include that as well.

Q. I've created a test suite that uses five threads per client. When I run this suite for more than about an hour, the scores start dropping. Do you know what's going on?
A.
We've noticed a problem on Windows 95 clients when we run them continuously with multiple threads for at least an hour and a half. At that time the PC begins doing a lot of disk activity that is usually associated with page swaps. If you want to run test mixes using multiple threads on each client, we recommend you use Windows NT clients.

Q. I've been running WebBench tests for several days and, not only are my results getting lower, but they're no longer repeatable. What's up?
A.
Check to see if logging is enabled on your Web server. If it is, you'll want to disable it. In our testing we've discovered that if you run WebBench on a Web server that has logging enabled, over time the WebBench scores may go down and the test results may not be repeatable. This drop in performance and repeatability occurs as the size of the log file increases. We recommend that you disable logging on your Web server software before you run WebBench.

Q. I'm running WebBench against a server running Windows NT (which WebBench supports for dynamic tests), but I'm seeing a lot of errors even at low client counts. The static test suites seem just fine. Why won't WebBench run the dynamic tests?
A.
Make sure you put the dynamic executables in the correct directory. If this directory isn't called CGI-BIN, you'll need to create an alias to CGI-BIN (this is the directory WebBench looks for). Some Web servers may not allow you to create a CGI-BIN alias to this directory. In this case, you'll need to edit the workload files by hand so that they point to the right URL for the dynamic executable. (See Chapter 6 in "The WebBench(TM) Tester's Handbook" for information on how to modify the workload files.)

Q. When I start WebBench on the clients and the controller, I get an error message on the clients. Why is this?
A.
WebBench requires you to start its executables in a specific order. First, start WebBench on the controller. When the controller displays a pop-up dialog box asking if you want to connect the clients, go to each client and start the WebBench client executable. If you don't start WebBench in this order, the clients will error out.

Q. Can I compare my WebBench 1.0 results with my WebBench 1.1 results?
A.
You can compare your WebBench 1.1 results with your WebBench 1.0 results for all common test suites except the Windows NT CGI test suites. We changed the Windows NT CGI executable that ZD_NT_SIMPLE_CGI20_V11.TST and ZD_NT_SIMPLE_CGI20_KEEPALIVE_V11.TST use because of a problem with some compiler libraries. As a result of the change, this executable no longer generates numerous errors when you run it on multiple processor configurations.

Go back to the top of the page.

Deciding when to use WebBench

Q. Why should I use WebBench?
A.
WebBench gives you a way to compare different Web servers to each other. For example, you can run different Web server packages on the same server hardware. Or you can compare how different server hardware platforms run a given Web server package.

To get a valid comparison, you must run WebBench the same way on all servers you include in the comparison: run the same test suite, use the same parameters (where possible) on all the servers, and use the same testbed.

Go back to the top of the page.

Running WebBench's test suites

Q. How many clients can I have connected during a test run?
A.
WebBench supports up to 1,000 clients; however, your operating system or network software may limit you to fewer than that. Don't worry, though. Because WebBench uses stress tests, you can get an accurate measure of your server's performance with as few as eight or ten clients. WebBench's standard test suites begin with one client and add clients incrementally until they reach a maximum of 60 clients.

Q. I'm confused by all these tests. Which ones should I run?
A.
WebBench provides numerous test suites. If you're not running WebBench against one of the server platforms it supports for dynamic tests, we recommend you execute ZD_STATIC_V11.TST. The tests in this test suite access static content on the server and are independent of the server architecture. In addition, this test suite doesn’t use keep-alives (an HTTP protocol that some servers ignore). It takes about an hour and a half to run this test suite.

Go back to the top of the page.

WebBench's results

Q. How does WebBench report its results?
A.
WebBench reports two overall results for the server: requests per second and throughput as measured in bytes per second. In its results workbook, WebBench also reports other information, including detailed results for the individual clients. To see the overall server scores, go to Table 1. To see the individual client scores, go to Table 4.

Q. What do my results mean?
A.
Your results indicate how well your Web server handles client requests. Higher numbers indicate better performance.

Q. I want to publish my WebBench results. What's involved in getting consent to publish these results?
A.
The WebBench License Agreement tells you what information you need to include when you publish your test results. We added Tables 5 and 6 to the results workbook to provide a central location for this information. You can set up your results database so that you'll have the information available when you need it. See Chapter 4 "Looking at Your Test Results" in "The WebBench(TM) Tester's Handbook" for information on creating a results database.

Q. I want to compare my server's results against results for other servers. Where can I find some comparison test results?
A.
PC Magazine®, PC WEEK® and other Ziff-Davis publications publish articles about WebBench.

Go back to the top of the page.

Questions about ZDBOp

Q. What is ZDBOp?
A.
ZDBOp is a division of Ziff-Davis dedicated to the research and development of the core, publicly available benchmarks ZD publications worldwide use for performance testing.

Q. Where can I get support for the older versions of the benchmarks?
A.
Ziff-Davis does not support older versions of the benchmarks. When we release an updated version of a benchmark, all ZD publications move to that new version. We encourage you to do likewise.

Q. Could I be involved in beta-testing the next releases of the benchmarks?
A.
Sorry, but we don't use external beta sites for our benchmarks. Although we welcome input from the user and vendor communities, we are first and foremost journalists and can't afford to let some--but not all--vendors see early releases of the benchmarks.

Go back to the top of the page.

Trademarks

MacBench®, NetBench®, PC Magazine®, PC WEEK®, ServerBench®, WinBench®, and Winstone® are registered trademarks and WebBench(TM), ZD INTERNET Magazine(TM), and ZDNet(TM) are trademarks of Ziff-Davis Inc.

Alpha(TM) and Digital(TM) are trademarks of Digital Equipment Corporation.

CompuServe® is a registered trademark of CompuServe, Inc.

Intel® and Pentium® registered trademark of Intel Corporation.

IRIX(TM) is a trademark of Silicon Graphics, Inc.

Mac® and Macintosh® are registered trademarks of Apple Computer, Inc.

Microsoft®, MS-DOS®, Windows®, and Windows NT® are registered trademarks of Microsoft Corporation.

Netscape(TM) and Netscape Navigator(TM) are trademarks of Netscape Communications Corporation.

NetWare® is a registered trademark and IntranetWare(TM) is a trademark of Novell, Inc.

Solaris(TM) is a trademark of Sun Microsystems, Inc.

SPARC® is a registered trademark of SPARC International, Inc. Products bearing the SPARC trademarks are based on an architecture developed by Sun Microsystems, Inc.

UNIX® is a registered trademark in the United States and other countries, licensed exclusively through X/Open Company, Ltd.

Go back to the top of the page.

=========================================================================

WEBBENCH(TM) 1.1 README.WRI

This is the README.WRI file for WebBench 1.1. You can view this file from the Windows Write program. This file contains the WebBench License Agreement, installation instructions, and important notes about WebBench.

Before you run WebBench, you must first read and agree to the License Agreement. We include this License Agreement in this file, in the WebBench documentation, and also when you run WebBench for the first time.

What's in This README.WRI File

This README.WRI file contains the following sections:

* License Agreement for Ziff-Davis' WebBench(TM) Version 1.1
* Getting the Files You Need
* Installing WebBench(TM) Version 1.1
* Where to Find Detailed Installation Instructions
* What's New About WebBench 1.1
* Important WebBench Notes
* Documentation
* Technical Support
* Trademarks

LICENSE AGREEMENT FOR WEBBENCH(TM) VERSION 1.1

READ THIS AGREEMENT CAREFULLY BEFORE USING THE SOFTWARE EMBODIED IN THE WEBBENCH(TM) 1.1 CD-ROM, OR, IF PRELOADED ON YOUR HARD DRIVE, DOWNLOADED OR IF PROVIDED AS PART OF A COLLECTION, THE PRELOADED, DOWNLOADED OR COLLECTED FILE(S) (the "Media"). Embodied in the WebBench(TM) 1.1 Media is the WebBench(TM) version 1.1 computer programs and related documentation (the "Software"). Ziff-Davis Inc., having a place of business at One Park Avenue, New York, New York 10016 ("Ziff-Davis"), is the licensor under this Agreement and you are the licensee. By using the Software, in whole or in part, you agree to be bound by the terms of this Agreement. If you do not agree to the terms of this Agreement, promptly return the Software to Ziff-Davis Benchmark Operation at 1001 Aviation Parkway, Suite 400, Morrisville, North Carolina 27560 (or, if downloaded or preloaded on your hard drive, delete the Software, or if provided as part of a collection, cease use of the Software). Title to the software and all copyrights, trade secrets and other proprietary rights therein are owned by Ziff-Davis. All rights herein, except those expressly granted to you in this Agreement, are reserved by Ziff-Davis.

1. Limited License

This Agreement grants you only limited rights to use the Software. Ziff-Davis grants you a non-exclusive, non- transferable license to use the Software on a Web server networked with multiple PC computers for the sole purpose of conducting benchmark tests to measure the performance of computer hardware and Web server software. You have the right to make a single copy of the Software for archival purposes and the right to transfer a copy of the Software across an internal local area network only to the PC computers attached to such network; provided, however, that all such copies are considered Software hereunder, that all uses of such copies are governed by the terms and conditions of this Agreement and that you shall be responsible for all uses of such copies in violation of the terms and conditions of this Agreement. Ziff-Davis hereby grants you the right to publish, except in any country where a third party claims during the term of this license that such publication infringes that party's proprietary rights, benchmark test results obtained by you from your use of the Software, provided that with the publication of each such result you:

A. Identify Ziff-Davis, the name and version number of the benchmark Software used and the name of the test suite used in the test (i.e., Ziff-Davis' WebBench(TM) version 1.1 with the standard test suite ZD_STATIC_V11.TST);

B. Identify for the Web server the exact name and version number of the Web server software package, any tunable parameters (e.g., logging) you set (unless you used the default values), the exact name, number of processors, processor speed (including clock speed) and type(s), amount of processor level 2 (L2) cache, amount of RAM, type of I/O bus, number and type of hard disk controller(s), size of hardware hard disk cache, if any, number and type of hard disks, hard disk size, type of disk organization (e.g., RAID 5), driver version of the disk controller(s), number and type of network controllers, driver version of network controller, network operating system name and version, and any relevant modifications to the default network operating system parameters (e.g., ABC Inc. Web X with logging disabled running on WXY Corp. Model 466 with a dual 166-MHz Pentium® processor, 512KB L2 cache, 128MB of RAM, bridged PCI/EISA I/O bus, WXY Integrated Drive Array disk controller, 4 ABC 1.2GB disk drives, RAID 0 hardware striping, a xxxxxxxx.dsk 12,621 4/29/96 disk driver, WXY 100BaseT network controller, a WXY 100BaseT driver version X.X, Microsoft® TCP/IP, Microsoft Windows® Sockets Version 1.1, Windows NT® Server 3.51);

C. Identify for the client testbed the network type, the number of clients, the client operating system version and service pack version, if any, (e.g., Microsoft Windows 95 or Windows NT 3.51 with Service Pack 1), the client CPU type and speed in percentages, amount of RAM, the number and type of hubs/concentrators, the number of clients per segment, client network software name and version (drivers, protocols, redirector; e.g., Microsoft TCP/IP, Microsoft Windows Sockets Version 1.1, > NE2000.386), client network card, size of the client network cache, if any (e.g., 100BaseT with 32 clients, 50% of the clients are 486/25-based PCs running Windows 95 with Service Pack 1, 50% of the clients are 100Mhz Pentium PCs running Windows NT 3.51 with Service Pack 4, 16MB RAM, 2 XYZ 100BaseT Ethernet hubs, 16 clients per segment, 50% of the clients have ABC 100BaseT network cards and 50% of the clients have GHI 100BaseT network cards, Microsoft Client for Microsoft Networks, Microsoft TCP/IP, Microsoft Windows Sockets Version 1.1, xxxxxx.SYS LAN driver);

D. Identify the controller operating system version, network software and version, and network card (e.g., Windows NT, Microsoft Client for Microsoft Networks, Microsoft TCP/IP, Microsoft Windows Sockets Version 1.1, xxxxxx.SYS LAN driver, ABC 100BaseT network card);

E. Identify any other special conditions used to achieve the result;

F. State that all products used in the test were shipping versions available to the general public;

G. State that the test was performed without independent verification by Ziff-Davis and that Ziff-Davis makes no representations or warranties as to the results of the test; and

H. Follow proper trademark usage and acknowledge Ziff-Davis' trademark rights (e.g., "[ ] achieved a WebBench(TM) overall score of X requests per second and a throughput score of X bytes per second. WebBench(TM) is a trademark of Ziff-Davis Inc. in the U.S. and other countries.").

This Agreement and your rights hereunder shall automatically terminate if you fail to comply with any provision of this Agreement. Upon such termination, you shall cease all use of the Software, cease the transfer of any copies of the Software and cease the publication of benchmark test results obtained by you from use of the Software. Further, you shall delete the Software and destroy all tangible copies of the Software and other materials related to the Software in your possession or under your control, or, if downloaded or preloaded on your hard drive or if provided as part of a collection, you shall cease use of and destroy any and all copies of the Software in your possession or under your control.

2. Additional Restrictions

A. You shall not (and shall not permit other persons or entities to) rent, lease, sell, sublicense, assign, or otherwise transfer the Software or this Agreement. Any attempt to do so shall be void and of no effect.

B. You shall not (and shall not permit other persons or entities to) reverse engineer, decompile, disassemble, merge, modify, include in other software or translate the Software, or use the Software for any commercial purposes, except for the publication of test results, as provided above.

C. You shall not (and shall not permit other persons or entities to) remove or obscure Ziff-Davis' copyright, trademark or other proprietary notices or legends from any of the materials contained in this package or downloaded.

D. You acknowledge that the Software contains Ziff-Davis' trade secret information and you shall not disclose or disseminate such information other than as provided herein.

3. Disclaimer of Warranty; Limitation of Liability

THE SOFTWARE AND THE MEDIA ARE PROVIDED "AS IS" WITHOUT WARRANTY OF ANY KIND, EITHER EXPRESS OR IMPLIED, INCLUDING, WITHOUT LIMITATION, ANY WARRANTY OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE.

THE ENTIRE RISK AS TO THE RESULTS AND PERFORMANCE OF THE SOFTWARE AND THE MEDIA IS ASSUMED BY YOU, AND ZIFF-DAVIS AND ITS AUTHORIZED DISTRIBUTORS ASSUME NO RESPONSIBILITY FOR THE ACCURACY OR APPLICATION OF OR ERRORS OR OMISSIONS IN THE SOFTWARE OR THE MEDIA. IN NO EVENT SHALL ZIFF-DAVIS OR ITS AUTHORIZED DISTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, SPECIAL, INCIDENTAL OR CONSEQUENTIAL DAMAGES ARISING OUT OF THE USE OR INABILITY TO USE THE SOFTWARE OR THE MEDIA, EVEN IF ZIFF-DAVIS OR ITS AUTHORIZED DISTRIBUTORS HAVE BEEN ADVISED OF THE LIKELIHOOD OF SUCH DAMAGES OCCURRING. ZIFF-DAVIS AND ITS AUTHORIZED DISTRIBUTORS SHALL NOT BE LIABLE FOR ANY LOSS, DAMAGES OR COSTS, ARISING OUT OF, BUT NOT LIMITED TO, LOST PROFITS OR REVENUE, LOSS OF USE OF THE SOFTWARE OR THE MEDIA, LOSS OF DATA OR EQUIPMENT, THE COSTS OF RECOVERING SOFTWARE, THE MEDIA, DATA OR EQUIPMENT, THE COST OF SUBSTITUTE SOFTWARE OR THE MEDIA, DATA OR EQUIPMENT OR CLAIMS BY THIRD PARTIES, OR OTHER SIMILAR COSTS.

SOME STATES DO NOT ALLOW EXCLUSION OR LIMITATION OF IMPLIED WARRANTIES OR LIMITATION OF LIABILITY FOR INCIDENTAL OR CONSEQUENTIAL DAMAGES; SO THE ABOVE LIMITATIONS OR EXCLUSIONS MAY NOT APPLY TO YOU.

4. U.S. Government Restricted Rights

The Software is licensed subject to RESTRICTED RIGHTS. Use, duplication or disclosure by the Government or any person or entity acting on its behalf is subject to restrictions as set forth in subdivision (c)(1)(ii) of the Rights in Technical Data and Computer Software Clause at DFARS (48 CFR 252.227-7013) for DoD contracts, in paragraphs (c)(1) and (2) of the Commercial Computer Software-Restricted Rights clause in the FAR (48 FR 52.227-19) for civilian agencies, or in the case of NASA, in Clause 18-52.227-86(d) of the NASA Supplement to the FAR, or in other comparable agency clauses. The contractor/manufacturer is Ziff-Davis Benchmark Operation, 1001 Aviation Parkway, Suite 400, Morrisville, North Carolina 27560.

5. General Provisions

Nothing in this Agreement constitutes a waiver of Ziff-Davis' rights under U.S. copyright laws or any other Federal, state, local or foreign law. You are responsible for installation, management, and operation of the Software. However, if you have questions or problems regarding the Software or Media, you can write to Ziff-Davis Benchmark Operation, 1001 Aviation Parkway, Suite 400, Morrisville, North Carolina 27560 Attn: Distribution Coordinator. This Agreement constitutes the entire agreement between the parties with respect to the Software and the Media and supersedes all prior or contemporaneous understandings or agreements, written or oral, regarding such subject matter. This Agreement shall be governed by and construed in accordance with the laws of the State of New York, applicable to agreements made and performed in New York. If any provision of this Agreement shall be held by a court of competent jurisdiction to be illegal, invalid or unenforceable, the remaining provisions shall remain in full force and effect and the unenforceable provision shall be reformed without further action by the parties and only to the extent necessary to make such provision valid and enforceable and to achieve the like economic intent and effect of such provision.

GETTING THE FILES YOU NEED

The \WB11\ directory contains the following subdirectories:

\CLIENT
\CONTROLR

The CLIENT subdirectory contains all the files you need to install WebBench on each client. To install your client files, you'll need to run the Windows-based SETUP.EXE program on each client.

The CONTROLR subdirectory contains all the files you need to install on the controller. To install your controller files, you'll need to run the Windows-based SETUP.EXE program on the controller.

You can also download WebBench from ZDNet(TM)/World Wide Web Edition or ZDNet/CompuServe® Edition. You'll need to download two files. The self-extracting file WBCONTLR.EXE contains the controller SETUP.EXE program and the self-extracting file WBCLIENT.EXE contains the client SETUP.EXE program. You'll need to expand these files to get the setup programs.

INSTALLING WEBBENCH(TM) VERSION 1.1

To run the WebBench test suites, you must have a Web server, a PC running Microsoft® Windows NT® 3.51 or 4.0 with an x86-compatible processor (this is the controller), at least one client running either Windows® 95 or Windows NT 3.51 or 4.0, and a network that allows these machines to communicate. WebBench provides both static test suites and dynamic test suites. You can run the static test suites against any server. The dynamic test suites are platform dependent.

For the dynamic test suites, WebBench provides a CGI application for the following server platforms: Apple’s Mac® OS 7.53, Digital(TM) UNIX® with Alpha(TM) processors, IBM’s OS/2® Warp Server, Linux with x86 processors, Microsoft’s Windows NT® 3.51/4.0 with x86 processors, Silicon Graphics’ IRIX(TM), and Sun’s Solaris(TM) 2.5 on SPARC®. In addition, we re-wrote the CGI application as a Windows NT Microsoft Internet Server API dynamically linked library (ISAPI.DLL), a Windows NT Netscape Server API dynamically loadable library (NSAPI.DLL), an IntranetWare Local-CGI Netware Loadable Module (L-CGI NLM), and a Solaris on SPARC, Digital Alpha, and IRIX NSAPI shared object (.so).

We recommend that you read the installation instructions in The WebBench(TM) Tester's Handbook before you install WebBench. The installation instructions contain information on WebBench's minimum hardware and software requirements and how to set up your machines for testing. The following is a quick overview of the WebBench installation process.

You install WebBench by running the controller SETUP.EXE program on the controller and the client SETUP.EXE program on each client. As part of the client installation, WebBench asks you for a unique ID number for each client. We recommend you number the clients sequentially, starting at 1. (WebBench's standard test suites use client IDs 1 through 60.)

The SETUP.EXE program creates an installation directory on each PC and places the appropriate files in it. To complete the client installation, you'll need to edit each client's HOSTS file to include the IP address for each server you want to test against and the IP address for the controller. While you can use any name you choose for the server (the standard WebBench test suites use the default name server), you must use the name controller for the Windows NT workstation you're using as the controller.

After you install the controller and client programs, you'll need to place the WebBench test files on the server. These files are in compressed format in the WORKLOAD directory in the WebBench installation directory on the controller.

If you plan to run the dynamic test suites, you’ll need to be sure to get the correct compressed file for your server operating system. The following list shows you which file to get for which operating system.

Move that file to your server and expand it. You'll then have the directory WB10TREE and the WebBench dynamic executable(s). Move WB10TREE to the HTML document root on the Web server (this is where the Web server looks for its HTML files). If you're running WebBench on one of the server platforms that it supports for its dynamic tests, you'll need to move the dynamic executable(s) to your Web server's CGI-BIN directory. This is the directory where the Web server looks for its CGI and other dynamic scripts. If this directory has a different name on your server, you'll need to create the alias CGI-BIN for that directory. Make sure CGI-BIN and the directory it is in have execute permission. Depending on which dynamic executable you want to use, you may also have to do some additional setup work. See The WebBench(TM) Tester's Handbook for information on installing the dynamic executables.

NOTE: If your Web server can’t expand one of the above compressed files, you can expand the NT_x86.EXE file on your controller and move the directory containing the test files, WB10TREE, to the correct directory on your server. The test files take up about 3 MB, so you should expand NT_x86.EXE on a network directory from which you can simply move the WB10TREE directory tree to your server. If you need to put WB10TREE on a diskette, you’ll have to compress WB10TREE using the compression tool of your choice. This way you'll be able to run WebBench's static test suites against any Web server platform.

WHERE TO FIND DETAILED INSTALLATION INSTRUCTIONS

We supply on-line copies of The WebBench(TM) Tester's Handbook. On the Ziff-Davis Server Benchmarks CD-ROM, this manual is in the directory \DOCS\WB11. The handbook uses the base name WBHANDBK.* . We supply this manual in three formats: Microsoft Word 7.0 for Windows® (.DOC), Rich Text Format (.RTF), and ASCII format (.TXT). This way, even if you don't have Word for Windows, you should still be able to print a copy of these manuals. (WebBench also places this manual in the \DOCS subdirectory on the controller when you install WebBench.)

If you've downloaded the WebBench controller and client files from ZDNe/World Wide Web Edition or ZDNet/CompuServe Edition, you can also download a copy of this manual in Microsoft Word format. (ZDNet is a Ziff-Davis on-line service.)

In addition, you can start WebBench on the controller and view the information in The WebBench(TM) Tester's Handbook from the on-line help.

WHAT'S NEW ABOUT WEBBENCH 1.1

WebBench 1.1 contains several enhancements over WebBench 1.0. These include:

For more details about these features, see the The WebBench(TM) Tester's Handbook.

IMPORTANT WEBBENCH NOTES

The following list provides information you need to know BEFORE you run WebBench.

  1. You must read and agree to the license information at the beginning of this file before you run WebBench. The same information appears on the controller's screen when you first run WebBench and in the front of the WebBench manual. If you do not agree to the licensing information, cease use of the benchmark and delete all copies of WebBench in your possession or under your control. If you received WebBench on a CD-ROM, return the CD-ROM and all accompanying materials (including any documentation) to the Ziff-Davis Benchmark Operation (ZDBOp) at the following address:

    Ziff-Davis Benchmark Operation
    1001 Aviation Parkway, Suite 400
    Morrisville, NC 27560

  2. We've noticed a problem on Windows 95 clients when we run them continuously with multiple threads for at least an hour and a half. At that time the PC begins doing a lot of disk activity that is usually associated with page swaps. If you want to run test mixes using multiple threads on each client, we recommend you use Windows NT clients.

  3. In our testing we've discovered that, if you run WebBench on a Web server that has logging enabled, over time the WebBench scores may go down and the test results may not be repeatable. This drop in performance and repeatability occurs as the size of the log file increases. We recommend that you disable logging on your Web server software before you run WebBench.

  4. Some Web servers may not allow you to create a CGI-BIN alias to the CGI root directory on the Web server (this is the directory where WebBench looks for dynamic executables). In this case, you will need to edit the workload files by hand so that they point to right URL for the dynamic executable. (See Chapter 6 in The WebBench(TM) Tester's Handbook for information on how to modify the workload files.)

  5. You can compare your WebBench 1.1 results with your WebBench 1.0 results for all common test suites except the Windows NT CGI test suites. We changed the Windows NT CGI executable that the test suites ZD_NT_SIMPLE_CGI20_V11.TST and ZD_NT_SIMPLE_CGI20_KEEPALIVE_V11.TST use because of a problem with some compiler libraries. As a result of the change, this executable no longer generates numerous errors when you run it on multiple processor configurations.

  6. The WebBench(TM) Tester's Handbook provides the minimum information you need to install and run WebBench, view your results, and create test suites. If you need additional information, look in the WebBench help file. Because WebBench and ServerBench share many of the same dialog boxes, you can also use the Understanding and Using ServerBench® 4.0 reference manual to get more details on some WebBench tasks. You'll find this manual in the ServerBench documentation directories on the Server Benchmarks CD-ROM(\DOCS\SB40\<ServerBench- platform>). You can also download this manual from the same ZDNet locations where you can download the benchmark programs (see the section "Technical Support" for the addresses to ZDNet/World Wide Web and ZDNet/CompuServe.)

DOCUMENTATION

WebBench 1.1 comes with the following documentation:

README.WRI This file contains the WebBench 1.1 license agreement, information on where to find detailed installation instructions, and important notes about WebBench 1.1. You should read this information before you run WebBench. (This is the file you are reading now.)

WBHANDBK.*** This is The WebBench(TM) Tester's Handbook. This book focuses on the basic steps you need to perform to install WebBench, run its standard test suites, and view your results. We include three formats of this manual in the \DOCS\WB11 directory on the CD-ROM. The .DOC version is in Word for Windows format; the .RTF version is in Rich Text Format; and the .TXT version is a plain ASCII text file.

WEBBENCH.HLP This is WebBench's on-line HELP information, which is accessible from the controller when WebBench is running.

TECHNICAL SUPPORT

If you have problems installing or running this release of WebBench, please contact the Ziff-Davis Benchmark Operation (ZDBOp).

===============================================================

 

LICENSE AGREEMENT FOR WEBBENCH™ VERSION 1.1

READ THIS AGREEMENT CAREFULLY BEFORE USING THE SOFTWARE EMBODIED IN THE WEBBENCHä 1.1 CD-ROM, OR, IF PRELOADED ON YOUR HARD DRIVE, DOWNLOADED OR IF PROVIDED AS PART OF A COLLECTION, THE PRELOADED, DOWNLOADED OR COLLECTED FILE(S) (the "Media"). Embodied in the WebBenchä 1.1 Media is the WebBenchä version 1.1 computer programs and related documentation (the "Software"). Ziff-Davis Inc., having a place of business at One Park Avenue, New York, New York 10016 ("Ziff-Davis"), is the licensor under this Agreement and you are the licensee. By using the Software, in whole or in part, you agree to be bound by the terms of this Agreement. If you do not agree to the terms of this Agreement, promptly return the Software to Ziff-Davis Benchmark Operation at 1001 Aviation Parkway, Suite 400, Morrisville, North Carolina 27560 (or, if downloaded or preloaded on your hard drive, delete the Software, or if provided as part of a collection, cease use of the Software). Title to the Software and all copyrights, trade secrets and other proprietary rights therein are owned by Ziff-Davis. All rights therein, except those expressly granted to you in this Agreement, are reserved by Ziff-Davis.

1. Limited License

This Agreement grants you only limited rights to use the Software. Ziff-Davis grants you a non-exclusive, non-transferable license to use the Software on a Web server networked with multiple PC computers for the sole purpose of conducting benchmark tests to measure the performance of computer hardware and Web server software. You have the right to make a single copy of the Software for archival purposes and the right to transfer a copy of the Software across an internal local area network only to the PC computers attached to such network; provided, however, that all such copies are considered Software hereunder, that all uses of such copies are governed by the terms and conditions of this Agreement and that you shall be responsible for all uses of such copies in violation of the terms and conditions of this Agreement. Ziff-Davis hereby grants you the right to publish, except in any country where a third party claims during the term of this license that such publication infringes that party's proprietary rights, benchmark test results obtained by you from your use of the Software, provided that with the publication of each such result you:

A. Identify Ziff-Davis, the name and version number of the benchmark Software used and the name of the test suite used in the test (i.e., Ziff-Davis' WebBenchä version 1.1 with the standard test suite ZD_STATIC_V11.TST);

B. Identify for the Web server the exact name and version number of the Web server software package, any tunable parameters (e.g., logging) you set (unless you used the default values), the exact name, number of processors, processor speed (including clock speed) and type(s), amount of processor level 2 (L2) cache, amount of RAM, type of I/O bus, number and type of hard disk controller(s), size of hardware hard disk cache, if any, number and type of hard disks, hard disk size, type of disk organization (e.g., RAID 5), driver version of the disk controller(s), number and type of network controllers, driver version of network controller, network operating system name and version, and any relevant modifications to the default network operating system parameters (e.g., ABC Inc. Web X with logging disabled running on WXY Corp. Model 466 with a dual 166-MHz Pentium® processor, 512KB L2 cache, 128MB of RAM, bridged PCI/EISA I/O bus, WXY Integrated Drive Array disk controller, 4 ABC 1.2GB disk drives, RAID 0 hardware striping, a xxxxxxxx.dsk 12,621 4/29/96 disk driver, WXY 100BaseT network controller, a WXY 100BaseT driver version X.X, Microsoft® TCP/IP, Microsoft Windows® Sockets Version 1.1, Windows NT® Server 3.51);

C. Identify for the client testbed the network type, the number of clients, the client operating system version and service pack version, if any, (e.g., Microsoft Windows 95 or Windows NT 3.51 with Service Pack 1), the client CPU type and speed in percentages, amount of RAM, the number and type of hubs/concentrators, the number of clients per segment, client network software name and version (drivers, protocols, redirector; e.g., Microsoft TCP/IP, Microsoft Windows Sockets Version 1.1, NE2000.386), client network card, size of the client network cache, if any (e.g., 100BaseT with 32 clients, 50% of the clients are 486/25-based PCs running Windows 95 with Service Pack 1, 50% of the clients are 100Mhz Pentium PCs running Windows NT 3.51 with Service Pack 4, 16MB RAM, 2 XYZ 100BaseT Ethernet hubs, 16 clients per segment, 50% of the clients have ABC 100BaseT network cards and 50% of the clients have GHI 100BaseT network cards, Microsoft Client for Microsoft Networks, Microsoft TCP/IP, Microsoft Windows Sockets Version 1.1, xxxxxx.SYS LAN driver);

D. Identify the controller operating system version, network software and version, and network card (e.g., Windows NT, Microsoft Client for Microsoft Networks, Microsoft TCP/IP, Microsoft Windows Sockets Version 1.1, xxxxxx.SYS LAN driver, ABC 100BaseT network card);

E. Identify any other special conditions used to achieve the result;

F. State that all products used in the test were shipping versions available to the general public;

G. State that the test was performed without independent verification by Ziff-Davis and that Ziff-Davis makes no representations or warranties as to the results of the test; and

H. Follow proper trademark usage and acknowledge Ziff-Davis' trademark rights (e.g., "[ ] achieved a WebBenchä overall score of X requests per second and a throughput score of X bytes per second. WebBenchä is a trademark of Ziff-Davis Inc. in the U.S. and other countries.").

This Agreement and your rights hereunder shall automatically terminate if you fail to comply with any provision of this Agreement. Upon such termination, you shall cease all use of the Software, cease the transfer of any copies of the Software and cease the publication of benchmark test results obtained by you from use of the Software. Further, you shall delete the Software and destroy all tangible copies of the Software and other materials related to the Software in your possession or under your control, or, if downloaded or preloaded on your hard drive or if provided as part of a collection, you shall cease use of and destroy any and all copies of the Software in your possession or under your control.

2. Additional Restrictions

A. You shall not (and shall not permit other persons or entities to) rent, lease, sell, sublicense, assign, or otherwise transfer the Software or this Agreement. Any attempt to do so shall be void and of no effect.

B. You shall not (and shall not permit other persons or entities to) reverse engineer, decompile, disassemble, merge, modify, include in other software or translate the Software, or use the Software for any commercial purposes, except for the publication of test results, as provided above.

C. You shall not (and shall not permit other persons or entities to) remove or obscure Ziff-Davis' copyright, trademark or other proprietary notices or legends from any of the materials contained in this package or downloaded.

D. You acknowledge that the Software contains Ziff-Davis' trade secret information and you shall not disclose or disseminate such information other than as provided herein.

3. Disclaimer of Warranty; Limitation of Liability

THE SOFTWARE AND THE MEDIA ARE PROVIDED "AS IS" WITHOUT WARRANTY OF ANY KIND, EITHER EXPRESS OR IMPLIED, INCLUDING, WITHOUT LIMITATION, ANY WARRANTY OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE.

THE ENTIRE RISK AS TO THE RESULTS AND PERFORMANCE OF THE SOFTWARE AND THE MEDIA IS ASSUMED BY YOU, AND ZIFF-DAVIS AND ITS AUTHORIZED DISTRIBUTORS ASSUME NO RESPONSIBILITY FOR THE ACCURACY OR APPLICATION OF OR ERRORS OR OMISSIONS IN THE SOFTWARE OR THE MEDIA. IN NO EVENT SHALL ZIFF-DAVIS OR ITS AUTHORIZED DISTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, SPECIAL, INCIDENTAL OR CONSEQUENTIAL DAMAGES ARISING OUT OF THE USE OR INABILITY TO USE THE SOFTWARE OR THE MEDIA, EVEN IF ZIFF-DAVIS OR ITS AUTHORIZED DISTRIBUTORS HAVE BEEN ADVISED OF THE LIKELIHOOD OF SUCH DAMAGES OCCURRING. ZIFF-DAVIS AND ITS AUTHORIZED DISTRIBUTORS SHALL NOT BE LIABLE FOR ANY LOSS, DAMAGES OR COSTS, ARISING OUT OF, BUT NOT LIMITED TO, LOST PROFITS OR REVENUE, LOSS OF USE OF THE SOFTWARE OR THE MEDIA, LOSS OF DATA OR EQUIPMENT, THE COSTS OF RECOVERING SOFTWARE, THE MEDIA, DATA OR EQUIPMENT, THE COST OF SUBSTITUTE SOFTWARE OR THE MEDIA, DATA OR EQUIPMENT OR CLAIMS BY THIRD PARTIES, OR OTHER SIMILAR COSTS.

SOME STATES DO NOT ALLOW EXCLUSION OR LIMITATION OF IMPLIED WARRANTIES OR LIMITATION OF LIABILITY FOR INCIDENTAL OR CONSEQUENTIAL DAMAGES; SO THE ABOVE LIMITATIONS OR EXCLUSIONS MAY NOT APPLY TO YOU.

4. U.S. Government Restricted Rights

The Software is licensed subject to RESTRICTED RIGHTS. Use, duplication or disclosure by the Government or any person or entity acting on its behalf is subject to restrictions as set forth in subdivision (c)(1)(ii) of the Rights in Technical Data and Computer Software Clause at DFARS (48 CFR 252.227-7013) for DoD contracts, in paragraphs (c)(1) and (2) of the Commercial Computer Software-Restricted Rights clause in the FAR (48 CFR 52.227-19) for civilian agencies, or in the case of NASA, in Clause 18-52.227-86(d) of the NASA Supplement to the FAR, or in other comparable agency clauses. The contractor/manufacturer is Ziff-Davis Benchmark Operation, 1001 Aviation Parkway, Suite 400, Morrisville, North Carolina 27560.

5. General Provisions

Nothing in this Agreement constitutes a waiver of Ziff-Davis' rights under U.S. copyright laws or any other Federal, state, local or foreign law. You are responsible for installation, management, and operation of the Software. However, if you have questions or problems regarding the Software or Media, you can write to Ziff-Davis Benchmark Operation, 1001 Aviation Parkway, Suite 400, Morrisville, North Carolina 27560 Attn: Distribution Coordinator. This Agreement constitutes the entire agreement between the parties with respect to the Software and the Media and supersedes all prior or contemporaneous understandings or agreements, written or oral, regarding such subject matter. This Agreement shall be governed by and construed in accordance with the laws of the State of New York, applicable to agreements made and performed in New York. If any provision of this Agreement shall be held by a court of competent jurisdiction to be illegal, invalid or unenforceable, the remaining provisions shall remain in full force and effect and the unenforceable provision shall be reformed without further action by the parties and only to the extent necessary to make such provision valid and enforceable and to achieve the like economic intent and effect of such provision.