WebLOAD Console Guide
WebLOAD Console Guide
Version 8.0
RadView Software
The software supplied with this document is the property of RadView Software and is furnished under a
licensing agreement. Neither the software nor this document may be copied or transferred by any means,
electronic or mechanical, except as provided in the licensing agreement. The information in this document is
subject to change without prior notice and does not represent a commitment by RadView Software or its
representatives.
WebLOAD User's Guide
© Copyright 2007 by RadView Software. All rights reserved.
January, 2007, RadView Publication Number WL-0305-UG72
WebLOAD, TestTalk, Authoring Tools, ADL, AppletLoad, WebFT, and WebExam, are trademarks or
registered trademarks of RadView Software IBM, and OS/2 are trademarks of International Business Machines
Corporation. Microsoft Windows, Microsoft Windows 95, Microsoft Windows NT, Microsoft Word for
Windows, Microsoft Internet Explorer, Microsoft Excel for Windows, Microsoft Access for Windows and
Microsoft Access Runtime are trademarks or registered trademarks of Microsoft Corporation. SPIDERSESSION
is a trademark of NetDynamics. UNIX is a registered trademark of AT&T Bell Laboratories. Solaris, Java and
Java-based marks are registered trademarks of Sun Microsystems, Inc. HP-UX is a registered trademark of
Hewlett-Packard. SPARC is a registered trademark of SPARC International, Inc. Netscape Navigator and
LiveConnect are registered trademarks of Netscape Communications Corporation. Any other trademark name
appearing in this book is used for editorial purposes only and to the benefit of the trademark owner with no
intention of infringing upon that trademark.
2 RadView Software
Table of Contents
1. Introduction...................................................................................15
WebLOAD Documentation .................................................. 15
Typographical Conventions ................................................ 16
Where to Get More Information........................................... 17
Online Help ............................................................................ 17
Technical Support .................................................................. 17
Technical Support Web Site................................................... 18
2. Introducing WebLOAD .................................................................19
3. Getting Started..............................................................................21
Welcome to WebLOAD ........................................................ 21
How Does WebLOAD Work? ............................................... 22
WebLOAD Architecture .......................................................... 23
How are Tests Defined?......................................................... 25
What is the Product of Your WebLOAD Testing?................... 26
WebLOAD Offers Two Client Type Modes............................. 26
WebLOAD Features.............................................................. 28
4. WebLOAD Quick Start..................................................................35
Quick Start Steps ................................................................. 35
Creating an Agenda ............................................................. 36
Configuring a Load Template using the WebLOAD
Wizard ................................................................................ 40
Opening the WebLOAD Wizard ............................................. 40
4 RadView Software
Status Bar .............................................................................. 81
The Session Tree................................................................... 81
The Results Window .............................................................. 89
Event Viewer .......................................................................... 94
Log Window ........................................................................... 94
The WebLOAD Console Shortcut Screen .............................. 95
Edit Mode and Run Modes of Operation ............................ 96
The Console in Edit Mode ...................................................... 96
The Console in Run Mode...................................................... 97
The Console Menu Options................................................. 98
File Menu ............................................................................... 98
Session Control Menu ............................................................ 99
Reports Menu....................................................................... 100
View Menu ........................................................................... 101
Tools Menu .......................................................................... 104
Window Menu ...................................................................... 105
Help Menu............................................................................ 105
The Console Toolbar.......................................................... 105
Toolbar Buttons.................................................................... 106
Extended Toolbar Segments................................................ 107
7. Creating Load Templates...........................................................111
WebLOAD Options for Creating Load Templates: .......... 111
8. Planning a Test Session ............................................................113
Load Session Workflow..................................................... 113
Forming a Test Plan ........................................................... 115
Planning a Test Program...................................................... 116
Creating the Test Agendas................................................... 116
Examining Your Test Resources.......................................... 117
Starting WebLOAD ............................................................. 117
9. Connecting to WebRM ...............................................................121
Getting Started ................................................................... 121
About WebRM...................................................................... 122
How Does WebRM Work ..................................................... 123
Working with WebRM......................................................... 123
Requesting Resources from WebRM ................................... 123
Updating the Number of Requested Resources................... 124
WebLOAD User's Guide 5
10. Managing Load Templates and Load Sessions.................... 125
Managing Load Templates ................................................ 125
Creating a New Load Template............................................ 126
Saving a Load Template ...................................................... 127
Opening a Saved Load Template......................................... 127
Saving Additional Information with the Load Template......... 127
Managing Load Sessions .................................................. 129
Saving a Load Session......................................................... 129
Saving Additional Information for the Load Session ............. 130
11. Creating Load Templates with the WebLOAD Wizard.......... 133
Using the WebLOAD Wizard to create Load
Templates ........................................................................ 133
The WebLOAD Wizard Workflow ...................................... 134
Opening the WebLOAD Wizard ......................................... 135
Selecting an Agenda or Mix .............................................. 136
Selecting an Agenda .......................................................... 137
Selecting a Mix ................................................................... 138
Creating a Mix ...................................................................... 139
Selecting Host Computers ................................................ 141
Adding Host Computers ....................................................... 142
Designating a Host Computer as a Load Machine ............... 142
Designating all Host Computers as Load Machines ............. 143
Designating a Host Computer as a Probing Client
Machine........................................................................... 143
Replacing a Load Machine or Probing Client Machine......... 143
Scheduling the Load Session ........................................... 144
Distributing the Load Equally................................................ 145
Manually Scheduling the Load to Generate on Each
Load Machine.................................................................. 148
Scheduling Models ............................................................... 150
Completing the WebLOAD Wizard.................................... 158
After Completing the WebLOAD Wizard .......................... 159
Editing the Load Template ................................................... 159
Viewing the WebLOAD Wizard Default Report .................... 159
Modifying the Host Selection ................................................ 160
Modifying the Load Schedule ............................................... 160
6 RadView Software
12. Creating Load Templates with Cruise Control ......................161
Using Cruise Control ......................................................... 161
How does Cruise Control Work? ...................................... 162
The Cruise Control Wizard Workflow ............................... 163
Opening the Cruise Control Wizard.................................. 163
Selecting an Agenda or Mix .............................................. 165
Running Single Agendas................................................... 166
Running a Mix of Agendas ................................................ 167
Defining a Mix ...................................................................... 169
Selecting Host Computers ................................................ 172
Adding Host Computers to the Host Computers List............ 173
Designating a Host Computer as a Load Machine ............... 173
Designating all Host Computers as Load Machines ............. 174
Designating a Host Computer as a Probing Client
Machine........................................................................... 174
Deleting a Host Computer .................................................... 174
Defining Performance Goals ............................................. 175
About the Goal Definition Dialog Box ................................... 175
Adding Performance Goals .................................................. 177
Deleting a Goal .................................................................... 180
Setting and Using Defaults................................................... 180
Defining the Acceleration Rate ......................................... 182
Defining WebLOAD's Action Once the Goals are
Reached ........................................................................... 183
Finishing the Cruise Control Wizard ................................ 184
Viewing the Cruise Control Default Report...................... 186
13. Configuring a Load Template Manually .................................189
The Manual Test Configuration Workflow........................ 190
Starting a New Template Manually ................................... 190
Selecting an Agenda .......................................................... 191
Creating Load Templates with One or More Single
Agendas .......................................................................... 191
Creating Load Templates with an Agenda Mix..................... 192
Selecting Hosts .................................................................. 193
Opening the Host Selection Dialog Box ............................... 193
Defining Host Computers ..................................................... 194
WebLOAD User's Guide 7
Defining Load Machines....................................................... 194
Defining Probing Client Machines ........................................ 195
Scheduling the Test Session............................................. 195
Creating a Load Schedule Manually..................................... 196
Creating a Load Schedule with the Load Profiler ................. 197
Completing Your Manual Load Template
Configuration................................................................... 198
14. Defining Test Session Options .............................................. 201
Session Options ................................................................. 201
15. Setting Global Options............................................................ 203
How to Set Global Options ................................................ 203
The General Tab .................................................................. 205
The Export Tab..................................................................... 208
The File Locations Tab......................................................... 211
The Data Drilling Tab ........................................................... 212
The Advanced Tab ............................................................... 213
The Defect Tracking Tab...................................................... 215
The Notification Manager .................................................. 215
16. Setting Agenda Options.......................................................... 221
About Agenda Options ...................................................... 222
Modifying Agenda Options................................................ 224
Setting the SSL Options .................................................... 225
Client Type Settings........................................................... 225
Selecting HTTP Client Type Mode ....................................... 227
Selecting Browser Client Type Mode ................................... 227
Setting Sleep Time Playback Settings.............................. 227
Setting the Browser Cache................................................ 229
Configuring Browser Parameter Settings ........................ 231
Setting the Browser Type and User-Agent ........................... 232
Setting Load Generator and Probing Client Thread
Properties ........................................................................ 235
Setting Redirection Limits..................................................... 235
Simulating DNS Cache......................................................... 236
Simulating SSL Cache ......................................................... 236
Enabling Persistent Connection (Keep-Alive)....................... 236
8 RadView Software
Setting the HTTP Parameters............................................ 237
Setting Authentication ....................................................... 240
Connection Speed Settings............................................... 242
Diagnostic Settings............................................................ 243
Enabling Syntax Checking ................................................... 244
Enabling Dynamic Object Recognition Diagnostic
Messages ........................................................................ 245
Log Report Diagnostics ........................................................ 246
Enabling RadView Support Diagnostic................................. 247
Enabling Save CSJS and CSJS Diagnostic Messages ........ 247
Functional Testing Options............................................... 247
Setting Automatic Data Collection........................................ 248
Setting Page Time Verifications ........................................... 249
Inserting a Text Verification Test.......................................... 250
Inserting Test Selections ...................................................... 251
Pass / Fail Definition Settings ........................................... 252
Configuring Reset Mechanism Settings........................... 254
Configuring Emulation Settings ....................................... 256
Web Services Settings....................................................... 258
17. Functional Verification Testing ...............................................259
The Functional Testing Tab............................................... 260
Opening the Functional Testing Tab .................................... 260
Selecting the Tests to Perform ............................................. 262
Functional Verification Test List ........................................... 264
User-Defined Transactions and Verification.................... 264
Including a User-Defined Transaction in your Agenda ......... 265
User-Defined Transactions and Verification Syntax ............. 265
User-Defined Transaction Example ..................................... 265
Verification Function Return Values................................. 266
Reporting Events ............................................................... 267
Tracking Reasons for Transaction Failure....................... 267
Viewing Verification Test Results ..................................... 269
Viewing Error Information in the Log Window ...................... 269
Viewing Error Information in the Event Viewer ..................... 270
Printing the Contents of the Log Window ............................. 271
Saving the Contents of the Log Window .............................. 271
10 RadView Software
Opening Reports ................................................................ 295
Opening an Existing Report .............................................. 298
Editing a Report ................................................................. 298
Renaming a Report ............................................................ 299
Deleting a Report ............................................................... 300
Configuring an Integrated Report ..................................... 300
Selecting Measurements for the Integrated Report.............. 302
Changing the Reports Parameters....................................... 302
Importing Report Parameters from a Saved Load
Template ......................................................................... 303
Regression Testing ............................................................ 305
Configuring an Integrated Report to Include Data from
Previously Saved Load Sessions .................................... 305
Removing a Previously Saved Load Sessions from the
Integrated Report Configuration ...................................... 306
Integrated Report Chart View............................................ 306
Changing the Scale of a Report for a Selected
Measurement .................................................................. 308
Viewing the True Value vs. Relative Value of a
Measurement .................................................................. 308
Integrated Report Toolbar in Chart View .............................. 309
Integrated Report in Spread(sheet) View ......................... 310
22. WebLOAD Statistics Reports ..................................................313
Statistical Measurements and Their Values..................... 313
Changing the Length of the Reporting Interval..................... 314
Current Value Computation.................................................. 315
Per Time-Unit Statistics Definitions ...................................... 315
Timer Definitions .................................................................. 317
Counter Definitions............................................................... 320
Statistics for Virtual Clients and Probing Clients .................. 322
Summary Reports .............................................................. 323
Using the Statistics Report ................................................... 324
Statistics Report-Summary by Agenda................................. 325
Statistics Report-Summary by Total ..................................... 327
Statistics Report-Summary by Measurement ....................... 329
Statistics Report-History by Measurement and Agenda ....... 331
Statistics Report-History by Measurement and Total ........... 332
WebLOAD User's Guide 11
23. Data Drilling and Table Compare Reports............................. 335
Data Drilling Reports.......................................................... 335
Opening a Data Drilling Report............................................. 336
Viewing the Transaction Grid ............................................... 337
Viewing the Transaction Reason Failure Grid ...................... 337
Viewing the Instance Grid .................................................... 338
Viewing the Parent Transaction Instance Tree..................... 339
Viewing the Child Transaction Instance Tree ....................... 340
Table Compare Reports..................................................... 341
Opening the Table Compare Viewer .................................... 343
24. Exporting Performance Reports ............................................ 345
Report Export Options....................................................... 345
Exporting Selective WebLOAD Report Data .................... 347
Exporting Performance Reports to Excel ............................. 348
Exporting Performance Reports to a Tab File ...................... 350
Exporting Performance Reports to HTML ............................ 352
Exporting All Statistical Data ............................................ 353
Exporting All Statistics to Excel ............................................ 354
Exporting All Statistics to a Tab File ..................................... 356
Exporting All Statistics to HTML ........................................... 357
Customizing the Export File .............................................. 357
Creating a Summary Report .............................................. 359
Customizing the Report Fields ............................................. 360
Report Destination................................................................ 361
Integrated Report Definition ................................................. 362
25. Performance Measurements Manager................................... 365
Using the Performance Measurements Manager ............ 365
About the Performance Measurements Manager ............ 366
Opening the Performance Measurements Manager........ 366
The Performance Measurements Manager Main
Window............................................................................ 367
The PMM Configuration Tree Structure................................ 370
Navigating the PMM Configuration Tree .............................. 372
Using PMM Configuration Files ........................................ 373
Creating a New Configuration File........................................ 373
Opening / Importing Saved PMM Configuration Files........... 373
12 RadView Software
Assigning a PMM Configuration to a WebLOAD
Template ......................................................................... 374
Saving and Exporting PMM Configurations.......................... 374
Configuring the Data Sources, Hosts, and
Measurements to Monitor .............................................. 375
Adding Data Sources ........................................................... 376
Adding Measurements to a Configured Data Source ........... 383
Adding Data Sources to a Configured Host.......................... 384
Deleting a Data Source ........................................................ 384
Deleting a Monitored Host.................................................... 385
Deleting a Measurement ...................................................... 385
Application Server Monitors ............................................. 385
WebLogic Monitor Configuration Tasks................................ 385
IBM WebSphere Configuration Tasks .................................. 387
Database Monitors ............................................................. 387
Oracle Monitor Configuration Tasks on Windows ................ 387
Oracle Monitor Configuration Tasks on UNIX ...................... 389
Web Server Monitors ......................................................... 391
Sun One (iPlanet) Configuration Tasks ................................ 391
Monitoring Generic SNMP Data Sources ......................... 392
Understanding MIB Files ...................................................... 392
The WebLOAD UserMibFilter.ini File Structure.................... 393
Adding SNMP Agents........................................................... 394
A. WebLOAD Error Messages .......................................................397
Message Severity ............................................................... 397
WebLOAD Error Messages................................................ 398
B. Reference Materials ...................................................................403
WebLOAD Files .................................................................. 403
WebLOAD Hot Key Commands ........................................ 404
Index ................................................................................................407
1. Introduction
This section provides a brief introduction to TestView technical support, including both
documentation and online support.
In This Chapter
WebLOAD Documentation ........................................................ 15
Typographical Conventions........................................................ 16
Where to Get More Information ................................................. 17
WebLOAD Documentation
WebLOAD is supplied with the following documentation:
Typographical Conventions
Before you start using this guide, it is important to understand the terms and typographical
conventions used in the documentation.
For more information on specialized terms used in the documentation, see the Glossary at the
end of this document.
The following kinds of formatting in the text identify special information.
16 RadView Software
Introduction
Online Help
TestView provides a comprehensive on-line help system with step-by-step instructions for
common tasks.
You can press the F1 key on any open dialog box for an explanation of the options or select Help
| Contents to open the on-line help contents and index.
Technical Support
For technical support in your use of this product, contact:
North American Headquarters
e-mail: [email protected]
Phone: 1-888-RadView (1-888-723-8439) (Toll Free)
781-238-1111
Fax: 781-238-8875
International Headquarters
e-mail: [email protected]
Phone: +972-3-915-7060
Fax: +972-3-915-7683
Note: We encourage you to use e-mail for faster and better service.
When contacting technical support please include in your message the full name of the product,
as well as the version and build number.
18 RadView Software
C H A P T E R 2
2. Introducing WebLOAD
The WebLOAD introductory material gives you an overall view of WebLOAD and consists of
three chapters:
Getting Started (on page 21) describes WebLOAD, how WebLOAD works, the WebLOAD
components and features.
WebLOAD Installation and Configuration (on page 51) provides installation and
configuration information.
WebLOAD Console Features (on page 77) describes the features available through the
Console toolbar and menus.
20 RadView Software
C H A P T E R 3
3. Getting Started
Welcome to WebLOAD, the premier performance, scalability and reliability testing solution for
Web applications - Internet and intranet.
WebLOAD accurately simulates Internet users behavior and models real-life demands on your
Web application to predict capacity requirements, report bottlenecks and weak links in your
application before deployment.
This chapter introduces WebLOAD, the WebLOAD test components and an overview of
WebLOAD's capabilities and features.
In This Chapter
Welcome to WebLOAD ............................................................. 21
How Does WebLOAD Work?.................................................... 22
WebLOAD Features ................................................................... 28
Welcome to WebLOAD
WebLOAD delivers the following for maximum testing performance, ease of use and value:
transaction, and per-instance level under defined load conditions. WebLOAD saves the test
results including data from the Load Machines (described in the paragraphs that follow) and the
hosting hardware's performance monitor. You can view all or part of the data in real time, or
after the test session is complete, in tabular format or in graphical format.
The WebLOAD test process includes:
1. Planning the test
2. Creating Agendas
3. Creating Load Templates
4. Running the test
5. Analyzing the test results
This guide describes steps 1, 3, 4 and 5 in the test creation process in detail. Information on
performing step 2, Creating Agendas is available in the WebFT User's Guide.
WebLOAD Architecture
The following diagram illustrates the configuration for a typical WebLOAD test.
The Console sets up, runs and controls a test session. At the Console, you can:
24 RadView Software
Getting Started
WebFT - a tool for recording your activity in a Web browser that automatically translates
that activity into a visual Agenda that can be used in WebLOAD for load testing. WebFT is
installed on your system with the Console installation. For more information, see the WebFT
User's Guide.
TestTalk - a network agent that facilitates communication between the Console and hosts,
either Load Machines or Probing Client Machines. TestTalk must be installed on both the
Console and the hosts. TestTalk runs in the background and does not interfere with any
other program running on your system.
WebLOAD REPORTER - enables you to generate a variety of reports based on performance
data and statistics gathered by WebLOAD. Use WebLOAD REPORTER to select a load
session and generate predefined or custom reports displaying the results of the test.
WebRM- distributes and circulates WebLOAD testing resources (Virtual Clients and Probing
Clients) amongst users on a “need to use” basis. WebRM is packaged with a maximum
number of Virtual Clients, Probing Clients and Connected Workstation ports, as defined by
the WebLOAD package. With WebRM, every WebLOAD Console can operate in
Standalone Workstation mode or Connected Workstation mode.
Standalone Workstation - an independent Console that runs WebLOAD and is limited to 20
Virtual Clients with one Load Machine running on the local machine. It is not connected to
WebRM.
Connected Workstation - a Console that runs WebLOAD, is connected to WebRM, and can
be connected to other Load Machines running on different machines.
The Connected Workstation sets up, runs, and controls the test. Using a Connected
Workstation, you can request resources (Virtual Clients and Probing Clients) from WebRM,
schedule tests and free the resources being used.
HTTP Client
The focus on user activities used by the Browser Client mode represents a high-level, conceptual
approach to test session design. This user-activity focus is not always appropriate for all testing
26 RadView Software
Getting Started
situations. Sometimes a tester may prefer to use a low-level, “nuts-and-bolts” approach that
focuses on specific internal implementation commands, such as HTTP transactions.
Agendas created for HTTP Client mode focus on HTTP transactions. HTTP Client Agendas are
created through the WebLOAD IDE. The WebLOAD IDE is an easy-to-use tool that facilitates
the writing of test Agendas for HTTP Client mode. The WebLOAD IDE records your actions as
you navigate through a Web application, automatically generating an Agenda that reflects your
actions in JavaScript, writing the GET and POST HTTP protocol commands that correspond to
your browser activities. You can then use these Agendas to test the performance of your Web
application during WebLOAD test sessions.
Working in HTTP Client mode provides users with the most efficient, portable, powerful work
mode. HTTP Client mode is portable to many operating systems, including all different flavors
of Windows, UNIX, Solaris, and so on. HTTP Client mode runs so efficiently that users can
achieve large benchmarks by simulating extremely heavy test loads of thousands of clients.
HTTP Client mode also provides the most comprehensive, detailed statistical analysis and test
reports.
Working in HTTP Client mode does require a certain level of programming sophistication, since
it can require hands-on programming of the JavaScript Agendas with HTTP protocol code, as
described in the TestView JavaScript Reference Manual. Since the Agendas for HTTP Client
mode are created through the WebLOAD IDE, they are not compatible or interchangeable with
Agendas created through WebFT, and can not be run on WebFT. HTTP Client mode Agendas
are designed to be run only on WebLOAD. HTTP Client mode is also not recommended for web
sites with complex User Interfaces or a heavy use of JavaScript. Test sessions that access such
web sites should be run in one of the other Client Type modes. WebLOAD provides these
different test modes to accommodate a wide range of testing requirements.
Browser Client
During test sessions run in Browser Client mode, Agendas that recreate user browser activities
are run through an emulation engine that simulates actual browser activities. WebLOAD offers
the option of running test session Agendas through an emulation engine to provide the full set of
complex functional verification tests and statistical analyses available through HTTP Client
mode.
Browser Client mode is the mid-range work mode for WebLOAD testing and provides a
powerful and sophisticated testing option. For example, Agendas running under Browser Client
mode can be configured to compensate for slow internet connections, or to filter out certain
URL sites, providing more control over the test session environment. Browser Client mode also
provides a simpler user interface than HTTP Client mode. Working in Browser Client mode
provides users with a work mode that is powerful enough for most work situations, while easy to
work with, with a user-friendly GUI that simplifies user input and minimizes any programming
requirements. Browser Client mode is portable to most operating systems, including both
Windows and UNIX systems. Browser Client mode also handles medium to high test loads of
100-300 clients, and provides clear, detailed statistical analysis and reports for a sophisticated set
of HTTP manipulations.
Browser Client mode Agendas can be customized to meet most common testing needs. Browser
Client mode is usually not recommended for web sites with complex User Interfaces or a heavy
use of JavaScript. Test sessions that access such web sites should be run in one of the other
Client Type modes. WebLOAD provides these different test modes to accommodate a wide
range of testing requirements.
WebLOAD Features
Using WebLOAD, you have access to all the features described in this section.
28 RadView Software
Getting Started
Feature Benefit(s)
HTTP 1.0 and 1.1 Comprehensive support for HTTP 1.0 and 1.1 including
cookies, proxies, SSL, client certificates, authentications,
persistent connections and chunked transfer coding.
SSL updates Support for the new SSL standard (TLS 1.0)
SSL Bit Limit Enables you to set a limit on the maximum number of bits
available to Virtual Client for SSL encryption.
WAP support WAP can now be selected as the browser emulation browser
type in the Console user interface. WAP sessions can be
recorded into scripts and WAP headers can be edited using
WebLOAD IDE for configuration and parameterization.
Various hand-sets can be emulated by defining custom user-
agents. Binary WAP (WAP1) can also be recorded and re-
played during a load session.
Internet Protocol Support Native support for FTP, SMTP, POP3, IMAP, NNTP, TCP,
UDP, and Telnet provides support for non-HTTP protocols.
COM support WebLOAD Java Script provides direct object access to any
component that has a COM wrapping and an Idispatch
interface.
Java from JavaScript WebLOAD supports full Java access from your JavaScript
Agendas. Full Java support means that your WebLOAD
Agendas not only test access time to an HTML page, they
can also invoke and run Java classes used by the Java
applications embedded within an HTML page.
ODBC Support Native support from WebLOAD Java Script for ODBC
commands to be integrated in an Agenda.
Throttle Control Vary the number of Virtual Clients on the fly to create
“what-if” scenarios as you change the load conditions.
Data Driven Load Testing Use real data to emulate realistic scenarios where many
different users are accessing your Web application
simultaneously. For example, you can use different
usernames, and passwords for different Virtual Clients.
Real-Time Server Side Statistics Using WebLOAD you can monitor a variety of server-side
application, database, system and Web-server statistics. You
can configure the WebLOAD performance monitor to
display real-time performance statistics for services
available from the server including percentage of CPU
usage, memory usage, and so on.
Feature Benefit(s)
Supported Server Side Statistics Server side statistics can be collected over generic protocols
for Open Interoperability and interfaces such as SNMP, JMX, Perfmon (Windows)
and a Generic Unix performance collector. Specific
pre-configured collectors exist for the following servers,
using the generic protocols: BEA WebLogic, IBM
WebSphere 5, Microsoft Active Server Pages, Microsoft
ASP .NET, Oracle, Microsoft SQL Server, Windows
NT/2000/XP, UNIX platforms, Windows .NET Framework,
Microsoft IIS, Sun One (former iplanet 6), and Microsoft
Media Server.
Agenda Authoring (as part of Automatically generates test scripts in JavaScript from a
WebLOAD IDE) standard Web browser.
JavaScript Agenda Definition Delivers the familiarity and scripting power of a full
(as part of WebLOAD IDE) programming language for maximum flexibility, including
writing, editing and extending.
Comprehensive Support for WebLOAD delivers comprehensive support for dynamic
Dynamic Content applications.
Capacity planning using Cruise Predictive feature for testing the performance thresholds of
Control Web applications.
Unlimited Load Generation Optimized, distributed architecture for unlimited load
generation with minimal resources.
Support for Keep-Alive Supports persistent connections with Web applications for
accurate simulation of Web browser behaviors.
Supports Multithreading Provides accurate real-world simulation of Web browser
behavior.
Browser Emulation Emulates Microsoft and Netscape browsers for maximum
real-world simulation.
Single-Point Management Console for managing WebLOAD setup, testing, monitoring
and reporting.
Real-time Dashboard View test execution statistics in real-time for a quick and
automated overview of test performance.
Real-Time Analysis Graphical and statistical analysis of test sessions while the
test session is running.
Real-Time Reports and Graphs WebLOAD's integrated reports let you see the results of
your test in real time. Test results can be viewed on the test,
Agenda, page and object level. Reports include over 75
metrics including response time, hit per seconds, pages per
second, and so on.
30 RadView Software
Getting Started
Feature Benefit(s)
Feature Benefit(s)
32 RadView Software
Getting Started
Feature Benefit(s)
Feature Benefit(s)
Error path zoom in/out and The path of the failed transaction is colored and expanded in
coloring the Data Drilling reports to emphasize the failed transaction.
A zoom in/out feature offers the ability to expand or
collapse the Data Drilling tree according to the error path.
WebLOAD Viewer Graphically see the difference between expected and actual
data.
First Byte Statistic First Byte is the time it takes a Virtual Client to receive the
first byte of data.
Agenda Level Options Unique options can now be set for each Agenda in the test
session.
Load Profiler Create schedules based on pre-defined visual templates for
easier and faster scheduling.
Sleep Recording Enables you to automatically record “think” time in the
Agenda to provide a more realistic simulation of users
accessing your Web application.
Sleep Time Control Settings Easily control sleep behavior on the fly to save script
development time.
Enhanced Statistic Setting ESS provides additional statistical information on the
WebLOAD metrics.
Pass/Fail Definitions Set rules to define the success or failure of test sessions.
TestView Scheduler Automates the process of scheduling WebLOAD tests and
test plans.
34 RadView Software
C H A P T E R 4
This section shows you how you can get started quickly using the RadView Software test site at
www.netizenbanking.com (http://www.netizenbanking.com). Then you can refer to the
complete WebLOAD documentation set to learn more techniques for using WebLOAD.
In This Chapter
Quick Start Steps ........................................................................ 35
Creating an Agenda .................................................................... 36
Configuring a Load Template using the WebLOAD Wizard ..... 40
Running the Test......................................................................... 44
Analyzing Test Results ............................................................... 45
Note: We recommend that you follow the steps in order. All examples are interrelated and
dependent on earlier steps.
Creating an Agenda
The first step in creating an Agenda is to record your actions as you interact with your Web
application.
1. Start WebLOAD IDE by selecting Start | All Programs | TestView | WebLOAD IDE.
WebLOAD IDE opens.
36 RadView Software
WebLOAD Quick Start
Each node in the Agenda Tree is actually a visual representation of JavaScript code. You
can view the contents of the nodes in the JavaScript view pane.
To the left of the Agenda Tree are WebLOAD IDE toolboxes that can be used to edit an
Agenda by dragging and dropping items from the WebLOAD IDE toolboxes into the
Agenda Tree. This makes programming easier by building the code behind an intuitive
drag-and-drop interface.
3. In the main window, in Editing Mode, click the Start Record toolbar button to begin
recording.
The following message appears.
4. Click OK.
WebLOAD IDE begins recording all of the actions you perform in the browser, as
indicated by the recording notification in the WebLOAD IDE status bar.
38 RadView Software
WebLOAD Quick Start
5. In the address bar, enter the Web address www.netizenbanking.com to go to the RadView
Software test site.
6. Navigate through the site, performing the actions you want to test.
For example:
Your actions are recorded and appear in the Agenda Tree as you navigate the site. (If you
see more nodes in the Agenda Tree with different URLs, this may be traffic generated by
browser plug-ins or extensions, for example, third-party toolbars).
7. Click the Stop Record toolbar button in WebLOAD IDE to stop the recording.
8. Click the Save toolbar button or click File | Save As to save your Agenda.
9. Type in Netizen Banking for the name of the Agenda in the Save As dialog box and click
Save.
The Agenda is saved with the extension *.wlp.
You now have a basic Agenda that can be used in a WebLOAD template. For complete
information on creating, editing, modifying Agendas, and adding functionality to your Agenda,
see the relevant sections of the WebLOAD IDE User's Guide.
perform during the test. WebLOAD supports both single Agendas (lone test scripts) and Mix
Agendas (a group of Agendas simulating groups of users performing different activities on the
application at the same time). In this Quick Start, we will create a template running a single
Agenda. For information on creating templates with multiple Agendas or Mix Agendas, see
Creating Load Templates with the WebLOAD Wizard (on page 133) in the WebLOAD User's
Guide.
1. Start WebLOAD.
Select Start | All Programs | TestView | WebLOAD Console.
The WebLOAD Console opens, and displays the WebLOAD Console dialog box.
2. Select Create a new template using the WebLOAD Wizard, and click OK.
The WebLOAD Wizard opens.
3. On the Welcome screen, click Next.
Note: In this simple example, we will monitor the CPU usage on your system during
runtime. During a real test, you would probably monitor the CPU performance of your
application server.
9. Expand the tree to select the measurement you want. Click Processor | _Total, select the
checkbox adjacent to % Processor Time, and click Next.
10. Click Finish.
11. Click the Close & update toolbar button to save the configuration with the template
and return to the WebLOAD Wizard.
12. On the Agenda / Mix Selection screen, click Next.
Selecting a Host
For the steps in this Quick Start, you use your own machine as the host load machine to generate
Virtual Clients. The number of Virtual Clients that can be generated depends on the power of
the host machine. For this test, you do not select a Probing Client.
1. On the Host Selection screen, note that your computer is automatically listed as the default
Load Machine.
2. Click Next to accept the default selection.
42 RadView Software
WebLOAD Quick Start
44 RadView Software
WebLOAD Quick Start
Select the Dashboard icon on the Console toolbar to view a quick overview of the
status of the test including the following:
Elapsed time
Current load size (number of running Virtual Clients)
Rounds executed
Failed rounds
Failed hits
Hits per second
Pages per second
Throughput
Select the Open Statistics icon on the Console toolbar. A Statistics tab appears in the
results window. Scroll through the data and double-click a cell to drill-down to further
detail.
Each of these views opens in a new tab window. Please close all windows except the Default
Report tab.
46 RadView Software
WebLOAD Quick Start
Total
Successful
Failed
Marked counts (indicates how many children of the transaction failed.)
Data Drilling provides both a global and detailed account of hit successes and failures, letting
you verify the functional integrity of your Web application at the per-client, per-transaction, and
per-instance level.
Note: You can continue exploring the reasons for the transaction failure by double-clicking
grid rows. For example, double-clicking a row in the Instance Grid opens a Parent
Transaction Instance Tree, which provides a detailed breakdown of all sub-events that
occurred as part of the selected transaction.
You can return to these test results to track progress as you tune performance and reopen the
session at a later point to perform more analysis on the data gathered. You can also export the
results.
1. Click the WebLOAD REPORTER icon on the Console toolbar or select Test View | Open
WebLOAD Reporter from the Console menu.
Data for the last session saved is automatically selected.
48 RadView Software
WebLOAD Quick Start
2. Double-click the Ten Slowest Transactions icon at the far left or the Ten Slowest
Transactions link in the Report List tree or select the Ten Slowest Transactions item from the
Reports menu.
The report will be generated in the main window.
Ten Slowest Transactions is an example of the many standard reports available. Reports are
displayed in separate windows. Use the Windows menu to move back and forth among
reports that you create.
3. Create a custom report by selecting Reports | Custom Session Report… or Reports | Custom
Agenda Report…, name the report, click the Save checkbox to save the report, select an axis
metric and statistics (from the tree), and click OK.
You can create both session and Agenda reports. Session reports display a higher level of
information for the entire session, including all the Agendas that were run. Agenda reports
display detailed data pertaining to one Agenda. In the steps outlined in this Quick Start, you
ran one Agenda so the only difference in these reports is the level of detail.
Each generated report can be either printed or exported to one of the formats that appear on
the Export toolbar.
4. Close WebLOAD REPORTER.
5. Close the WebLOAD Console.
This concludes our Quick Start guide for WebLOAD. For additional information please refer to
the WebLOAD User's Guide or to RadView's web site at http://www.RadView.com
(http://www.radview.com/support).
50 RadView Software
C H A P T E R 5
This chapter contains the instructions for installing WebLOAD on a PC running Windows 2000,
Windows 2003, or Windows XP systems.
Communication between hosts in a test session is a critical aspect of WebLOAD testing. This
chapter also describes communications between systems. It discusses TestTalk, the WebLOAD
network agent, and gives you tips on troubleshooting network problems.
In This Chapter
Installing WebLOAD.................................................................. 51
License Registration and Update ................................................ 52
Uninstalling WebLOAD............................................................. 61
Upgrading WebLOAD................................................................ 62
Communication Between Host Computers................................. 62
Configuration.............................................................................. 69
Working with a Firewall ............................................................. 73
Installing WebLOAD
You must have at least 260 MB of available disk space for the complete installation of
WebLOAD.
When you install WebLOAD on your computer, the installation program asks you for the
components to install.
Install the WebLOAD components as follows:
WebLOAD User's Guide 51
License Registration and Update
Install the full WebLOAD product (with or without WebLOAD REPORTER) on computers
that run the WebLOAD Console.
Install just the Probing Client or the Load Machine software on the network host computers.
Do not install the full product on the network host computers.
Notes:
WebLOAD components in each major version will work with all other components of the same
version, no matter what the minor revision number. WebLOAD components are not compatible
with other components of a different version. For example, all components of WebLOAD
Version 7.x will work with each other. Version 7.0, 7.1, etc, are all compatible with each other.
However, components of Version 7.x are not compatible with components of Version 6.x or
Version 5.x. Verify that the WebLOAD components have been upgraded to Version 7.x on all
participating hosts. Running WebLOAD Console Version 7.x with an earlier version of Load
Generator (running on different Hosts) may cause unexpected behavior.
The WebLOAD license file is limited to the computer system (machine) on which WebLOAD is
initially installed. Before installing WebLOAD, make sure you are installing on the machine
with which you intend to work.
When installing WebLOAD with Unicode support on a system running an English MS
Windows, Japanese must be configured as the default Windows language.
52 RadView Software
WebLOAD Installation and Configuration
The Update License dialog box can be opened in one of two ways:
At the time of WebLOAD installation, when the WebLOAD Installation Wizard is finished,
the License dialog box opens automatically,
Note: After a WebLOAD installation, you may be prompted to restart your computer before
the License dialog box appears. In this case, the License dialog box will appear
automatically after the computer restarts.
-Or-
At any other time, from the Windows desktop, click on Start | All Programs | TestView |
Tools | Update License.
The License dialog box includes three tabs:
Host ID
Select License
License Information
These tabs are used to register and update your WebLOAD authorization license information.
a. Copy the Host ID displayed in the text box into the email.
b. Add your name, company, address, and phone number to the email message.
4. Send the email to [email protected].
A WebLOAD license file (*.lic) will be sent to you.
5. After receiving the file, save it to your hard drive in the following directory:
Program Files\RadView\WebLOAD\License
54 RadView Software
WebLOAD Installation and Configuration
6. Select Start | All Programs | TestView | Tools | Update License to open the Update License
dialog box and install the license key.
7. Select the Select License tab.
8. By default, the License file radio button is selected. This assumes that your license file is
located on your local computer system.
For information on installing a floating license, or connecting to a license server, see
Floating License Server (on page 57).
9. Click on the Browse (…) button next to the License file text box to browse to the location of
the license (*.lic) file sent to you by a RadView representative.
10. Select the correct license file and click Open to return to the Update License dialog box.
56 RadView Software
WebLOAD Installation and Configuration
If you are not able to successfully register a valid license file, contact RadView Support for
assistance.
The number of concurrent applications for each product (RM Workstation, WebFT,
FTRunner, and TestView Manager).
The number of concurrent Virtual and Probing Clients.
Installation also creates a FlexLM Service that controls the operation.
The license server displays a list of currently connected RM Workstations and the number of
resources (Virtual and Probing Clients) each is using. A Floating License Server enables a
testing team to share resources.
This section describes how to connect to the Floating License Server and how to view License
details. For more information about running WebRM, WebFT, and FTRunner in Floating
License Mode, updating Floating License Resources, revoking Resources, or working with the
Floating License Server Console, see the TestView User's Guide.
58 RadView Software
WebLOAD Installation and Configuration
3. Select the License Server radio button, and press the browse button to select the license
server.
The Host Selection dialog box opens.
4. Select the license server host and click Select to return to the Update License dialog box.
5. Click OK.
A connection to the selected license server is attempted.
60 RadView Software
WebLOAD Installation and Configuration
For more information about running WebRM, WebFT, and FTRunner in Floating License
Mode, updating Floating License Resources, revoking Resources, or working with the Floating
License Server Console, see the TestView User's Guide.
Uninstalling WebLOAD
To uninstall WebLOAD:
1. Click Start | Settings | Control Panel.
2. Double-click the Add/Remove Programs icon.
3. Click the Install/Uninstall tab.
4. Click WebLOAD, and then click Add/Remove.
The Uninstall Wizard appears.
Note: It is only necessary to delete the WebLOAD program files and not the user files.
WebLOAD Load Templates, Load Sessions, and Agenda files can be used with all WebLOAD
upgrades.
Caution: Close TestTalk before uninstalling WebLOAD. Failure to close TestTalk results in
outdated files remaining on your disk, which could adversely affect future installations of the
program.
Upgrading WebLOAD
To upgrade WebLOAD:
1. Close TestTalk.
2. Install the new version of WebLOAD.
For installation instructions, see Installing WebLOAD (on page 51).
Note: WebLOAD components in each major version will work with all other components of the
same version, no matter what the minor revision number. WebLOAD components are not
compatible with other components of a different version. For example, all components of
WebLOAD Version 7.x will work with each other. Version 7.0, 7.1, etc, are all compatible with
each other. However, components of Version 7.x are not compatible with components of
Version 6.x. This applies to executable components only. All WebLOAD data files, including
WebLOAD Load Templates, Load Sessions, and Agenda files, can be used with all WebLOAD
upgrades.
SNMP
Simple Network Management Protocol (SNMP) enables communication between systems in
your Load Session. If SNMP service is installed on your system, it runs automatically in the
background.
62 RadView Software
WebLOAD Installation and Configuration
Opening TestTalk
TestTalk is opened automatically when you open the Console.
64 RadView Software
WebLOAD Installation and Configuration
Caution: Do not close TestTalk in the middle of a test session. This can cause unexpected
behavior. When shutting down the system, be sure to shut down the Console first and only then
shut down TestTalk.
Menu Function
File Contains the Exit option. Use this option to Close the TestTalk program. Clicking
the Windows Close button does not close TestTalk. This is a safety feature that
prevents you from accidentally closing TestTalk when the Console is active.
View Contains the TestTalk display configuration options.
The first two options in this menu are:
Toolbar
Status Bar
Toggle these options on/off to display the Toolbar or the Status Bar
The Clear View Window option clears the display window.
Help Contains the options for viewing the program information and information about
your system. The Help menu contains two options:
Select About TestTalk to display the information about this program.
Select Net Configuration to display your system configuration information.
Field Description
Computer's name The name of your computer. You must know the name of each of
the hosts to configure your test session. If you are not familiar
with the name of your host, open TestTalk at that host and view
this name in this message box.
Computer's address The IP address. You need this information to ping a host if you
want to verify that it is connected to the network.
While TestTalk is running, the TestTalk icon is displayed on the Status bar in the lower
right hand corner of the screen.
2. Right click on the TestTalk icon to display the following menu:
66 RadView Software
WebLOAD Installation and Configuration
Pinging a Host
You can test a host's physical connection to the Console by pinging the remote host. A ping is a
signal that is sent from the Console to a remote host. The response that the Console receives
from the ping indicates the status of the connection to the remote computer.
From the operating system, type ping systemname.
Example:
c:> ping systemname
Troubleshooting Communications
On occasion, you may have trouble with a WebLOAD host. Refer to the following table for
possible solutions.
Problem Solution
At the beginning of a test session, the host prompts: Test your connection, as described
in Testing Your Connection (on
All hosts are stopped.
page 67).
Indicating that none of the hosts are accessible and the
Domain Name Server (DNS) cannot recognize a host
name.
At the beginning of a test session, a host prompts: Test your connection, as described
in Testing Your Connection (on
Load Session Setup -timed out, not completed.
page 67).
Indicating that at least one host is not accessible and the
Domain Name Service (DNS) cannot recognize a host
name.
68 RadView Software
WebLOAD Installation and Configuration
Problem Solution
Your system starts, but the test session times out during Test your connection, as described
operation. in Testing Your Connection (on
page 67).
Check the names of the computers
in the TestTalk Help menu, under
Net Configuration. Verify that there
is complete consistency between the
names of the computers and the
names used in WebLOAD.
WebLOAD reports unexpected results from a test. Access the Application Being
Tested (ABT). If you have a good
connection, the test results are
accurate.
The system issues the message: The Console must specify the
complete path of the Agenda that is
Error loading Agenda…
used by the remote host. For
example, if the Agenda is on a
shared disk, the name of the disk in
the path specified by the Console,
must be the same name used by the
remote host.
Configuration
A Load Session includes the following components:
Component Description
Console The Console controls set up, manages a Load Session, and
displays the Load Session results.
It is recommended that the Console reside on its own separate
system and not be installed on systems running Load
Generators or Probing Clients.
ABT The Web application to be tested. WebLOAD can test any Web
application that can be accessed through a Web browser, on the
(Application Being Tested)
Internet or intranet. The ABT requires no special software.
Load Machine The host(s) that runs Load Generators. The Load Generators
bombards the ABT with user-defined transactions for testing
scalability and integrity of the ABT.
Component Description
Probing Client Machine The host(s) that run Probing Clients. A Probing Client runs at
(optional) the same time as load testing. WebLOAD generates exact
values for the Probing Client performance as opposed to
averages for Load Generator performance.
WebRM WebRM distributes and circulates WebLOAD testing resources
(Virtual Clients and Probing Clients) amongst users on a “need
to use” basis.
With WebRM, every WebLOAD Console can operate in
Standalone Workstation mode or Connected Workstation mode.
For more information, see Connecting to WebRM (on page
121).
The following diagram shows a simple WebLOAD test configuration. The Load Machine and
the Probing Client Machine both access the ABT. The Console controls the test from a separate
system. All three computers are networked.
Console TestTalk
Console
WebFT (optional)
ABT (Application Being Tested) None
Load Generator TestTalk
Probing Client (optional) TestTalk
Shared disk (optional) None
70 RadView Software
WebLOAD Installation and Configuration
This test session has three participating host computers. Two hosts act as Load Machines, and
one host is a Probing Client Machine. In this configuration, the Console is run on a separate
system. All systems are on a network with a shared disk that contains the Agenda.
The Console
In the diagram above, the Console is a node on the network along with the host computers in the
test session, and the shared disk. The Console communicates with the hosts that are testing the
ABT through the TestTalk network agent. The Console controls the test setup, management and
reporting.
Note: Running a Load Machine on the same system running the Console software is not
recommended.
Probing Client
The Probing Client resides on the network with the other hosts in the test session. The Probing
Client tests the performance of the ABT while it is being bombarded by the Load Machines. In
WebLOAD User's Guide 71
Configuration
this test session there is one Probing Client, however WebLOAD can support multiple Probing
Clients in a test session. WebLOAD generates exact values for the Probing Client performance
as opposed to averages for Load Machine performance.
Sharing an Agenda
You can increase the size of the test load by distributing the load generation tasks over multiple
Load Machines. You may want all of these Load Machines to test the ABT using the same
Agenda. There are several ways to share an Agenda over multiple systems.
Shared disk:
You can configure a shared disk that can be accessed by all the systems on the network, as
shown in the previous diagram. The Agenda can then reside on the shared disk, and can be used
by all systems for testing. Thus there is only one copy of the Agenda, and all the relevant hosts
run it from the shared disk.
In this configuration, you must make sure that in the Agenda configuration, all hosts refer to the
shared disk and are mapped to the same logical name. On one host the logical name of the
shared disk may be “K” while on another host the logical name may be “L.” Each host must
configure the Agenda using the correct logical name.
72 RadView Software
WebLOAD Installation and Configuration
In some cases you may want to use machine resources in the test session that reside on different
sides of a firewall. In this case, you must change your initial settings to enable WebLOAD to
accommodate this unique configuration.
The following diagrams illustrate situations where the Load Machine and the Console are on
opposite sides of the firewall.
The following configuration is required to enable data transfer through the firewall. Generally,
the rules that need to be added to the firewall are:
74 RadView Software
WebLOAD Installation and Configuration
Allow outgoing TCP from the Console all ports to Probing Client port 9000.
Allow incoming TCP from Probing Client port 9000 to Console all ports (options=ackonly).
Allow outgoing UDP from Console port 137 to Probing Client port 137.
Allow incoming UDP from Probing Client port 137 to Console port 137.
76 RadView Software
C H A P T E R 6
The WebLOAD Console is where you perform the majority of your load / performance testing
activities.
In This Chapter
The WebLOAD Console ............................................................ 77
Console Elements ....................................................................... 79
Edit Mode and Run Modes of Operation.................................... 96
The Console Menu Options ........................................................ 98
The Console Toolbar .................................................................. 105
78 RadView Software
WebLOAD Console Features
The Console is operated through its toolbar and menus at the top of the screen. Tests are
configured through a series of dialog boxes.
Console Elements
When you first start WebLOAD, you see the Console interface in Edit mode. The Console
window displays different components in the two different modes:
Edit Mode where you create your templates.
Run Mode where you run your sessions and see the results.
The Console interface consists of the following elements.
Title Bar
Displays the name and path of the current template or session.
Menu Bar
Displays the commands you use to work with WebLOAD. Beside the standard File, View,
Window and Help menus, menus are provided to access functions specific to template creation
and session management such as Session Control and Reports. For more information on the
menus, see Edit Mode and Run Modes of Operation (on page 96).
Context Menus
Contain shortcuts to frequently performed actions.
Toolbar
Provides quick access to commonly used commands. You click a button on the toolbar once to
carry out the actions represented by that button.
For more information on the Console toolbar, see The Console Toolbar (on page 105).
80 RadView Software
WebLOAD Console Features
Status Bar
At the bottom of the Console window, the status bar displays information about operations in
progress.
Note: You can also access the configuration dialog boxes through the toolbar and the menus.
While your test session is running, the session tree displays its progress. The session tree
displays the Agendas configured for running, the participants in the test, and their status. For
more information on the status of the test, see Status Icons (on page 82).
Status Icons
During runtime, status icons appear adjacent to each load session item (hosts, Agendas) in the
session tree, indicating the item's status.
The status icons are described in the following table.
82 RadView Software
WebLOAD Console Features
Icon Description
No current activity.
Key Result
Display Explanation
Agenda1 Describes one of the Agenda files (*.pft / *.js) used for running this
test session.
TestLab 22 Describes one of the Load Machines running the Agenda, Agenda1.
The Load Machine icon is followed by the host name, in this case
“TestLab 22”. A test session may include multiple Load Machines.
TestLab 16 Describes one of the Load Machines running the Agenda, Agenda1.
The Load Machine icon is followed by the host name, in this case
“TestLab 16”. A test session may include multiple Load Machines.
TestLab 1 Describes one of the Probing Client Machines used in this test
session. The Probing Client Machine icon is followed by the host
name, in this case “TestLab 1”. A test session may include multiple
Probing Client Machine hosts
Agenda4 Describes the one of the Agenda files (*.js) used for running this test
session.
TestLab 9 Describes one of the Load Machines running the Agenda, Agenda4.
The Load Machine icon is followed by the host name, in this case
“TestLab 9”. A test session may include multiple Load Machines.
Mix1 The Mix file (*.mix) to be run during the load session. The icon is
followed by the name of the Mix, in this case “Mix1”.
TestLab 24 Describes one of the Load Machines running the Mixed Agenda,
Mix1. The Load Machine icon is followed by the host name, in this
case “TestLab 24”. A test session may include multiple Load
Machines.
84 RadView Software
WebLOAD Console Features
Using the session tree you can define and modify load templates.
The following options can be accessed by right clicking on a session tree component in edit
mode:
86 RadView Software
WebLOAD Console Features
modes of operation, the current load size will never exceed the maximum load size scheduled
for that Agenda listed to the right of the current value.
Note: Using Throttle Control, you can increase the current load size to exceed the maximum
load size - but not while running a Cruise Control session.
In run mode you can right-click on an item in the session tree to view the test parameters defined
in the session.
While in run mode, the following options can be accessed through the right mouse button:
Load Session View Agendas Selection Open the Agenda / Mix Selection dialog box.
Stop All / Pause / Resume Control the test in progress.
88 RadView Software
WebLOAD Console Features
Report Description
Integrated Reports A report that can be configured to combine Performance data from
the Load Machines and Probing Clients, Performance data on the
ABT, gathered from Performance Manager and existing
performance data from previously saved load sessions.
Report Description
Statistic Reports A real-time monitor that displays the results of your test while you
run it that can be saved and exported.
WebLOAD Dashboard Provides a summary of the performance statistics generated during
the test session.
Data Drilling Reports Provides a detailed description of all user-defined and named
transactions to the instance level, including all hit failures and the
reasons for them.
Viewer Provides a graphic presentation of event failures displaying both
the expected and actual events.
Integrated Report
WebLOAD Integrated Reports provide both a graphical and statistical view of the performance
of your application as it is being tested. Integrated Reports can be viewed while the test is in
progress or saved for later analysis.
90 RadView Software
WebLOAD Console Features
Statistics Report
WebLOAD collects approximately 35 different statistics during a test. The Statistics Report
displays the values for all of them.
Dashboard
The Dashboard displays real-time statistical information about the test session including the
number of Virtual Clients running, hits per day, pages per day and throughput.
92 RadView Software
WebLOAD Console Features
Event Viewer
The Event Viewer displays an illustration of each event triggered during the test.
Log Window
The Log Window displays a summary of the test including all log messages detected by
WebLOAD in run time that are generated by the Console, the JavaScript compiler and any user
messages programmed in the test script.
94 RadView Software
WebLOAD Console Features
Item Description
Cruise Control Opens the Cruise Control Wizard for creating a new load template.
Wizard
WebLOAD Wizard Opens the WebLOAD Wizard for creating a new load template.
Manual Opens a new blank session tree for creating a new template.
Load Template Opens a saved load template (*.tpl). Enter the name of the saved load
template in the edit box.
Load Session Opens a saved load session (*.ls) including template data from a
completed test. Enter the name of the saved load session in the edit box.
Use either the Cruise Control Wizard to develop a goal-seeking performance test or the
WebLOAD Wizard to develop a basic load template. Both of these wizards enable you to save
your test parameters as a template that you can later edit manually using the Console toolbar and
menus.
After developing a test with either of the wizards, you can save the test parameters as a load
template. After running the test you can save the test parameters and results as a load session.
You can then edit the test parameters by editing the load template or view the load session test
results. For more information on load sessions and load templates, see Managing Load
Templates and Load Sessions (on page 125).
If you are in Run mode and you wish to edit the test parameters, press the Edit Template
toolbar button to switch to edit mode.
96 RadView Software
WebLOAD Console Features
The following table describes the segments of the Console in Run mode:
Segment Function
Session Run Displays only during a test run. It contains buttons for controlling the test
Toolbar session.
Session Tree Presents a graphic display of your test session.
Results Window Displays all of the reports opened during a test session. Use the tabs
located at the top and the bottom of the window to view different reports.
Log Window Displays all of the error messages recorded during a test session. You can
toggle the Log Window display on/off through the View menu.
Status Bar Indicates the program status. You can toggle the Status Bar display on/off
through the View menu.
This section discusses each of the WebLOAD menus, their options and their related toolbar
buttons.
File Menu
The File menu includes the following functions for WebLOAD file management. The Console
toolbar has an icon corresponding to menu options.
98 RadView Software
WebLOAD Console Features
Save, Save As (none) Saves the current load configuration as a load template file. The
Load Template following information is saved:
Selected hosts, both Probing Clients and Load Machines
Assigned Agendas for each host
Test session schedule
Report configuration
Edit Template Transfers from Run mode to Edit mode.
Print Prints the report displayed with all contents and formats.
Modify Hosts Opens the Host Selection dialog box. In Run mode, you can view
Selection the contents of the Host Selection dialog box. In Edit mode, you can
define the hosts you wish to designate as Probing Client Machines
or Load Machines.
Open Session Opens a graphic representation of the current load session
Tree configuration. Double-clicking on components before the test
session begins lets you modify the selected component's
configuration.
Modify (none) Opens the Agenda / Mix Selection dialog box. In Run mode you can
Agenda view the selected Agenda and Mix files. In Edit mode you can add
Selection or delete Agenda and Mix files to the session.
Performance (none) Opens the Performance Statistics Manager dialog box for
Statistics configuring the statistics parameters to be monitored.
Manager
Modify (none) Opens the Modify Schedule dialog box that corresponds to the
Schedule currently selected host or Agenda. In Run mode you can view the
contents of the Modify Schedule dialog box. In Edit mode you can
change the schedule.
Throttle Opens the Throttle Control dialog box. Use Throttle Control to
Control dynamically change the load while the session is in progress.
Delete Deletes the currently selected load session component. A
component may be an Agenda, or a host. A message box pops up,
asking for confirmation.
Start Session Starts executing the current load session.
Stop All Stops all load session activities. A message box pops up, asking for
confirmation.
Stop (none) Stops the currently selected load session components. If the root is
currently selected, all components are stopped. Selecting stop
during a Cruise Control test session stops all components.
Pause Temporarily stop execution of the load session component. If the
root is selected, all components are paused
Resume Resumes execution of a paused load session component.
Note: Pausing and continuing does not automatically reschedule
load session Agendas. Scheduling continues in relation to the start
of the session.
Update Open the Update Resources dialog box.
floating license
resources
Reports Menu
The Reports menu provides options for creating real-time and summary reports showing the
performance data gathered by WebLOAD. The following is a brief description of each report. A
full description of all WebLOAD reports can be found in Running a Load Session (on page
275).
View Menu
Menu Entry Icon Description
Tab Description
Example: You can remove all the file options from the toolbar by disabling the file entry in the
toolbar dialog box. The file-specific toolbar segment is not displayed.
The first five toolbar segments are checked by default. The Export toolbar is not displayed by
default. When using the WebLOAD export features, check this option to use the convenient
Export toolbar buttons.
Toolbar buttons are moved or removed by dragging and dropping them where desired. To
remove a toolbar button drag and drop it from the Console toolbar to this dialog box.
Tools Menu
Menu Entry Icon Description
Global Options Opens the Global Options dialog box. Use the global options to
customize WebLOAD to your specifications, see Setting Global
Options (on page 203).
Default Options Opens the Default Options dialog box. Use the default settings to
customize the Agenda settings to your specifications. When
building a new template or session, or editing an existing one,
each new single Agenda or Agenda participating in a Mix is
automatically assigned these defaults, see Setting Agenda
Options (on page 221).
Current Session (none) Opens the Current Session Options dialog box. Use the session
Options settings to customize the current session to your specifications.
This option will update the options for all Agendas in the
current session, see Setting Agenda Options (on page 221).
Window Menu
Use the Window menu to configure the look of the Results window in the Console.
Help Menu
Use the Console Help menu to access:
Online help
License information
WebLOAD Wizard
Copyright and version information about the Console
Options Description
Toolbar Buttons
The following figure details each button on the Console toolbar. (Some items, such as Throttle
Control, Stop, Pause and Go display when you start a test or select them from Tools |
Customize.)
Button Description
New Starts a new load template. A new empty session tree displays.
Open Load Opens an existing load template.
Template
Save Load Session Saves a load session. This includes the load template and the data
accumulated running the test.
Edit Template Transfers from Run mode to Edit mode.
Print Prints the active chart or grid displayed in the Results window, with all
contents and formats.
Button Description
Options Description
Options Description
Reports Options Commands used to open, delete and create new reports.
Export Options Commands used to export Integrated and Statistic reports to Excel,
HTML and tab files, and to create and export a summary report.
Window Options Commands used to define multiple window presentations. Windows can
be displayed tiled, cascaded, or adjacent to one another. This toolbar can
also be used to close windows.
Button Description
Button Description
Using WebLOAD, you create load templates to define the events to occur in the load session.
This part describes how to quickly set up a load template. Load templates define the participants
(Clients and Servers) in the load session, the Agendas to run, the number of Virtual Clients to
emulate and the test schedule.
In This Chapter
WebLOAD Options for Creating Load Templates: .................... 111
Using the WebLOAD Wizard to build a basic load template. You define the Agendas to run,
the Load Machines on which to run the Virtual Clients and the test schedule. After creating
a basic template with the wizard, you can change aspects of the configuration through the
Console toolbar and menus.
Using the Cruise Control Wizard to create goal-seeking performance tests. Rather than
running multiple tests to measure the performance at different loads, you can run Cruise
Control to measure the performance goal that you desire. You define the Agendas to run and
the goals to achieve and WebLOAD automatically schedules the test session and increases
the number of Virtual Clients running until your goals are achieved.
Manually using the Console toolbars and menus.
The load template material is divided into five chapters:
Planning a Test Session (on page 113) describes the stages involved in planning and
executing a test session.
Connecting to WebRM (on page 121) describes the process of connecting to WebRM and
requesting testing resources.
Managing Load Templates and Load Sessions (on page 125) provides information about
working with Load Templates and Load Sessions.
Creating Load Templates with the WebLOAD Wizard (on page 133) guides you through a
basic template configuration using the WebLOAD Wizard.
Creating Load Templates with Cruise Control (on page 161) teaches you to configure a
goal-seeking performance test.
Configuring a Load Template Manually (on page 189) provides instructions for configuring
a test template manually.
After completing these chapters, you will have the basic skills necessary to create and run your
own test.
Planning your test is imperative for developing tests that accomplish your load testing
objectives. Planning your test helps you:
Plan the resources required for testing.
Create templates that accurately emulate your users typical working environment.
Define test success criteria.
This chapter describes the stages involved in planning and executing a test session. It shows you
where to find details in this guide about each concept so you can quickly find information of
particular interest to your needs.
In This Chapter
Load Session Workflow ............................................................. 113
Forming a Test Plan.................................................................... 115
Starting WebLOAD .................................................................... 117
What you want to test - what application are you going to test?
What functionality do you want to test - what actions will the users perform?
How many Virtual Clients you want to simulate?
How long your test will run?
What are acceptable results? Acceptable results are defined by your test objective. For
example, you can verify:
Acceptable user response times
Reliability by running stress tests
Performance degradation after updates
What resources are required for performing the test?
Creating Visual Test Agendas
Use WebFT to record the activity you want to test in a Web browser and save it as an
Agenda. If you plan to test users performing different activities, you need to record
separate visual Agendas for each activity. See the WebFT User's Guide.
Creating Load Templates
Creating load templates involves defining the participants (Clients and Servers) in the
load session. The definition includes selecting the Agendas to run, the host computers
and number of Virtual Clients participating in the load session, and scheduling the test.
After you create a template it is saved in a *.tpl file. Use the commands in the Session
Control menu to create, open and save template files.
WebLOAD provides three ways for creating load templates:
Using the WebLOAD Wizard to build a basic load template. You define the Agendas to
run, the Load Machines on which to generate the Virtual Clients and the test schedule.
After creating a basic template with the wizard, you can change aspects of the
configuration through the Console toolbar and menus. For information on creating a
template with the WebLOAD Wizard, see Creating Load Templates with the
WebLOAD Wizard (on page 133).
Using the Cruise Control Wizard to create goal-seeking performance tests. Rather than
running multiple tests to measure the performance at different loads, you can run Cruise
Control to measure the performance goal that you desire. You define the Agendas to run
and the goals to achieve, and WebLOAD automatically schedules a test and increases
the load until your goals are achieved. For information on creating a template with the
Cruise Control Wizard, see Creating Load Templates with Cruise Control (on page
161).
Manually using the Console toolbars and menus. See Configuring a Load Template
Manually (on page 189) for information on defining a template manually.
Running a Load Template
After you create the load template, you run it. As your test runs, you receive various
reports and messages from the various participating hosts. The activity in your test
114 RadView Software
Planning a Test Session
session is displayed on the Console screen. This screen is updated continuously in real
time as you work. See Running a Load Session (on page 275).
Analyzing the Test Results
WebLOAD displays the results of your load session in tabular format and in easy to
read graphs.
What type of connections speeds do you want to simulate? Using the connection speed
settings available with WebLOAD you can simulate users connecting through different
connections.
3. What types of systems are testing these aspects?
Your test can include Load Machines and Probing Clients. A Load Machine bombards the
ABT with a load of Virtual Clients. Load Machine testing gives you an indication of the
way your program behaves under the stress of a heavy load.
Probing Client testing is usually performed simultaneously with Load Machine testing.
Using a Probing Client, you can test the system performance with other functionality, while
the system is undergoing stress testing. For example, you can measure the time required to
submit a form, while the ABT is being bombarded by the Load Machine simulating multiple
clients calling up a page.
WebLOAD generates exact values for Probing Client performance as opposed to averages
for Load Machine performance.
the basic activity, you can add transaction statements to measure the time it takes to display the
page.
You can tailor the recorded Agenda, by adding commands and timers available in WebFT. For
more information on creating an Agenda with WebFT, see the WebFT User's Guide.
Starting WebLOAD
When your test plan is complete and your Agendas are ready, you can open WebLOAD and
begin test configuration process.
To open WebLOAD:
1. Verify that all the participating Host computers are properly connected. See Pinging a Host
(on page 67).
2. Select Start | All Programs | TestView | WebLOAD Console.
The Console calls up TestTalk automatically, on its system.
The Console opens with the Console shortcut screen displaying a variety of shortcuts to the
various WebLOAD components.
To create a new WebLOAD template, select one of the three methods available on the
shortcut screen.
Manual - create the load template by yourself, defining the Agendas to run, the number
of Virtual Clients to run, the Load Machines on which to generate the load and build a
schedule for running the test.
WebLOAD Wizard - the wizard helps you through the process of configuring the load
template.
Cruise Control wizard - you define the Agendas to run, the Load Machines to run the
Virtual Clients on and the test goals to achieve and WebLOAD automatically builds a
schedule based on your goals.
The easiest way to configure a basic test is through the WebLOAD Wizard. Save your work
in the Wizard as a load template. After you have closed the WebLOAD Wizard, you can use
the Console menus to add functionality to your load template (*.tpl file).
Creating Load Templates with the WebLOAD Wizard (on page 133) discusses the
WebLOAD Wizard. Setting Global Options (on page 203) provides instructions for using
the Console features to add functionality to the load template created in the WebLOAD
Wizard. Creating Load Templates with Cruise Control (on page 161) provides instructions
for configuring a test session running the Cruise Control feature. Connecting to WebRM (on
page 121) provides instructions for connecting the Console to WebRM using the WebRM
Wizard. The Console can then request resources from WebRM so it can work in Connected
Workstation mode.
3. Click the TestTalk icon on each host computer participating in the test. (TestTalk is
automatically enabled only on the host running the Console.)
If you are using a Solaris or Linux system as a Load Machine, verify that the TestTalk
software is running on the Solaris or Linux system. If you have closed TestTalk after the
installation, run TestTalk again.
All of the systems participating in the test session, including the system running the
Console, the Load Machines, and the Probing Clients must be running the network agent,
TestTalk. This establishes communication between participating clients and servers. The
ABT does not require TestTalk or any other WebLOAD-specific software.
Note: Do not close TestTalk in the middle of a test session. This can cause unexpected results.
When shutting down the system, be sure to shut down the Console first. Then shut down
TestTalk.
9. Connecting to WebRM
In This Chapter
Getting Started............................................................................ 121
Working with WebRM ............................................................... 123
Getting Started
WebRM addresses corporate functional and performance testing efforts by managing
WebLOAD Virtual Clients, Probing Clients, and Connected Workstations for powerful
collaborative testing and analysis, improved communications and faster issue resolution.
WebRM enables multiple users involved in various stages of application development and
testing to share testing resources. By distributing WebLOAD testing resources each developer
can run a test session to simulate, validate, and pinpoint where performance problems occur at
any stage of the development life cycle, thus eliminating design flaws and ensuring product
quality. Using WebRM, WebLOAD resources can be used optimally to fulfill an organizations
goals and priorities.
WebRM can either run on a separate machine from the other WebLOAD Consoles/Load
Machines or it can run on a machine that is also running WebLOAD Consoles/Load Machines.
About WebRM
WebRM enables developers and QA teams to perform functional and performance testing
throughout the development cycle, to ensure that applications respond quickly in the way they
were designed to work.
Using WebRM, WebLOAD testing resources (Virtual Clients and Probing Clients) can be
distributed and circulated amongst users on a “need to use” basis. WebRM is packaged with a
maximum number of Virtual Clients, Probing Clients and Connected Workstation ports, as
defined by the WebLOAD package.
The following diagram illustrates a typical WebRM configuration:
3. Enter the number of Virtual Clients and Probing Clients you will need for your test. If you
request more than the number of resources allowed for your license, or more than are
currently available, you will be prompted with a dialog stating that the resources you
requested are unavailable. This dialog will also list what resources are currently available so
that you can resubmit your request.
4. If you know that you will require the same number of resources each time you connect to
WebRM, you can enable the checkbox to Always request same number of clients.
5. Click OK.
Load templates define the events that occur in the test including:
The Agendas to run
The hosts participating in a test session
The test schedule
The number of Virtual Clients to run
Load templates save you the time of reconfiguring the test when you are repeating tests. Load
sessions include the test definitions in the load template along with test results.
In This Chapter
Managing Load Templates ......................................................... 125
Managing Load Sessions ............................................................ 129
Agendas
Information on participating hosts
Host schedules
Virtual Clients to run
Run time configuration options
Report configurations
Load templates include all test information except for the test results. Load template files are
saved as *.tpl files. Load templates can be edited using the Console toolbar and menus. You
can then save the load template for future test iterations. For example, if you build a test in the
WebLOAD Wizard you can call it up in the Console and edit the load template manually,
changing perhaps the hosts, the number of Virtual Clients and even the Agenda. Use the
commands in the File and Session Control menus to create, open, save and close template files.
Descriptive title
Author name
Subject of the test
Application being tested
Other important information about the load template
Use the Additional Information dialog box to display information about the load template.
2. Complete the fields to save additional information, useful for later reference, with the load
template.
The following fields are available:
Field Description
Title Provides a space for you to type a title for this load template.
The title can be different then the template file name.
Subject Provides a space for you to type a description of the subject
of the load template. Use this property to group similar load
templates together.
Created by Provides a space for you to type the name of the person who
authored this load template.
Test Description Provides a space for you to type a description of the test
objectives and what the load template is designed to test.
Application being tested - Provides a space for you to type the name, version and build
version and build number of the application being tested.
Template comments Provides a space for you to type any comments regarding the
load template.
Custom Provides a space for you to type any comments you want
saved with this load template.
3. Click OK.
-Or-
2. Complete the fields to save additional information, useful for later reference, with the load
session. Some of the fields, such as title, subject and application being tested, may already
be completed if the information was saved with the load template.
The following fields are available:
Field Description
Title Provides a space for you to type a title for this load session.
The title can be different from the template file name.
Subject Provides a space for you to type a description of the subject
of the load session. Use this property to group similar load
sessions together.
Field Description
Created by Provides a space for you to type the name of the person who
authored this load session.
Test Description Provides a space for you to type a description of the test
objectives and what the load session is designed to test.
Application being tested - Provides a space for you to type the name, version and build
version and build number of the application being tested.
Results Description Provides a space for you to type a description of the test
results.
Template comments Provides a space for you to type any comments regarding the
load session.
Session comments Provides a space for you to type any comments regarding the
load session.
Custom Provides a space for you to type any comments you want
saved with this load session.
3. Click OK.
The easiest way to configure a load template is using the WebLOAD wizard.
In This Chapter
Using the WebLOAD Wizard to create Load Templates ........... 133
The WebLOAD Wizard Workflow ............................................ 134
Opening the WebLOAD Wizard ................................................ 135
Selecting an Agenda or Mix ....................................................... 136
Selecting an Agenda ................................................................... 137
Selecting a Mix ........................................................................... 138
Selecting Host Computers .......................................................... 141
Scheduling the Load Session ...................................................... 144
Completing the WebLOAD Wizard ........................................... 158
After Completing the WebLOAD Wizard.................................. 159
Agenda(s) to run
Load machines on which the load is generated
Number of Virtual Clients to run
Load schedule
You can also specify Agenda options including the type of browser to emulate connection
speed, and playback sleep time options.
The WebLOAD wizard walks you through the configuration process step-by-step. Each screen
of the WebLOAD wizard contains text explaining the configuration process.
After creating a load template with the WebLOAD Wizard, you can add functionality, not
available in the wizard, through the Console menus. The Console allows you access to any part
of the configuration process at any time. For details about the Console menus, see The Console
Menu Options (on page 98).
Note: You cannot use the WebLOAD wizard if you have not yet recorded (or otherwise created)
any Agendas. If you do not have any Agendas to work with, you must exit the wizard and create
an Agenda using one of the authoring tools, such as WebLOAD IDE.
2. Click Next.
The WebLOAD wizard progresses to the Agenda/Mix Type dialog box.
Selecting an Agenda
The WebLOAD wizard enables you to create a test using previously recorded Agendas. If you
do not have any Agendas to work with, you must exit the wizard, and create an Agenda using
one of the authoring tools, such as WebLOAD IDE.
To select an Agenda:
1. Click the browse button to select the source of the Agenda to be used for the test.
2. (Optional) To open and view or edit the selected Agenda, press View.
WebFT opens displaying the selected Agenda.
3. (Optional) To configure runtime options exclusive to this Agenda, press Options.
The Agenda Options dialog box opens enabling you to define the runtime options for the
selected Agenda. For more information on Agenda options, see Setting Agenda Options (on
page 221).
4. (Optional) Open the Performance Measurements Manager and configure WebLOAD to
collect performance statistics from your application server, database, Web server and stream
technologies while the test is running. Click Measurements Manager to configure the
performance statistics to monitor and display in real-time during the test session.
The Performance Measurements Manager opens. For more information on configuring the
statistics to monitor, see Performance Measurements Manager (on page 365).
5. Click Next.
The Wizard progresses to the Host Selection dialog box. See Selecting Host Computers (on
page 141).
Selecting a Mix
The WebLOAD wizard enables you to run your test using a saved Mix or to create a new Mix of
Agendas through the wizard. If you do not have any Agendas to work with, to create a Mix, you
must exit the wizard, and create an Agenda using one of the authoring tools, such as WebLOAD
IDE.
To select a mix:
1. Select the source of the Mix to be used for the test on the Mix Selection screen.
Select Use an Existing Mix and click the browse button to choose an existing Mix
(*.mix) from your system.
Select Create a Mix to build a new Mix.
2. (Optional) You can set WebLOAD to monitor and collect statistics from the application
server, database, Web server and stream technologies. Click Measurements Manager to
configure the performance statistics to monitor and display in real-time during the test
session.
The Performance Measurements Manager opens. For more information on configuring the
statistics to monitor, see Performance Measurements Manager (on page 365).
3. Click Next.
WebLOAD progresses to the Mix Definition dialog box. See Creating a Mix (on page 139).
Creating a Mix
A Mix is a set of Agendas, each performing different activity, to simulate groups of users
performing different activities on the ABT at the same time. If you selected Use an Existing Mix
on the Agenda / Mix Selection screen, the selected mix appears on the screen, enabling you to
modify the Mix. You can also create a new Mix.
Saving a Mix
After defining your mix you must save it as a *.mix file.
To save a mix:
1. Click Save or Save As.
2. In the File name field, type a descriptive name for the Agenda and then click Save.
The mix configuration is saved with the extension .mix, and the wizard returns to the Mix
Definition dialog box.
Load Machines
By selecting a group of Load Machines, you define the pool of computers WebLOAD uses to
generate the Virtual Clients. The number of Virtual Clients that can be generated on each host is
dependent on the power of each host machine.
Probing Client Machines
Probing Clients are single instances of Virtual Clients running WebLOAD while the Load
Machines bombard your Web application. WebLOAD reports exact measurements for Probing
Clients.
This dialog box has three windows. The large window on the left, entitled Host Computers,
displays all the host computers defined for testing. The two windows to the right contain the
systems defined as Load Machines and Probing Client Machines. Use the arrow buttons between
the windows to change the roles of the host computers.
WebLOAD User's Guide 141
Selecting Host Computers
Notes:
In Standalone Workstation mode only the local machine is available to act as a Load Machine
At least one system must be configured as a Load Machine to continue working with the wizard.
After configuring the hosts participating in the test session, the WebLOAD Wizard progresses to
the Schedule dialog box. See Scheduling the Load Session (on page 144).
2. Insert the host computer name or IP address in the Host Name / IP field.
3. Click OK.
The host is added to the Host Computers window.
For defining host computers as Load Machines and Probing Clients, see Designating a Host
Computer as a Load Machine (on page 142), and Designating a Host Computer as a Probing
Client Machine (on page 143).
2. Press the single arrow key adjacent to the Load Machine window.
The selected host moves to the Load Machines window.
Press the double arrow key adjacent to the Load Machines window.
All of the host computers move to the Load Machines window.
2. Press the single arrow key adjacent to the Probing Client Machines window.
The selected host moves to the Probing Client Machine window.
3. After assigning all Load Machines and Probing Client Machines, click Next.
The wizard progresses to the Schedule screen. See Scheduling the Load Session (on page
144).
Manually defining time frames and the load to generate over each frame
Using the Load Profiler
A combination of both
Note: Days are numbered 00 through to 99. Following them are hours:minutes:seconds.
Seconds are numbered 1 through 59. To enter one minute, use the 1:00 syntax not 00:60.
2. Enter the number of days and the ending time to generate the load in the To field.
The ending time is defined for each particular Agenda relative to the beginning of the test
session.
If the To field is left blank, WebLOAD continues generating the load until the From time in
the next entry for this Agenda, or, if this is the last entry for this Agenda, until the test
session is stopped manually.
3. Under Load Size enter the number of Virtual Clients to generate during the time frame
defined in steps 1 and 2. The load is distributed equally among the Load Machines defined
for the Agenda.
4. (Optional) Schedule additional Virtual Client loads for this Agenda:
3. Select a schedule pattern from the Scheduling model field at the top of the dialog box. For
an explanation of each model, see Scheduling Models (on page 150).
4. Fill in the fields at the bottom of the dialog box.
5. Click Append to include the load definition in the schedule,
-Or-
Click Replace to substitute the load definition for the one appearing in the Schedule dialog
box.
Note: Days are numbered 00 through to 99. Following them are hours:minutes:seconds.
Seconds are numbered 1 through 59. To enter one minute, use the 1:00 syntax not 00:60.
3. Enter the number of days and the ending time, to generate the load in the To field.
The ending time is defined for each particular Agenda relative to the beginning of the test
session. If the To field is left blank, WebLOAD continues generating the load until the From
time in the next entry for this Load Machine, or if this is the last entry, until the test session
is stopped manually.
4. In the Load Size field enter the number of Virtual Clients to generate during the time frame
defined in steps 2 and 3.
5. (Optional) Schedule additional Virtual Client loads for this Load Machine:
2. Select a pattern from the Scheduling model field at the top of the dialog box.
For an explanation of each model, see Scheduling Models (on page 150).
3. Fill in the fields at the bottom of the dialog box.
4. Click Append to include the load definition in the schedule,
-Or-
Click Replace to substitute the load definition for the setting in the Schedule dialog box.
Scheduling Models
WebLOAD includes numerous scheduling models to provide a wide variety of options and aid
you in quickly achieving a schedule that meets you testing goals.
The linear model increases the load with linearly increasing values.
Using this model you need to define:
Total time in minutes - the number of minutes to generate the load using this model.
Starting Load Size - the load size to begin from.
Concluding Load Size - the load size to end with.
For example, using this model the load will start at the value defined in the Starting Load Size
field and increase the load (automatically calculating the intervals when the load should
increase) over the time specified in the Total time in minutes field to the load size defined in the
Concluding Load Size field.
Random
relationship or trend between each interval until the time defined in the Total time in minutes
field has expired.
Incrementing Intervals
The incrementing intervals model generates an increasing load in intervals while returning to the
base load between intervals.
Using this model you need to define:
Total time in minutes - the number of minutes to generate the load using this model.
Base Load Size - the load size to begin from, and to return to between each interval.
Time Between Each Interval-T1 - the amount of time to wait before increasing the load for
each interval.
Time of Each Interval-T2 - the length of time for each interval.
Load to Increase Each Interval - the number of users to add each interval (over the load
generated in the last interval).
For example, using this model the load will start at the Base Load Size, increase for the time
defined as T2, return to the base load and then increase again. Each subsequent increase will be
the Load Size from the previous interval plus the increment.
The incrementing intervals (time calculate) model generates an increasing load in intervals while
returning to the base load between intervals. The schedule stops when the concluding load size
is achieved.
Using this model you need to define:
Concluding Load Size - the load size to end with.
Base Load Size - the load size to begin from, and to return to between each interval.
Time Between Each Interval-T1 - the amount of time to wait before increasing the load for
each interval.
Time of Each Interval-T2 - the length of time for each interval.
Load to Increase Each Interval - the number of users to add each interval (over the load
generated in the last interval).
For example, using this model the load will start at the Base Load Size, increase for the time
defined as T2, return to the base load and then increase again. Each subsequent increase will be
the Load Size from the previous interval plus the increment until the concluding load size is
achieved. The time it takes to complete the task is calculated automatically by WebLOAD based
on the parameters defined.
Step Increments
The step interval model generates an increasing load in intervals until a defined load size is
reached. The time to complete the scheduled model is calculated automatically by WebLOAD
based on the parameters defined.
Using this model you need to define:
Concluding Load Size - the load size to end with.
Starting Load Size - the load size to begin from.
Time of each Interval - the length of time for each interval.
Load to increase each interval - the number of Virtual Clients to increase each interval.
For example, using this model the load will start at the value defined in the Start Load Size field,
and increase gradually the amount of user defined in the Load to increase each interval field until
the concluding load size is achieved.
Ramp Up
The ramp up model generates an increasing load over a set time frame to until reaching a
defined maximum load size. The maximum load size is then run for a set period and then the
load is reduced back to the base load again over a time frame defined by the user. The load
increments are calculated automatically by WebLOAD based on the parameters defined.
Using this model you need to define:
Base Load Size - the load size to begin with.
Max Load Size - the maximum number of Virtual Clients to generate using this model.
Ramp UpTime - the time interval over which to increase the load in order to reach the max
load size.
Time to Run Max Load Size - the time frame to run the load at the max load size.
Ramp Down Time - the time interval over which to decrease the load in order to return to the
base load size.
For example, using this model the load will start at the value defined in the Base Load Size field,
and increase over the time frame defined in the Ramp Up Time field to reach the load size
defined in the Max Load Size field. The test will then continue running with the max load size for
the period of time defined in the Time to Run Max Load Size field after which, load will return to
the base load size value over the time frame set in the Ramp Down Time field.
User Defined
The user-defined model generates the load based on values extracted from a data file. The time
to complete the scheduled model is defined by the user as is the minimum and maximum load
size.
Using this model you need to define:
Total time in minutes - the number of minutes to generate the load using this model.
Min Load Size - the absolute minimum load size to generate over the schedule. At no point in
the schedule should the load size dip below this value.
Max Load Size - the absolute maximum load size to generate over the schedule. At no point
in the schedule should the load size raise above this value.
Data file name - the name and location on the data file to read values from for generating the
load.
Delimiter - the delimiter used in the data file from which values are read to generate the load
size.
Using this model the TestView Scheduler will read values from the assigned data file and
generate the load based on the values in the file. The load size will fluctuate based on the
number of values in the file and the time frame for which the schedule is set to run.
Note: If after closing the WebLOAD wizard you want to edit parameters in the Host Selection
dialog box, select Session Control | Modify Host Selection.
WebLOAD offers the Cruise Control Wizard, a unique feature that enables goal-seeking
performance testing.
In This Chapter
Using Cruise Control .................................................................. 161
How does Cruise Control Work?................................................ 162
The Cruise Control Wizard Workflow ....................................... 163
Opening the Cruise Control Wizard ........................................... 163
Selecting an Agenda or Mix ....................................................... 165
Running Single Agendas ............................................................ 166
Running a Mix of Agendas......................................................... 167
Selecting Host Computers .......................................................... 172
Defining Performance Goals ...................................................... 175
Defining the Acceleration Rate................................................... 182
Defining WebLOAD's Action Once the Goals are Reached ...... 183
Finishing the Cruise Control Wizard .......................................... 184
Viewing the Cruise Control Default Report ............................... 186
performance falls below the user-defined profile. Cruise Control incorporates intelligent,
distributed load generation capabilities for virtually unlimited load generation.
Using the Cruise Control wizard to create a load template, you define:
The Agenda(s) to run.
The load machines on which the load is generated.
The performance goals to achieve.
You can also specify Agenda options including the type of browser to emulate, connection
speed, and playback sleep time options.
Note: You cannot use the Cruise Control wizard if you have not yet recorded (or otherwise
created) any Agendas. If you do not have any Agendas to work with, you must exit the wizard
and create an Agenda using one of the authoring tools, such as WebLOAD IDE.
Click the Cruise Control Wizard icon from the Console toolbar.
2. Click Next.
The Cruise Control Wizard progresses to the Agenda Type screen.
To run an agenda:
1. Select the source of the Agenda to be used for this test. Press the Browse button and choose
an existing Agenda from your system.
2. (Optional) To open and view or edit the selected Agenda, press View.
WebFT opens displaying the selected Agenda.
3. (Optional) To configure runtime options exclusive to this Agenda, press Options.
The Agenda Options dialog box opens enabling you to define the runtime options for the
selected Agenda. For more information on Agenda options, see Setting Agenda Options (on
page 221).
166 RadView Software
Creating Load Templates with Cruise Control
Mix
Host Selection
Definition
Select Mix
Browse
File
Using Mix you can prepare a combination of Agendas, each performing a specific activity. Then
you can tune your testing by changing the combination of Agendas, and changing the weight of
each Agenda in the Mix.
Combining Agendas through a Mix can simulate different groups of users performing different
activities on the ABT at the same time. Using a Mix is another way you can create test scenarios
emulating real-life behavior.
Select Use an Existing Mix and click the browse button to select a *.mix file
from your system.
Select Create a Mix to configure a new mix.
2. (Optional) Open the Performance Measurements Manager and configure WebLOAD to
collect performance statistics from your application server, database, Web server and stream
technologies while the test is running. Click Measurements Manager to configure the
performance statistics to monitor and display in real-time during the test.
The Performance Measurements Manager opens. For more information on configuring the
statistics to monitor, see Performance Measurements Manager (on page 365).
3. Click Next.
WebLOAD progresses to the Mix Definition dialog box. See Defining a Mix (on page 169).
Defining a Mix
A Mix is a set of Agendas, each performing different activity, to simulate groups of users
performing different activities on the ABT at the same time. If you selected Use an Existing Mix
on the Agenda / Mix Selection screen, the selected mix appears on the screen, enabling you to
modify the Mix. You can also create a new Mix.
Creating a Mix
To create a mix:
1. Click Add Script.
The Open dialog box appears, allowing you to select the Agenda to add to the mix.
2. Select an Agenda from your file system and click Open.
The Agenda is added to the mix.
3. (Optional) Select the Agenda and click Options to customize the settings (such as the
browser to emulate, connection settings and playback sleep options) for only the current
Agenda without effecting settings in other Agendas.
The Agenda Options dialog box opens. For information on setting the Agenda Options, see
Setting Agenda Options (on page 221).
Saving a Mix
After defining your mix you must save it as a *.mix file.
To save a mix:
1. Click Save or Save As.
2. In the File name field in the Save As dialog box, type a descriptive name for the Agenda and
then click Save.
The mix configuration is saved with the extension .mix.
3. Click Next.
WebLOAD displays the Host Selection dialog box. See Selecting Host Computers (on page
172).
Hosts are the computer systems participating in the test session. In this dialog box you define the
hosts to participate in the Cruise Control test. Hosts can act as Load Machines or Probing Client
Machines.
Load Machines:
By selecting a list of Load Machines, you define a pool of Hosts that WebLOAD uses to
generate a load until the Cruise Control goal is reached. To reach your defined goal, WebLOAD
may need to generate a large number of users accessing the ABT. The number of users that can
be generated on each host depends on the power of each host machine. Defining many machines
as Load Machines allows for the simulation of more users throughout the Cruise Control test.
Note: In Standalone Workstation mode only your local system is available as the Load Machine.
172 RadView Software
Creating Load Templates with Cruise Control
2. Insert the host computer name or IP address in the Host Name / IP field.
3. Click OK.
The host is added to the Host Computers dialog box.
For defining Host Computers as Load Machines and Probing Clients, see Designating a Host
Computer as a Load Machine (on page 173) and Designating a Host Computer as a Probing
Client Machine (on page 174).
Note: If you define a Probing Client host, the Probing Client Agenda Selection dialog box
displays.
1. Press the double arrow key adjacent to the Load Machine window.
All of the host computers move to the Load Machine window.
Note: At least one system must be configured as a Load Machine to continue working with the
wizard.
2. Press the single arrow key adjacent to the Probing Client Machines window.
The selected host moves to the Probing Client Machine window.
The Goal Definition dialog box enables you to define the performance goals for the Cruise
Control test. The defined goals are displayed at the top of the dialog box where you can see at a
glance all of the participants in the test, the measurements being tested and the goals to be
achieved.
On the bottom of the dialog box define the (OR/AND) conditions of the goals where you specify
whether at least one or all of the goals must be met. The buttons on the lower half of the dialog
box are used for defining and editing these goals.
Source The source you are tracking to determine when your goal is
reached. The source can be either:
A Probing Client Machine
A Load Machine
The source for Load Machines is “Total” because your results are
the average of all Virtual Clients. The Source for Probing Client
Machines is the name of the Probing Client.
Measurement The measurement or activity that you are testing such as:
Response Time
Hits Per Second
Round Time
Operator The operator is the condition of the test. This may be either >= or
<= depending on the measurement being tracked. Here you
indicate whether the results should be greater than, or less than
the desired value for the goal to be reached. For example, if the
measurement being tracked were Response Time, you would
choose the operator >= to achieve the goal when the time needed
to receive a response from the server exceeds the limit for this
goal.
Value The limit that determines when this goal is achieved.
Report When this field is checked, the default Integrated Report includes
the measurement for which the performance goal is defined. The
Report field is selected by default.
Operator
Below the Goal Table are the Operator radio buttons. When more than one goal is defined, the
Operator enables you to establish the conditions under which the goal will be achieved.
Select the OR radio button to require only one expression to be true for the goal to be
achieved.
Select the AND radio button to require all expressions to be true for the goal to be achieved.
Control Buttons
Below the Operator radio buttons, find the following control buttons:
Button Description
Add Goal Open the Add Goal dialog box to add a new measurement to the test
goals.
Delete Delete a test goal.
Defaults Open the Default Goals dialog box. For more information on using
Defaults, see Setting and Using Defaults (on page 180).
Set As Defaults Set the current goals as default measurements and save them for
future use.
Note: A gray circle next to a measurement indicates that no default subcomponents are
defined (the upper level component cannot be selected). To select subcomponents, click on
the + to expand the tree.
3. Click OK.
WebLOAD closes the list and opens a row for each of the selected measurements in the
Goal Definition dialog box.
4. Click in the Op field and select an operator from the drop-down list.
5. Click in the Value field and type a boundary value for the goal.
6. Ensure that the Report field is checked to include the goal measurement in the Cruise
Control default Integrated Report.
7. Click Next.
WebLOAD displays the Acceleration dialog box. See Defining the Acceleration Rate (on
page 182).
Deleting a Goal
To delete a goal:
1. Select the goal to be deleted on the Goal Definition screen by clicking on the number to the
left of the goal.
2. Click Delete.
The goal is removed from the Goal Definition dialog box.
Setting Defaults
The Set As Default button sets all of the Goals located on the Goal Definition dialog box as
defaults. This operation overrides any previous default configurations.
To set defaults:
1. Configure the test goals that you want to save as the default.
2. Click Set As Defaults.
Using Defaults
To use defaults:
1. On the Goal Definition dialog box, click Default.
The Goal Defaults dialog box appears.
2. Select the goal to be included in the Cruise Control test session by selecting the number to
the left of the goal (use the Ctrl or Shift keys to select more than one goal).
3. Click OK.
The goal(s) are added to the test definition.
Note: The number of users that WebLOAD can generate on any given Load Machine in an
efficient manner (so that the addition of users will not reduce the performance on that
machine) is dependent on the power of the host machine. When WebLOAD estimates that it
has reached the maximum output on a specific host, a log message is sent. WebLOAD
continues to accelerate the load until the end of the Cruise Control test session.
Cruise Control can be set to stop, to continue testing under the same load, or to continue
accelerating the number of Virtual Clients.
Real-time results are displayed in graphs on the right side of the Console screen.
The following elements are displayed on the screen:
Elements Description
Menu Bar Contains the menus for configuring and controlling the Cruise
Control test session.
Toolbar Contains buttons for controlling the Cruise Control test session.
Gauges Display the Load Size and Rounds Per Second during the Cruise
Control test session.
Results Window All of the reports opened during the Cruise Control test session
will appear in this window. Use the tabs located at the top and
bottom of the window to view different reports.
Elements Description
Session Tree Presents a graphical display of your Cruise Control test session.
Status Bar Indicates the program status. When running with the Resource
Manager the Status Bar displays the current mode, Standalone
Workstation mode or Connected Workstation mode.
The Cruise Control Session continues running until the performance goals are achieved. When
the performance goals are reached, WebLOAD displays a message box with the test results.
The following is an example of such a message box:
Click Details to examine the original goal settings and statistics for them. A complete report
displayed as follows:
Note: You can view this goals message box while the test is running by right clicking on the load
session icon in the session tree.
This chapter describes the process of manually configuring a load template. Load templates are
configured in the Console through a sequential series of dialog boxes. Through these dialog
boxes, you can select and schedule the Agenda or Mix files to run during the test, define the
number of Load Machines and Probing Clients to run each Agenda and the number of Virtual
Clients to run.
Note: You cannot create load templates if you have not yet recorded (or otherwise created) any
Agendas. If you do not have any Agendas to work with, you must create an Agenda using one of
the authoring tools, such as WebLOAD IDE.
In This Chapter
The Manual Test Configuration Workflow ................................ 190
Starting a New Template Manually ............................................ 190
Selecting an Agenda ................................................................... 191
Selecting Hosts ........................................................................... 193
Scheduling the Test Session ....................................................... 195
Completing Your Manual Load Template Configuration........... 198
Manually Schedule
Session
Load Machine /
Agenda Selection Host Schedule Session Finish
Selection
Make Equal
2. Click OK.
The WebLOAD Console displays.
Selecting an Agenda
Agendas are test scripts that describe user activity and are run during WebLOAD test sessions.
WebLOAD supports two types of Agendas:
Single Agendas, which are lone test-scripts.
Mix of Agendas, which are a combination of existing single Agendas. Using a mix you can
simulate different groups of users performing different activities on the application being
tested (ABT) at the same time.
Select the Agendas to run through the Agenda / Mix Selection dialog box.
a. Click in the Agenda Path field and enter the name and the full path for the Agenda.
The path name can be entered manually, or through browsing the directory structure.
Selecting Hosts
Hosts are the computer systems participating in the test session. Each host has an assigned task,
acting as either a Load Machine or a Probing Client. Both Load Machines and Probing Clients
are configured in the same Host Selection window. Hosts must be defined for each Agenda or
Mix file running in the test session.
In this dialog box, you define the host computers participating in the test. There are three
windows in this dialog box. The large window on the left, entitled Host Computers, lists all
of the host computers available for testing. The two windows to the right contain the
systems defined as Load Machines and Probing Client Machines. Use the arrow buttons
between the windows to change the roles of the host computers.
Note: In Standalone Workstation mode only your local machine is available to act as a Load
Machine.
2. Enter the name or IP address of the host computer you want to add.
3. Click OK.
WebLOAD adds the machine to the Host Computer list.
2. Press the single arrow key adjacent to the Load Machine window.
The selected Host moves to the Load Machine window.
3. Click OK.
1. Press the double arrow key adjacent to the Load Machine window.
All of the Host Computers move to the Load Machine window.
2. Click OK.
2. Press the single arrow key adjacent to the Probing Client Machines window.
The selected Host moves to the Probing Client Machine window.
3. Click OK.
Manually defining time frames and the load to generate over each frame.
Using the Load Profiler.
A combination of both.
2. Expand the Load Machine tree using the + / - button adjacent to the Load Machine you want
to schedule.
3. Enter the number of days and the starting time to generate the load in the From field.
The starting time is defined for each particular Agenda relative to the beginning of the test.
Note: Days are numbered 00 through to 99. Seconds are numbered 1 through 59. To enter
one minute, use the 1:00 syntax not 00:60.
4. Enter the number of days and the ending time to generate the load in the To field.
The ending time is defined for each particular Agenda relative to the beginning of the test. If
the To field is left blank, the WebLOAD continues generating the load until the From time in
the next entry for this Load Machine, or if this is the last entry, until the test is stopped
manually.
5. Enter the number of Virtual Clients to generate during the time frame defined in steps 3 and
4.
6. Schedule additional Virtual Client loads for this Load Machine:
4. Select a pattern from the scheduling model field at the top of the dialog box.
5. For an explanation of each model, see Scheduling Models (on page 150).
6. Fill in the fields at the bottom of the dialog box.
7. Click Append to include the load definition in the schedule,
-Or-
Click Replace to substitute the load definition for the one appearing in the Schedule
Manually dialog box.
In This Chapter
Session Options........................................................................... 201
Session Options
The session option material shows you how define test session options and introduces you to
some of WebLOAD's more advanced features. This material is divided into three chapters:
Setting Global Options (on page 203) guides you through the process of setting WebLOAD
global options.
Setting Agenda Options (on page 221) guides you through the process of setting WebLOAD
default Agenda options, individual Agenda options and session Agenda options.
Functional Verification Testing (on page 259) details how to set WebLOAD to perform
Functional Verification tests on your Web application.
Global Options enable you to modify WebLOAD defaults to your specifications. All WebLOAD
tests use these defaults.
In This Chapter
How to Set Global Options......................................................... 203
The Notification Manager........................................................... 215
Goals reached delay - the number of samples required to declare the Cruise Control goals as
achieved. Cruise Control declares the defined goals reached only after receiving the desired
goal results in several samples.
Acceleration point measurements - the number of sample measurements to read before
automatically accelerating the load if the Cruise Control goal has not been reached by then.
reflect a sampling every 40 seconds. They are the measurements gathered at the intervals 20, 60,
100, 140…
This option is useful in cases where a large amount of data results from a long test session.
When the Open viewer when exporting option is enabled (default), WebLOAD launches the
program which serves as the default viewer (e.g., Notepad, Excel) for an exported Tab file. The
default viewer is defined in the Tab File default viewer field (see below).
These options define the Tab File that can be exported to other programs.
When the Open viewer when exporting option is enabled (default), WebLOAD launches the
program, which serves as the default viewer.
These options define the HTML files that can be exported to other programs.
3. Select the item whose default location you want to change, and click Modify.
The Modify Location dialog box opens.
4. Select a new default location and click OK.
The dialog box closes and the new location is listed.
Note: If you disable the Collect Instances checkbox, the Data Drilling options of the Functional
Testing tab are also disabled.
3. Under Automatic Load Generator Spawning, click Enable Load Generator Spawning (default).
4. Type or select the number of threads at which WebLOAD should automatically spawn Load
Generators for the appropriate operating systems.
5. Click OK.
Notes:
Only client threads running within a single spawned process, on the same Load Generator, are
able to share global variables. So if, for example, you have spawning set to 100 and you are
running a total of 300 threads, realize that you are actually running three spawned processes on
three separate Load Generators.
Only client threads running within a single spawned process, on the same Load Generator, can
be synchronized.
3. Select the tracking system to use. If you select a generic tracking system enter the location
to save the defects.
2. Click the Add button to define a new rule or condition under which someone should receive
notification.
The Add Notification Rule and Action dialog box appears.
3. Create a new Mail Notification Rule by entering the following information into the Rule
Definition input text boxes in the Add Notification Rule and Action dialog box.
a. Select the measurement to be monitored from the list of built-in measurements that
appears in the drop-down Measurement field, illustrated in the following figure:
b. Select the operator from the list of operator options that appears in the drop-down
Operator field, illustrated in the following figure:
c. Specify the threshold value in the input text box of the Value field.
4. Define the parameters of the email notification to be sent out if the Rule conditions are met.
Enter the email information in the Action Definition section of the Add Notification Rule
and Action dialog box.
a. Define the official sender of the email in the input text box of the From Mail field.
b. Define the official recipient of the email in the input text box of the To Mail field.
c. Define the email subject line in the input text box of the Subject field.
d. Type the message text body in the input text box of the Message Body field.
5. Click OK to return to the main WebLOAD Notification Manager dialog box.
6. To send an additional email notification when the test is completed, check the End of Test
Notification checkbox in the main WebLOAD Notification Manager dialog box and click on
the Configure button.
The Action Definition dialog box appears.
7. Define the parameters of the email notification to be sent out when the test is completed.
a. Define the official sender of the email in the input text box of the From Mail field.
b. Define the official recipient of the email in the input text box of the To Mail field.
c. Define the email subject line in the input text box of the Subject field.
d. Type the message text body in the input text box of the Message Body field.
Agenda Options enable you to modify Agendas to your specifications, and define the way
WebLOAD behaves during runtime. Setting Agenda options allows you to emulate different
types of user activity. For example, you can set WebLOAD to emulate users connecting to your
application at different modem speeds.
Note: Agenda options, set through the Console, are not uploaded when editing the Agenda in
WebFT. Debugger options are available (in WebFT) for debugging the Agenda in WebFT.
In This Chapter
About Agenda Options ............................................................... 222
Modifying Agenda Options ........................................................ 224
Setting the SSL Options.............................................................. 225
Client Type Settings.................................................................... 225
Setting Sleep Time Playback Settings ........................................ 227
Setting the Browser Cache.......................................................... 229
Configuring Browser Parameter Settings ................................... 231
Setting the HTTP Parameters...................................................... 237
Setting Authentication ................................................................ 240
Connection Speed Settings ......................................................... 242
Diagnostic Settings ..................................................................... 243
Functional Testing Options......................................................... 247
Pass / Fail Definition Settings..................................................... 252
Configuring Reset Mechanism Settings...................................... 254
Configuring Emulation Settings ................................................. 256
Web Services Settings ................................................................ 258
Note: To modify Agenda options, you must be in Edit mode (select File | Edit Template, or click
Right-click on the root node (Load Session node ) and select Default Options.
Current Session Options- It is possible to modify Agenda options for all Agendas
participating in a specific load session or part of a specific load template without changing
the default settings for all sessions.
To activate this mode:
Select Tools | Current Session Options,
-Or-
Right-click the root node (Load Session node ) and select Current Session
Options.
Modify the defaults as desired. All participating Agendas currently listed in the Session
Tree are automatically assigned the modified default values.
Agenda Options- It is possible to modify Agenda options for an individual Agenda within a
new or existing load session or load template.
To activate this mode:
Select Tools | Agenda Options,
-Or-
Right-click the desired Agenda in the Session Tree and select Agenda Options.
The modifications you make here only affect that specific Agenda.
For Current Session Options and Agenda Options, WebLOAD automatically identifies the Client
type (HTTP Client or Browser Client) which is associated with the Agenda file type and selects
the right engine for execution.
3. Check SSL Bit Limit and select the highest SSL Bit Limit available between the Virtual
Clients and the server.
4. Click OK to accept the modifications, or Cancel to reject them, and exit the Default / Current
Session Options dialog box.
transactions in as short a time as possible to see how your application copes when pushed to the
limit and beyond. And sometimes, you don't want to take any extra time or resources to measure
anything specific, you just want to see how your Web application behaves for a typical user
under standard operating conditions.
Multiple Client Type modes provide the ability to test your application using greater or lesser
load to ensure optimal WebLOAD performance by defining the behavior of clients according to
your specific requirements. The advantages of the different Client Type modes are explained in
Getting Started (on page 21).
Notes:
1) This option is not available for individual Agendas. It can only be set for the entire session
though the Default Options dialog boxes.
2) The Client Type dialog box is disabled for Current Session Options.
3. From the Load Machine Type drop-down list, select Load Generator or Probing Client.
4. Select the type of cache by clicking one of the following: 5.0 Cache, Browser Cache, or
None.
5. If you selected Browser Cache, do the following:
a. Select a scope by clicking one of the following: Per Request, Per Round, or Global.
b. Select a filter by clicking one of the following: Default or User Filter.
If you selected User Filter, the filter checkboxes are enabled.
c. If you selected User Filter, select one or more User Filter checkboxes.
6. Click OK.
The following table describes the fields and buttons in the Browser Cache dialog box.
Field Description
5.0 Cache All Virtual Clients simulate the browser cache the same way
that WebLOAD 5.0 used to:
1. Only images were cached.
2. The content in the cache was not stored, just an indication
that the image was read.
3. Whenever the engine had a request for a cached image, it
simply sent nothing to the server.
4. If the same image appeared more than one time in the
same page, it would always be cached only once, even if the
cache is turned off. Of course, it would be cached in further
pages if the cache was turned on.
Browser Cache Stores the most recent downloaded document in a virtual
browser cache. This speeds up document accesses, just as
real browser cache speeds up access time. The browser cache
is automatically cleared at the end of each round.
WebLOAD behaves as follows:
1. WebLOAD can cache anything.
2. WebLOAD stores actual HTML pages except for other
resources, which is only an indication.
3. Whenever the engine has a request for a cached resource,
it will send the request with an "if-modified-since" header. If
the server responds with a 200 status, the engine will refresh
the cache.
4. If the same resource appears more than one time in the
same page, it will be requested with an "if-modified-since"
header.
Scope Defines when the cache will be cleared.
Per Request Defines that cache be cleared after each request.
Note:
Cache will be cleared after each request on the Agenda level.
It will not be cleared after internal requests.
Per Round Defines that cache be cleared after each round.
Field Description
Global Defines that the cache will never be cleared. Each client
maintains its own cache by the way and this is true for the
old cache as well.
Cache Content Filter The type of content to filter.
Default Select to use the default, which is that WebLOAD will cache
only images and static pages.
User Filter You can specify which content types to filter.
Note:
If you selected User Filter and do not select any of the
content types, then WebLOAD will cache all content types.
None All Virtual Clients simulate a browser with no available
cache.
Note: The script in the Agenda (coded to account for a particular option) always overrides the
Browser Parameters options. WebLOAD recommends always setting these options through the
Console user interface.
can specify any specific application by supplying a custom user-agent that is included in all
HTTP headers. By default, all Virtual Clients use the WebLOAD Default browser agent.
4. You can select an alternative user-agent from the drop-down list, or click Update to add
additional user-agents. See Adding a User-Agent (on page 233).
If you select WAP as the browser type, the user-agent field changes to WAP-emulation.
When WAP-emulation is selected, WebLOAD automatically sends all necessary headers,
including the screen size.
Note: For the WebLOAD 7.x protocol engine, the default User Agent is Mozilla/4.0
(compatible; MSIE 6.0; Windows NT 5.0). For the WebLOAD 5.x protocol engine, the default
HTTP version was Mozilla/4.0 (compatible; MSIE 5.01; Windows NT; DigExt).
Adding a User-Agent
The available user-agent list is appropriate for the browser type you select. You can add to or
delete from the user-agent list.
To add a user-agent:
1. Click the Update button beside the user-agent field on the Browser Parameters tab.
The Add User Agent dialog box opens.
2. Type or select the browser type to which you want to add a user-agent.
3. Type or select the new user-agent.
4. Click Add.
Deleting a User-Agent
To delete a user-agent:
1. Click the Update button beside the user-agent field on the Browser Parameters tab.
The Add User Agent dialog box opens.
2. Type or select the browser type from which you want to delete a user-agent.
3. Type or select the user-agent to be deleted.
4. Click Delete.
4. In the Redirection Limit box, type or select the desired redirection limit.
The default limit is 10.
However, some HTTP servers may refuse to keep a connection alive. You should not keep a
connection alive if establishing the connection is part of the performance test.
3. Set the HTTP version by clicking HTTP Version 1.0 or HTTP Version 1.1.
Note: For the WebLOAD 7.x protocol engine, the default HTTP version is 1.1. For the
WebLOAD 5.x protocol engine, the default HTTP version is 1.0.
Field Description
Field Description
HTTP Properties
Multi IP Support Sets the wlGlobals.MultiIPSupport flag.
GZip Support Sets the wlGlobals.AcceptEncodingGzip flag.
WebLOAD will behave as follows:
For each request, sends the header "Accept-Encoding:
gzip, deflate". This tells the server that the client can accept
zipped content.
When this header is turned on, the server MAY send a
response with the header "content-encoding: gzip" or
"content-encoding: deflate". If either of these headers is
sent, it means that the response is zipped/deflated and
WebLOAD will unzip/inflate the content.
Notes:
Most servers will work correctly even if the client does not
send the "Accept-Encoding: gzip, deflate" header.
Therefore, unless needed, it is recommended not to set the
wlGlobals.AcceptEncodingGzip flag because it is
performance heavy. However, some servers will fail if it is
not sent. Microsoft Internet Explorer/Mozilla sends it by
default.
You can manually code the agenda to send the "Accept-
Encoding: gzip, deflate" header even if the
wlGlobals.AcceptEncodingGzip flag is not set. In this
case, if the server returns zipped content, WebLOAD will
not unzip it, so the content will not be available for
WebLOAD to function properly.
Encode Form data Sets the wlGlobals.EncodeFormdata flag.
In general, when an HTTP client (Microsoft Internet
Explorer/Firefox or WebLOAD) sends a post request to the
server, the data must be HTTP encoded. Special characters like
blanks, ">" signs, and so on, are replaced by "%xx". For
example, space is encoded as "%20".
This encoding can be performance heavy for large data, so
WebLOAD allows you to turn it off.
This should ONLY be done if you are sure that the data does
not contain special characters. If it does, and the customer
wants to improve performance via this flag, then you should
replace the special characters within the script or use
wlHttp.EncodeFormdata to set the flag for specific requests.
Field Description
Accept Language Sets the wlGlobals. AcceptLanguage flag. This flag defines a
global value for the "Accept-Language" header which will be
sent with each request. Some applications/servers will behave
differently depending on the setting. It is a simple string and
WebLOAD does not enforce any checks on the values. It is
similar to the wlGlobals.UserAgent property in the sense that it
is a wlGlobals/wlHttp setting that affects the value of request
headers.
Setting Authentication
The Authentication option enables you to define the Global and Proxy authentication settings.
To set authentication:
1. Select Tools | Default/Current Session Options.
2. Select the Authentication tab.
3. Type the appropriate user names and passwords into the corresponding Global
Authentication Settings input text boxes:
Field Description
UserName and PassWord The user name and password the Agenda
should use to log onto restricted HTTP sites.
NTUserName and The user name the Agenda should use for
NTPassWord Windows NT Challenge response
authentication.
ProxyUserName and The user name and password the Agenda
ProxyPassWord should use for proxy servers that require
user authorization.
Field Description
The Variable Connection Speed feature enables you to set the connection speed for the load
and emulate users connecting through different modem connections (14400, 28800…).
3. Select the Load Generator Connection Speed checkbox and type or select the appropriate
speed from the list.
4. Select the Probing Client Connection Speed checkbox and type or select the appropriate
speed from the list.
5. Click OK to accept the modifications, or Cancel to reject them, and exit the Default / Current
Session Options dialog box.
Diagnostic Settings
Diagnostic options can be enabled while developing an Agenda or for tracking problems in
existing Agendas.
Note: Running a load session with the diagnostic options enabled may affect system
performance. Therefore, we recommend that these options remain disabled while running a
load session with multiple threads.
Note: This option is for backward compatibility purposes only, to continue support of scripts that
were developed using the AAT.
4. Click in the Output File Directory field and select the directory for saving Log Report files.
The default directory is: C:\Program Files\RadView\WebLOAD\User\Log.
5. Select which rounds should generate the Log Report in the Start round and End round fields.
The log report displays the time a request was made, the round number, the request number, the
URL received and the response headers. If GetContents = “Yes” the same information is
displayed for each frame.
A message is sent to the Log Window for each request. The message appears in the following
format:
### Thread:000, Round Num:001, Request Num:001 ### URL
Each request is numbered sequentially for each thread and round number.
4. Check Object level and/or HTTP level to specify whether to collect data at object or HTTP
level or both.
Note: For the WebLOAD 7.x protocol engine, the default automatic data collection setting is
Pages Off. For the WebLOAD 5.x protocol engine, the default automatic data collection setting
is Pages On.
4. Specify the maximum amount of time that a page should take to load in the Max Time field.
5. Specify the severity level if the page exceeds the specified time in the Severity Level field.
Note: It is important to specify a meaningful time period. For example, if you specify 3 seconds
for a page which is likely to take longer, numerous essentially meaningless failures will result.
Conversely, if you specify 10 seconds for a page which should take less than 1 second, the test
becomes meaningless.
Note: You may select the Verification checkbox without selecting the Text Search checkbox
and the Advanced Emulation Verifications checkbox in the frame beneath it. It is for
backward compatibility for scripts that were recorded with the AAT and support table
comparisons.
Note: Selecting a lower level checkbox selects and highlights the upper level checkbox.
6. Select the severity level from the drop-down list next to the test. For further information, see
Verification Function Return Values (on page 266).
7. Enter a literal string with the text for which you are searching in the Text field.
8. Choose the type of test to run from the drop down list in the Option field:
Text to Find - the verification test will pass if the selected text is found in the current
Web page.
Text not to Find - the verification test will fail if the selected text is found in the current
Web page.
For example, to ensure that the word "error" does not appear on the page during runtime,
enter the word "error" in the Text field, and select Not to Find. If the text error is found on
the page, WebFT will report the error.
9. Click Add to add another string to the list of Text Search strings.
10. Click Delete to delete a selected string from the list of Text Search strings.
To select a test:
1. Select Tools | Default/Current Session Options.
2. Select the Functional Testing tab.
3. Under Verification, select the Verification checkbox.
The Advanced Emulation Verifications checkbox is enabled.
4. Select the Advanced Emulation Verifications checkbox, and click the [...] button.
The Emulation Verifications dialog box appears.
5. Select one or more individual tests to perform by marking the checkbox adjacent to the test
(for information on each test, see Functional Verification Test List (on page 264)).
6. Select the severity level for each selected test (for information on severity levels, see
Verification Function Return Values (on page 266)).
7. Click OK.
For a full description of the tests in Functional Testing and their results, see Functional
Verification Testing (on page 259).
Note: This option is not available for individual Agendas. It can only be set for the entire session
though the Default Options or the Current Session Options dialog boxes.
You can set WebLOAD to declare a test as failed based on a number of criteria. First, you
must decide if you want the test to be professed a failure after an absolute number of failures
or a relative number of failures. Then you set the conditions upon which a test is declared a
failure.
3. Select Absolute (default) to set the total number of failures required for a test to be declared
failed,
-Or-
Select Relative to set the percent of failed transactions need for a test to be declared failed.
4. Set the conditions upon which to declare a test failed and the action to take if a test fails.
Fail test on severe error - Sends a failure message and declares the test failed in the
event of a severe error.
Fail the test if the number / percent of errors is higher than - Declares the test failed if the
number of errors received is higher than the number stated. If the pass / fail definition is
set to relative then the percentage of errors is considered rather than the absolute value.
Fail the test if the number / percent of warnings is higher than - Declares the test failed if
the number of warnings received is higher than the number stated. If the pass / fail
definition is set to relative then the percentage of warnings is considered rather than the
absolute value.
Check rule only after X minutes of execution - This option is only enabled when the pass
/ fail definition is set to relative. Use this option to set a stabilizing period for the test
when it starts. Errors or warning received during this time frame will not fail the test,
and the rules set for failing the test will only be checked after this timeframe has
expired.
Stop test on failure - Stops the test if the test is declared failed.
5. Click OK to accept the modifications, or Cancel to reject them, and exit the Default / Current
Session Options dialog box.
Note: This option is not available for individual Agendas. It can only be set for the entire session
though the Default Options or the Current Session Options dialog boxes.
3. Select the Enable Load Generator Reset Mechanism checkbox to allow WebLOAD to restart
the Load Generators.
4. Set the number of minutes for Reset Frequency (Every) and Stop Wait (Timeout).
5. Set the number of times for Max attempts to reconnect.
6. Set the number of seconds for Unresponsive LG for more then (Timeout).
Note: The Load Generator Timeout is valid only if the Reset Mechanism is enabled.
7. Click OK to accept the modifications, or Cancel to reject them, and exit the Default / Current
Session Options dialog box.
3. For each type of supported dialog select the action to take if the dialog should appear
unexpectedly during runtime. Define the answer by clicking on the value next to the dialog
box name and selecting a value from the drop down list, clicking or clearing a checkbox or
entering the value in the field box provided.
4. Select Use History to instruct WebLOAD to store previous commands. Set a history limit to
define an upper limit to the number of commands to be saved in a history file.
5. CSJS Recovery is selected by default.
6. Click OK to accept the modifications, or Cancel to reject them, and exit the Default / Current
Session Options dialog box.
3. Enter the appropriate values in the basic and proxy authentication fields.
4. Select a Web Service client.
5. Click OK to accept the modifications, or Cancel to reject them, and exit the Default / Current
Session Options dialog box.
WebLOAD combines unified load, performance, and functional testing into a single process.
allowing you to verify the scalability and integrity of Web applications at the per-client, per-
transaction and per-instance level under user-defined load conditions. By creating and analyzing
the Document Object Model (DOM) for every Virtual Client during a test session, WebLOAD is
able to verify each success and failure and present detailed information to you about each
transaction.
The WebLOAD Functional Verification Test Tool is an easy-to-use tool which automates the
process of inserting verification tests into the Agenda and verifies that the links, images, forms,
tables and frames in your Web application continue to act as expected.
WebLOAD also enables you to insert application level transactions into the Agenda and define
verification criteria, which compare expected outcomes against actual ones. WebLOAD tracks
the success and failure rates of the transactions for each Virtual Client and integrates the data
into the WebLOAD Statistics Report. You can find additional information about functional
testing in the TestView Programmer's Guide.
In This Chapter
The Functional Testing Tab........................................................ 260
User-Defined Transactions and Verification .............................. 264
Verification Function Return Values .......................................... 266
Reporting Events ........................................................................ 267
Tracking Reasons for Transaction Failure.................................. 267
Viewing Verification Test Results.............................................. 269
Click the Functional Verification icon on the Console toolbar to set the verification
tests to run for the current session.
Note: Tests set through the Current Sessions Options override the Default Options settings.
Verifying Text
Use Text verifications to verify the absence or presence of selected text expressions within your
Web page.
Note: Marking a lower level checkbox selects and highlights the upper level checkbox.
5. Select the severity level from the drop-down list next to the test. For further information, see
Verification Function Return Values (on page 266).
6. Enter a literal string with the text for which you are searching in the Text field.
7. Choose the type of test to run from the drop down list in the Option field:
Text to Find - the verification test will pass if the selected text is found in the current
Web page.
Text not to Find - the verification test will fail if the selected text is found in the current
Web page.
For example, to ensure that the word "error" does not appear on the page during runtime,
enter the word "error" in the Text field, and select Not to Find. If the text error is found on
the page, WebFT will report the error.
8. Click Add to add another string to the list of Text Search strings.
9. Click Delete to delete a selected string from the list of Text Search strings.
10. Click OK.
Note: Marking a lower level checkbox selects and highlights the upper level checkbox.
5. Select the one or more checkboxes (for information on each test, see Functional
Verification Test List (on page 264)).
6. Select the severity level for each selected test (for information on severity levels, see
Verification Function Return Values (on page 266)).
7. Select individual tests to perform by marking the checkbox adjacent to the test.
For an explanation of each of the tests available, see Functional Verification Test List (on
page 264).
8. Select the severity level from the drop-down list next to the test. For further information, see
Verification Function Return Values (on page 266).
9. Click OK.
Test Description
Compare Forms Verifies the number of forms on each page, the URL of each form, and the
number of elements per form.
Compare Images Validates the images found on each page by image count or by URL.
Compare Links Checks the number of links on each page and verifies the link URL.
Compare Number Verifies that the number of frames received by the Virtual Clients on each
of Frames page while the test is running is the same as the number of frames found
on each page when the Agenda was recorded.
Compare Tables Verifies the number of tables on each page, as well as the structure of each
table.
Compare Titles Verifies that the page titles received by the Virtual Clients while the test is
running are the same as the “expected” page titles which were recorded
for the same transaction.
WebLOAD tracks all of the results from user-defined transactions and integrates the data into
the WebLOAD Statistics Report, including the timer and counter statistics as well as statistics
on specific failures. The results are then automatically displayed with the other WebLOAD
statistical information.
EndTransaction("UpdateBankAccount",
UpdateBankAccount_VerificationFunction())
function UpdateBankAccount_VerificationFunction()
{
return WLSuccess;
}
/* The user-defined transaction “UpdateBankAccount” ends*/
If during the test the above transaction succeeded 15 times and failed 5 times, and it took an
average of 1.2 seconds for the transaction to execute, the Statistics Report displays as follows:
Reporting Events
WebLOAD enables you to record specific events and transaction instances as they occur. This
information is very useful when analyzing Web site performance with Data Drilling, described
in Data Drilling and Table Compare Reports (on page 335).
Record transaction events with the following syntax:
ReportEvent (“<event-name>”, [“<description>”])
Where:
event-name is a string that identifies the specific event.
description is an optional string providing more information about the specific event.
For example:
Continuing the UpdateBankAccount transaction example described in User-Defined
Transactions and Verification (see "User-Defined Transactions and Verification Syntax" on
page 265), the UpdateBankAccount verification function might appear as follows:
function UpdateBankAccount_VerificationFunction()
{
VerifyUpdateBankAccount () {
if <Condition-success> then
return(WLSuccess)
else
if (document. URL == “../no-login”) then {
SetFailureReason(“User Not Logged”)
return(WLError) }
else
if (document.URL == “...wrong- password”) then {
SetFailureReason(“Wrong Password”)
return(WLError); }
else
return(WLSevereError)
}
Explanation:
Return(WLSuccess) - the actual transaction succeeded, it is added to the
SuccessfulUpdateBankAccount transactions counter.
The SetFailureReason() function accepts a string as a parameter. This string is used
to identify the cause of the transaction failure (“User Not Logged”, “Wrong Password”, etc.).
This data enables tracking of the number of failures that occur for a specific reason, as well
as the time that the failures occurred.
When the reason for failure is “User Not Logged” or “Wrong Password”, the severity level is
Error. Consequently, when the main Agenda resumes control the current round aborts and a
new round commences.
If you mark the transaction as failed, but do not specify any failure reason, the system
registers a “General Failure” which is the default failure severity.
If during the execution of a test the above transaction succeeded 15 times and failed 5, and the
SetFailureReason() function was called 3 times with the string “Wrong Password” and 2
times with the string “User Not Logged”, the Statistics Report appears as follows:
The Log Window displays the following additional information about your test session results:
Message Status - the result and severity of each message.
Messages fall into three categories:
Success / Info Message
Minor Error
Error
Note: The Event Viewer is only available after the test session is complete.
Double click on the description below the viewer to highlight any changes relating to the error.
If the object you are testing for is not found, it is highlighted on the expected page only. If the
object is changed, it is highlighted on both pages.
Data Drilling
When testing the integrity of your Web application, it is important to see not only if a given
transaction succeeded or failed, but also the reasons for the failure to pinpoint the transaction's
exact weak points and bottlenecks. Data Drilling enables you to display a detailed description of
all user-defined and named transactions to the instance level.
With Data Drilling you can view all transaction failures and the reasons for them. Data Drilling
reports provide a complete breakdown of all information about a selected event. Click on an
event entry to bring up detailed information about that event in the Event Viewer. The Event
Viewer provides a graphic illustration of the verification failure event, displaying the “expected”
results, the “actual” results, and a description of the error.
Data Drilling is described in detail in Data Drilling and Table Compare Reports (on page 335).
Running test and analyzing the results introduces you to WebLOAD's performance reports.
In This Chapter
Explanation of Creation and Analysis of Performance Reports . 273
Running a Load Session (on page 275) provides instructions for running a WebLOAD test
session.
Viewing the Log Window (on page 287) describes the WebLOAD Log window.
WebLOAD Performance Reports (on page 293) provides instructions for creating and
viewing WebLOAD Integrated Reports.
WebLOAD Statistics Reports (on page 313) describes the statistics reports provided by
WebLOAD.
Data Drilling and Table Compare Reports (on page 335) describes how to view all
transactions failures and the reasons for them.
Exporting Performance Reports (on page 345) describes how to export all of the data
displayed in the Integrated Reports and Statistics to HTML, Microsoft Excel and to Tab
files.
Performance Measurements Manager (on page 365) describes how to configure the
Performance Measurements Manager to collect data from different sources during the test.
After completing this material, you will have the skills necessary to create and analyze all of the
performance reports provided by WebLOAD.
After configuring a Load Session, you can run your Load Session. This chapter describes the
procedure for running your Load Session, and the way WebLOAD behaves while a Load
Session is in progress.
WebLOAD collects a complete set of test data while a Load Session is running. You can
configure reports that include only the data needed. Reports can be configured during or after a
test.
WebLOAD run commands can start, pause, and stop your Load Session while it is in progress.
These commands are detailed in this chapter.
In This Chapter
Starting a Load Session .............................................................. 275
Establishing Communication...................................................... 278
The Console Screen in Session Mode......................................... 281
WebLOAD Operating Commands.............................................. 282
Throttle Control - Changing the Load Size Dynamically During Runtime
.................................................................................................... 284
For information on scheduling WebLOAD tests with TestView Scheduler, see the TestView
Suite User's Guide.
Note: Verify that the Agenda used with the template specified, and any included files, are
accessible to the Load Template or Load Session file that will be run.
Parameters
When running a test invoked by the executable, you can specify the following parameters:
Parameter Description
File name to open The name of the TPL or LS file (Load Template or Load Session
file) to run.
File name to save The name of the LS file containing the test data. This file will be
saved in the current directory unless otherwise specified.
Flags /ar - auto run
Automatically run the WebLOAD test without waiting for user
input. If this flag is not specified, the Console is called up with the
specified Load Template/Load Session but the test is not
automatically run. The system waits for user input.
Time The length of time (in seconds) that the test is to run.
Parameters are all optional. If no parameters are entered, the executable launches the Console
and does not run a test. If the autorun flag </ar> flag is not set, the <file name to save> and the
<time> parameters are ignored.
Note: If there is a conflict between the time defined in this command and the time defined in the
TestView Scheduler, the load test runs for the shorter of the two periods defined.
Syntax
Use the following syntax to define the parameters for running a WebLOAD test through a
command line interface:
WebLOAD.exe [<file name to open>][<file name to
save>][<flags>][<time>]
Examples
Example 1:
WebLOAD.exe test1.tpl
This command opens the Console and the test1.tpl template. The Console waits for user
input.
Example 2:
WebLOAD.exe test1.tpl march9.ls /ar 30
This command opens the Console and automatically run a test using the test1.tpl template
file. The test results are saved in the Load Session file march9.ls, which includes all of the
test data and results. This file is saved in the current directory, unless otherwise specified. The
autorun flag is set, meaning that the test runs without user intervention. The test will run for 30
seconds.
Establishing Communication
WebLOAD begins executing a test by verifying the test parameters and attempting to
communicate with the hosts participating in the test session.
While WebLOAD is preparing to run the test, the following message appears:
Errors in Communication
The following errors may occur during Load Session startup:
The following table describes the segments of the Console in session mode:
Segment Function
Session Run Toolbar Displays only during a test run. It contains buttons for controlling
the test session.
Session Tree Presents a graphic display of your test session. See Viewing the
Session Tree in Session Mode (on page 282) for a complete
discussion of the Session Tree.
Results Window Displays all of the reports opened during a test session. Use the
tabs located at the top and the bottom of the window to view
different reports.
Log Window Displays all of the error messages recorded during a test session.
You can toggle the Log Window display on/off through the View
menu.
Status Bar Indicates the program status. You can toggle the Status Bar
display on/off through the View menu.
When running the WebLOAD Workstation, the Standalone
Workstation icon or the Connected Workstation icon displays in
the lower right-hand corner.
Note: Using Throttle Control, you can increase the current Load Size to exceed the maximum
Load Size - but only if you are not using Cruise Control.
Resuming a Host that is generating a heavy load causes an immediate drop in system response
time.
The Log Window displays log messages detected by WebLOAD in run time which are
generated by the Console, the JavaScript compiler and any user messages programmed in the
test script.
In This Chapter
Setting the Log Window Options ............................................... 287
Opening the Log Window .......................................................... 288
Stop Session after n log Stop the Load Session after a defined number of log
messages messages have been received.
Once the maximum defined number of log messages is
received, a message box displays stating that the Load
Session is stopped.
Stop sending log messages after Stop displaying log messages in the log window after a
n log messages defined number of log messages have been received.
Once the maximum defined number of log messages is
received, a message box displays stating that no more log
messages are displayed.
Command Description
Command Description
User Note
Warning
Error
Severe
The log messages include messages generated by the Console, the JavaScript compiler and any
human errors or user messages detected by WebLOAD at run-time. You can also add log
messages to the Agenda.
User Note
User Notes are messages added to the Agenda by the user. User Notes can be used in order to
facilitate debugging.
They do not stop an active Load Session.
Warning
A Warning indicates a suspicious condition. The source of the problem may be generated by the
application being tested, the connection to the server, networking, etc.
Warnings do not stop an active Load Session.
Error
When an Error is detected WebLOAD will stop the round and restart the Agenda for the Virtual
Client that encountered the error.
Errors do not stop an active Load Session.
Severe
Severe errors cause the Load Generator, on which the error was generated, to stop immediately.
An error message indicating that the process was stopped is displayed.
The WebLOAD Dashboard lets you view a summary of performance statistics generated during
the test session. Integrated Reports enable you to configure a report that combines
Performance data from the Load Machines and Probing Clients.
Performance data on the ABT, gathered from Performance Monitors.
Existing performance data from a previously saved Load Session with data from the
currently running Load Session.
Combining several different measurements into a single report gives you a more complete
picture of your Web application's performance.
In This Chapter
Viewing Test Results on the WebLOAD Dashboard ................. 294
Opening Reports ......................................................................... 295
Opening an Existing Report........................................................ 298
Editing a Report.......................................................................... 298
Renaming a Report ..................................................................... 299
Deleting a Report........................................................................ 300
Configuring an Integrated Report ............................................... 300
Regression Testing...................................................................... 305
Integrated Report Chart View..................................................... 306
Integrated Report in Spread(sheet) View.................................... 310
Category Description
Time The time during the test that the values displayed on the Dashboard were
sampled.
Running Virtual The total number of Virtual Clients currently running.
Clients
Rounds The total number of times the Agenda has been executed by all Virtual
Executed Clients.
Failed Rounds The total number of rounds with errors.
Failed Hits The total number of hits with errors.
Hits Per Second The number of HTTP requests for information (pages, individual images,
objects, frames) made by Virtual Clients to the ABT per second.
Calculated Hits The number of HTTP request for information made by Virtual Clients to
Per Day the ABT per day.
Pages Per The number of upper level request for information made by a Virtual Client
Second to the application being tested (the ABT) per second.
Calculated Pages The number of upper level request for information made by a Virtual Client
Per Day to the application being tested (the ABT) per second.
Throughput Per The average number of bytes per second transmitted from the ABT to the
Second Virtual Clients running this Agenda.
Calculated The average number of bytes per day transmitted from the ABT to the
Throughput Per Virtual Clients running this Agenda.
Day
Opening Reports
To open reports:
1. Select Reports | Report Manager,
-Or-
2. Click New.
-Or-
Select a report and click Edit.
Report data is organized in a tree structure. Report fields are organized into different “report tree
branches”, illustrated in the preceding figure.
Report tree branches include:
Total-data collected for the session as a whole, including total load size, total number of
rounds, and total number of pages.
AgendaName@LoadGenerator-data collected for all instances of the named Agenda, running
under the named Load Generator. Multiple branches may appear for a single Agenda name,
each one linked to a different Load Generator.
AgendaName:: Total-data collected for all instances of the named Agenda, running under all
Load Generators participating in this test session.
Using the Report dialog box you can:
Configure the items displayed in the test report.
Rename the report.
Import an existing report configuration.
Set a default report configuration.
Apply a default report configuration.
Configure a report that compares existing performance data from a previously saved Load
Session with data from the currently running Load Session.
The following table describes the fields in the Report dialog box.
Button Functionality
2. Select a report.
The Open button is enabled and a description of the measurements included in the report
appears at the bottom of the dialog box.
3. Click Open.
The Integrated Report displays in the Console Results Window. WebLOAD displays the
results of your test in both graphical and tabular format. Click the Chart View and Spread
View tabs at the top of the Results Window to switch between the two views.
Editing a Report
To edit a report:
1. Select Reports | Report Manager,
-Or-
Renaming a Report
You can assign a meaningful name to a report. The default name for reports is “ReportX” where
X = a number that is incremented with each generated report.
To rename a report:
1. Select Reports | Report Manager,
-Or-
3. Click Rename.
The Rename dialog box appears.
Deleting a Report
To delete a report:
1. Select Reports | Report Manager,
-Or-
2. Click New.
The Report dialog box appears.
The full tree of measurements, including all of the items that can be included in the report, are
displayed in the Report dialog box. In addition, items configured to be collected in the Server
Monitor Definition are also displayed. Data for all of the items is accumulated during the test
run.
Select the parameters to view in the Integrated Report from the tree structure displayed in the
Report dialog box. Only the results for the items that are checked are display in the Integrated
Report.
Note: A gray circle next to a measurement indicates that no default subcomponents are
defined (the upper level component cannot be selected). Right-click on a measurement to be
included in the report to prompt the pop-up menu that enables you to select the item.
Select a WebLOAD default configuration by clicking the Apply Defaults button. WebLOAD
selects the set of default measurements.
Select Import All to apply a previously configured Integrated Reports test set, saved as a
Load Template to your current test.
Note: The configuration is imported, even if it was designed for testing a different
application.
Note: If the imported configuration contains a measurement not configured in the current
report, WebLOAD ignores the measurement in the Load Template file. It is not added to the
report.
This function imports the Integrated Report test configuration from the selected Load Template
file. None of the other information in the Load Template file is imported.
Note: If the imported component contains a measurement that is not configured in the
current report, WebLOAD ignores the measurement in the Load Template file. It does not
add it to the report.
This function imports the Integrated Report test configuration from the selected Load Template
file. None of the other information in the Load Template file is imported.
Regression Testing
Regression testing enables you to compare Load Sessions while tests are running, or after the
test session is completed. WebLOAD regression testing reports enable you to visualize, verify
and compare results over time to confirm that results only vary in ways that you expect them to.
Regression testing enables you to configure a report that compares:
Existing performance data from a previously saved or “expected” Load Session with data
from the currently running test session.
Existing performance data from two previously saved Load Sessions.
Combining performance results from several different Load Sessions into a single report enables
you to view, verify and compare results over time and ensure that results only vary in ways that
you expect them to.
To configure an Integrated Report t include data from previously saved Load Sessions:
1. Select Reports | Report Manager,
-Or-
current Load Session it is automatically selected in the Load Session selected to be included
in the report.
6. Click OK.
The report displays in the Console Results window.
To remove a previously saved Load Sessions from the Integrated Report configuration:
1. Select the Load Session you want to remove from the Integrated Report configuration in the
Reports dialog box.
2. Click Remove Session.
The Load Session is removed from the Report dialog box, and any measurements displayed
for that Load Session in the Console Results window are removed.
The report above shows an Integrated Report that includes four measurements. Each is tracked
on a different line graph. The line graphs are color-coded to match the measurements on the
bottom half of the report.
Features of this graph:
The upper part of the report displays a line graph for each measurement tracked in the
report. The X-Axis represents the elapsed time (in seconds) for this test session. The Y-Axis
represents the actual value for each measurement, displayed as a percentage relative to the
defined maximum Scale Value.
Right-click on a colored line graph to view the exact value for the selected point.
The scroll bar at the center of the screen provides you with a quick and easy way to access
points in the test session currently not in focus.
The lower part of the report displays a list of the measurements being tracked for each host
and each session, the current value of the measurement and, in the last column on the right,
the maximum Scale Value set for each measurement.
The Time Range box in the top right-hand corner displays the time scale for the X-Axis. In
this graph, the X-Axis displays a window of 200 seconds out of the total time span of the
Load Session. Using the arrows on the Time Range box, you can change the number of
seconds to achieve the report display you desire.
If you are tracking more than one measurement, the top graph shows multiple graph lines.
To bring up a history focusing on a specific measurement, double click on a single
measurement line in the Integrated Report.
See Integrated Report Toolbar in Chart View (on page 309) for information on the toolbar at the
top of the Integrated Report screen in Chart View.
value of that measurement rather than the true value of a measurement. View the real value in
the display at the bottom of the screen.
Note: The data displayed in Spread View tabular format is the same data displayed graphically in
the Integrated Report Chart View.
Example:
The last column of this table displays all Total Successful Connections Current Slice Sum
(Current Value) values recorded in the current test session. The report resolution is set to 20
seconds. Therefore values are recorded starting at 20 seconds after the test session began and
gathered every 20 seconds after that. This is called a reporting interval.
The Statistics Report is both a real-time monitor that displays the results of your test while you
run it and a report that you can save and export.
WebLOAD collects approximately 30 different statistics during a test. The Statistics Report
displays the values for all of them. You can also create Integrated Reports that display only the
statistics that you choose. For information about Integrated Reports, see WebLOAD
Performance Reports (on page 293).
In This Chapter
Statistical Measurements and Their Values................................ 313
Summary Reports ....................................................................... 323
Total-sums the values for each statistic across all the Agendas running during the test.
AgendaName::Total-sums the values for each statistic across all instances of the specified
Agenda running during the test.
AgendaName@LoadGenerator-sums the values for each statistic across all instances of the
specified Agenda running under the specified Load Generator during the test.
The statistics displayed in each row are grouped into the following categories:
Per time unit statistics are ratios that calculate an average value for an action or process. For
example: Hits Per Second, Rounds Per Second.
Timers track the amount of time it took the Virtual Clients to complete an action or process.
For example: Connect Time, Round Time.
Counters track the number of times the Virtual Clients completed an action or process. For
example: Hits, Rounds.
By default, WebLOAD displays values for its statistics at 20-second intervals. The reported
value at each reporting interval is called the current value.
Statistic Definition
Hits Per Second A hit is a request for information made by a Virtual Client to
the application being tested (the ABT). In other words, an
HTTP request.
For example, a Get statement for a URL retrieves a page. The
page can include any number of graphics and contents files.
Each request for each gif, jpeg, html file, etc., is a single hit.
Hits Per Second is the number of times the Virtual Clients
made an HTTP request divided by the elapsed time, in
seconds.
The value posted in the Current Value column is the number
(sum) of HTTP requests per second during the last reporting
interval.
Successful Hits Per Second The number of times the Virtual Clients obtained the correct
HTTP response to their HTTP requests divided by the elapsed
time, in seconds.
The value posted in the Current Value column is the number
(sum) of successful HTTP requests per second during the last
reporting interval.
Failed Hits Per Second The number of times the Virtual Clients did not obtain the
correct HTTP response divided by the elapsed time, in
seconds.
The value posted in the Current Value column is the number
(sum) of unsuccessful HTTP requests per second during the
last reporting interval.
Statistic Definition
Pages Per Second A page is an upper level request for information made by a
Virtual Client to the application being tested (the ABT). For
example, a Get statement for a URL retrieves a page.
Pages Per Second is the number of times the Virtual Clients
made upper level requests divided by the elapsed time, in
seconds.
The value posted in the Current Value column is the number
(sum) of requests per second during the last reporting interval.
Successful Pages Per Second The number of times the Virtual Clients obtained the correct
HTTP response to their upper level requests divided by the
elapsed time, in seconds.
The value posted in the Current Value column is the number
(sum) of successful requests per second during the last
reporting interval.
Failed Pages Per Second The number of times the Virtual Clients did not obtain the
correct response to an upper level request divided by the
elapsed time, in seconds.
The value posted in the Current Value column is the number
(sum) of unsuccessful requests per second during the last
reporting interval.
Rounds Per Second The number of times the Virtual Clients attempted to run the
Agenda divided by the elapsed time, in seconds.
The value posted in the Current Value column is the number
(sum) of attempts (both successful and unsuccessful) per
second during the last reporting interval.
Successful Rounds Per The number of times the Virtual Clients completed an entire
Second iteration of the Agenda divided by the elapsed time, in
seconds.
The value posted in the Current Value column is the number
(sum) of successful iterations of the Agenda per second during
the last reporting interval.
Failed Rounds Per Second The number of times the Virtual Clients started but did not
complete an iteration of this Agenda divided by the elapsed
time, in seconds.
The value posted in the Current Value column is the number
(sum) of failed iterations of the Agenda per second during the
last reporting interval.
Statistic Definition
Throughput (bytes per The average number of bytes per second transmitted from the
second) ABT to the Virtual Clients running this Agenda during the last
reporting interval.
In other words, this is the amount of the Response Data Size
(sum) divided by the number of seconds in the reporting
interval.
Connection Speed (Bits Per The number of bits transmitted back and forth between the
Second) Virtual Clients and the ABT divided by the time it took to
transmit those bits, in seconds.
You can set the Virtual Clients to emulate a particular
connection speed during the test either by using the Variable
Connection Speed settings (Tools | Default Options |
Connection tab) or by coding the connection speed in the
Agenda.
If a connection speed is specified for the test, WebLOAD
reports it in the Statistics Report.
The value posted in the Current Value column is the number
(sum) of bits passed per second during the last reporting
interval. It should match, very closely, the connection speed
you specified for the test.
Timer Definitions
The following table presents definitions of the timer statistics:
Statistic Definition
Round Time The time it takes one Virtual Client to finish one complete
iteration of an Agenda, in seconds.
The value posted in the Current Value column is the average time
it took the Virtual Clients to finish one complete iteration of the
Agenda during the last reporting interval.
Hit Time The time it takes to complete a successful HTTP request, in
seconds. (Each request for each gif, jpeg, html file, etc. is a single
hit.) The time of a hit is the sum of the Connect Time, Send Time,
Response Time, and Process Time.
The value posted in the Current Value column is the average time
it took to make an HTTP request and process its response during
the last reporting interval.
Statistic Definition
Connect Time The time it takes for a Virtual Client to connect to the application
being tested (the ABT), in seconds. In other words, the time it
takes from the beginning of the HTTP request to the TCP/IP
connection.
The value posted in the Current Value column is the average time
it took a Virtual Client to connect to the ABT during the last
reporting interval.
If the Persistent Connection option is enabled (Tools | Default
Options | Browser Parameters tab), there may not be a value for
Connect Time because the HTTP connection remains open
between successive HTTP requests.
Send Time The time it takes the Virtual Clients to write an HTTP request to
the ABT, in seconds.
The value posted in the Current Value column is the average time
it took the Virtual Clients to write a request to the ABT during the
last reporting interval.
Response Time The time it takes the ABT to send the object of an HTTP request
back to a Virtual Client, in seconds. In other words, the time from
the end of the HTTP request until the Virtual Client has received
the complete item it requested (Time to first byte + Receive Time
- Process Time).
The value posted in the Current Value column is the average time
it took the ABT to respond to an HTTP request during the last
reporting interval.
Process Time The time it takes WebLOAD to parse an HTTP response from the
ABT and then populate the document-object model (the DOM), in
seconds.
The value posted in the Current Value column is the average time
it took WebLOAD to parse an HTTP response during the last
reporting interval.
Automatic Data If you have Automatic Data Collection enabled (Tools | Default
Collection timers Options | Functional Testing tab, Automatic Data Collection)
WebLOAD creates a timer for each GET and POST statement in
the Agenda.
The row heading is the URL of the page. The value posted is the
average time it took a Virtual Client to complete the GET or
POST statement, in seconds, during the last reporting interval.
Time to First Byte The time it takes from when a request is sent until the Virtual
Client receives the first byte of data.
Statistic Definition
Receive Time The elapsed time between receiving the first byte and the last
byte.
<User-Defined> timer You can add timers to Agendas to keep track of the amount of
time it takes to complete specific actions. (See the JavaScript
Programming for WebLOAD Test Agendas guide.) If there are any
timers in the Agendas that you are running, WebLOAD reports
their values in the Statistics Report.
The row heading is the name (argument) of the timer, that is, the
string in parenthesis in the SetTimer() function call. The timer
represents the time it takes to complete all the actions between the
SetTimer() call and its corresponding SendTimer() call, in
seconds.
The value posted is the average time it took a Virtual Client to
complete the actions between the pair of timer calls, in seconds,
during the last reporting interval.
<User-Defined You can add user-defined transaction functions to Agendas for
Transaction> timers functional tests. (See the JavaScript Programming for WebLOAD
Test Agendas guide.) If there is a user-defined transaction function
in the Agenda that you are running, WebLOAD reports a timer for
it in the Statistics Report.
The row heading is the name (argument) of the user-defined
transaction, that is, the string in parentheses in the
BeginTransaction() function call.
The timer represents the average time it took to complete all the
actions between the BeginTransaction() call and its corresponding
EndTransaction() call, in seconds, during the last reporting
interval.
Java and ActiveX timers You can add function calls to your Agendas that allow you to
instantiate and call methods and properties in Java and ActiveX
components. (See the JavaScript Programming for WebLOAD
Test Agendas guide.) If there are ActiveX or Java function calls in
the Agenda you are running, WebLOAD reports timers for them
in the Statistics Report.
The row heading is the name of the function call. The value
posted is the average amount of time it took to complete the
function call, in seconds, during the last reporting interval.
Statistic Definition
Server Performance If you selected Performance Monitor statistics for the report,
Measurements WebLOAD creates a row for them and reports their values in the
Statistics Report.
(For example: Sessions
Errored Out, % Processor For definitions of the statistics, see the Server Monitor Definition
Time, Current Disk dialog box (click the Pencil icon so that you are in edit mode, then
Queue Length, and so on) select Tools | Performance Measurements Manager).
Be selective when choosing server performance
measurements or the system resources required to
manage the data might affect the Console.
Counter Definitions
The following table presents definitions of the counter statistics:
Statistic Definition
Load Size The number of Virtual Clients running during the last
reporting interval.
Rounds The total number of times the Virtual Clients attempted to run
the Agenda during the last reporting interval.
Successful Rounds The total number of times the Virtual Clients completed one
iteration of the Agenda during the last reporting interval.
Failed Rounds The total number of times the Virtual Clients started but did
not complete the Agenda during the last reporting interval.
Hits The total number of times the Virtual Clients made an HTTP
request to the application being tested (ABT) during the last
reporting interval. (Each request for each gif, jpeg, html file,
etc. is a single hit.)
Successful Hits The total number of times the Virtual Clients made an HTTP
request and received the correct HTTP response from the ABT
during the last reporting interval. (Each request for each gif,
jpeg, html file, etc. is a single hit.)
Failed Hits The total number of times the Virtual Clients made an HTTP
request but did not receive the correct HTTP response from the
ABT during the last reporting interval. (Each request for each
gif, jpeg, html file, etc. is a single hit.)
Attempted Connections The total number of times the Virtual Clients attempted to
connect to the ABT during the last reporting interval.
Statistic Definition
Successful Connections The total number of times the Virtual Clients were able to
connect to the ABT during the last reporting interval.
This number is always less than or equal to the number of
successful hits because several hits might use the same HTTP
connection if the Persistent Connection option is enabled
(Tools | Default Options | Browser Parameters tab).
Failed Connections The total number of times the Virtual Clients tried to connect
to the ABT but was unable to, during the last reporting
interval.
This number is always less than or equal to the number of
failed hits because hits can fail for reasons other than a failed
connection.
Responses The number of times the ABT responded to an HTTP request
during the last reporting interval.
This number should match the number of successful hits.
Return Status WebLOAD creates a row in the Statistics Report for each kind
of HTTP status code it receives as an HTTP response from the
ABT (redirection codes, success codes, server error or client
error codes).
The value posted is the number of times the Virtual Clients
received that status code during the last reporting interval.
Response Data Size The size, in bytes, of all the HTTP responses sent by the ABT
during the last reporting interval.
WebLOAD uses this value to calculate Throughput (bytes per
second).
Automatic data collection If you have Automatic Data Collection enabled ( Tools |
counters Default Options | Functional Testing tab, Automatic Data
Collection) WebLOAD creates three counters for each GET
and POST statement in the Agenda.
The counters are the total number of times it occurred, the
number of times it succeeded, and the number of times it failed
during the last reporting interval.
Statistic Definition
User-defined counters You can add your own counters to Agendas with the
SendCounter() and the SendMeasurement() functions. (See the
JavaScript Programming for WebLOAD Test Agendas guide.)
If there is a user-defined counter in the Agenda that you are
running, WebLOAD reports their values in the Statistics
Report.
The row heading is the name (argument) of the counter, that is,
the string in parenthesis in the SendCounter() or
SendMeasurement() function call.
The value reported is the number of times the counter was
incremented during the last reporting interval.
<User-Defined Transaction> You can add user-defined transaction functions to Agendas for
counters functional tests. (See the JavaScript Programming for
WebLOAD Test Agendas guide.) If there is a user-defined
transaction function in the Agenda that you are running,
WebLOAD reports three counters for it in the Statistics
Report.
The row heading is the name (argument) of the user-defined
transaction, that is, the string in parenthesis in the
BeginTransaction() function call.
The counters are the total number of times it occurred, the
number of times it succeeded, and the number of times it failed
during the last reporting interval.
Java and ActiveX counters You can add function calls to your Agendas that allow you to
instantiate and call methods and properties in Java and
ActiveX components. (See the JavaScript Programming for
WebLOAD Test Agendas guide.) If there are ActiveX or Java
function calls in the Agenda that you are running, WebLOAD
reports three counters for them in the Statistics Report.
The row heading is the name of the function call.
The counters are the total number of times it
occurred, the number of times it succeeded, and
the number of times it failed during the last
reporting interval.
The values reported for Virtual Clients are average values based on the data reported by the
total number of Virtual Clients running that Agenda.
The values reported for a Probing Client are the specific values for that single,
representative user.
By default, Virtual Clients are single-threaded so the load machines can support more Virtual
Clients overall. Probing Clients run with four threads in order to emulate a single user as closely
as possible.
Therefore, you should use the statistics that you obtain from the different types of emulated
users in different ways. For example, if you are interested in the exact experience of a typical
user of your site, examine the timers like Round Time, Hit Time, Connect Time, and so on from
a Probing Client. For performance or load issues, however, examine statistics like Successful
and Failed hits, Successful and Failed Rounds, Rounds Per Second, Hits Per Second, and so on
from Virtual Clients.
Summary Reports
Each cell in the Statistics Report is an access point for another report:
Double-click on a row heading to display a Summary by Measurement Report.
Double-click on a column heading to display a Summary by Agenda or Summary by Total
Report.
Double-click on a cell to display a summary report for that statistical measurement.
Value Definition
Current Slice/Current The value posted for this reporting interval - current slice.
Value
Min The lowest value reported for this statistic since the beginning of the
test.
Max The highest value reported for this statistic since the beginning of the
test.
Count (For timers only.) The total number of readings (the number of times
the item being timed occurred) for the timed statistic since the
beginning of the test. For example, for Hit Time, Count shows the
number of hits completed.
Average For per time unit statistics and counters, average is the total of all of
the current values divided by the number of reporting intervals.
For timers, average is the total amount of time counted by the timer
(not the elapsed time) divided by the Count (the total number of
readings). For example, the average for Hit Time is the amount of
time it took to complete all the successful hits divided by the number
of successful hits (the Count).
Sum The aggregate or total value for this statistic in this Agenda since the
beginning of the test.
STDev The average amount this statistic varies from the average number in
this Agenda.
The Statistics Report displays all of the statistics that WebLOAD collects unless there were no
readings for a statistic during a test. For example, if there were no unsuccessful rounds during a
test, the table would not display a row for Failed Rounds.
Additionally, if a measurement has no meaning for a specific statistic, WebLOAD displays a
dash (-) in the cell. For example, if there were two Agendas running but only one of them
returned an HTTP Response Status code, the table would include a row for that Response Status
and would display a dash in the cell for the Agenda that did not return the code.
Double click on To
Double click on To
The following example contains all of the values collected for the Agenda, SearchAgenda:
The names of all the measurements tracked during the current Load Session are listed in the
column at the far left. The names of all the Attributes totaled for the relevant measurements are
displayed in the header at the top of the table.
Example:
The screen above displays the Response Time measurement values for the Agenda,
SearchAgenda.
Note: The MINIMUM time needed for the ABT to respond to requests from the Agenda during
the current test session is 1.246 seconds. The MAXIMUM time needed for the ABT to respond
to the Agenda run by Generator1 during the current slice is 0.461 seconds. The AVERAGE
Response Time to the Agenda in the current slice is 1.398 seconds.
accessed from the Statistics Report main window by double-clicking on the button located at the
intersection of the Response Time row and SearchAgenda column.
This table displays the sum of all measurement values for all current Agendas.
The names of all the measurements tracked during the current Load Session are listed in the
column at the far left. The names of all the attributes totaled for the relevant measurements for
all current Agendas are displayed in the header at the top of the table.
Double-click on Response Time in the left-hand column to call up a History Report. This
History Report lists all of the Response Time Attribute values recorded for all Agendas from the
start of the current test session. This is the same Summary Report accessed from the Statistics
Report main window by double-clicking on the button located at the intersection of the
Response Time row and Total column.
This table displays all attribute values for the selected measurement for the current Agendas.
The screen above shows the Response Time values for Agenda, SearchAgenda. The attributes
displayed are for Current Slice, Min, Max Count and Average. The third line displays the total
values for these attributes for all running Agendas.
Example:
Double-click on SearchAgenda in the left-hand column to call up a History Report. This History
Report lists all of the Response Time Attribute values recorded for all Agendas from the start of
the current test session. This is the same History Report accessed from the Statistics Report main
window by double-clicking on the button located at the intersection of the Response Time row
and Agenda column.
The following table displays all of the Response Time Attribute values for the Agenda,
SearchAgenda.
The times that the attribute values were recorded since the start of the current Load Session are
listed in the column at the far left. The names of all the attributes totaled for the selected
measurement and Agenda are displayed in the header at the top of the table.
Example:
This table displays all Total Load Size, Current Value, Min, Max, count and average values
recorded for Response Time for SearchAgenda. Values were recorded starting at 20 seconds
after the Load Session began.
The History by Measurement and Total Report displays all the Attribute values recorded from
the start of the current Load Session for the selected measurement. These values are summed for
all Agendas:
Listed in the far left-hand column are the times that attribute values were recorded since the
beginning of the current Load Session. The names of all the attributes totaled for the selected
measurement are displayed in the header at the top of the table.
When testing the integrity of your Web application, it is important to see not only if a given hit
succeeded or failed, but also the reasons for the failure to pinpoint the hits exact weak points and
bottlenecks. Data Drilling enables you to display a detailed description of all user-defined and
named hits to the instance level. With Data Drilling you can view all hit failures and the reasons
for them.
The WebLOAD Data Drilling reports and the Table Compare Viewer provide a graphic
illustration of the problematic tables, making it simple to locate and correct the source of the
error.
Data Drilling and Table Compare reports are available after the Load Session has completed.
In This Chapter
Data Drilling Reports.................................................................. 335
Table Compare Reports .............................................................. 341
A named entry for each HTTP action in the Agenda. Information is automatically saved for
all actions tested using the Functional Verification Wizard. Information is also saved for all
user-defined or manually named transactions.
Click on a specific entry to zoom in on the detailed information available for that action. For
each HTTP action tested and verified by your test Agenda, WebLOAD provides a complete
breakdown of all relevant information, including:
A hierarchical tree of all the 'sub-steps' needed to complete the specified action. For
example, a single Get of a simple Web page may include multiple 'sub-Gets' or hits of
the various frames found on that page, each with its own pictures and tables, each of
which is listed with its own 'sub-sub-Get' as well.
Note: WebLOAD works with the 'visual set' of Web page elements associated with each
action. Only the targeted frames that were actually accessed during a specific action will be
saved as part of the DOM for that action, and not necessarily every single peripheral frame
associated with the parent Web page.
Detailed information about the time required for each action and sub-transaction,
including low-performance flags for actions that complete correctly, but take longer
than a user-set time threshold to complete.
A hierarchical tree of all the objects accessed or otherwise associated with the specified
action, together with detailed information about the time required to access each object.
The timing information is broken down into the time required for each HTTP activity
associated with that object, including connect time, send time, wait time (time to first
byte), and final receive time.
Entries for every timer triggered during the test session, including user-defined timers,
WebLOAD-defined built-in timers, and automatic timers triggered each time ActiveX or
Java object methods are activated. Click on a timer entry to bring up detailed information
about that timer.
Entries for every error event triggered during the test session by both WebLOAD and user-
defined verification functions. Click on an event entry to bring up detailed information
about that event.
The Transactions Grid displays summary information for all user-defined and named
transactions incorporated in the Agenda including:
Attribute Explanation
In this example the UpdateAccount transaction is highlighted in the Transaction Grid. The
Reason for the UpdateAccount transaction failure, Transfer incomplete displays immediately
below the Transaction Grid. The Transaction Failure Reason Grid displays why two of the
UpdateAccount transactions failed.
The Instance Grid lists the individual instances of transaction failure, together with specific
details about each instance.
The following information is included for each transaction instance displayed in the grid:
Attribute Explanation
Severity The severity level of this Transaction Instance failure, based on the
return code set by the Agenda in the Verification function.
Reason The reason that the Transaction Instance failed based on information
supplied by the Agenda in the SetFailureReason function.
Time The time that this Transaction Instance began execution, relative to
the beginning of the current test session.
Duration The time it took to complete the Transaction instance. If the
Transaction Instance failed, the duration is the time until the
transaction failed.
Total Load Size The total number of Virtual Clients running on the Load Generator
during this Transaction Instance.
Thread Number The number of the current thread running during this Transaction
Instance.
Round Number The number of the current round running during this Transaction
Instance.
Load Generator The name of the Load Generator that executed the Transaction
Instance.
The Instance Grid incorporates functional testing with performance testing and provides you
with a detailed description of exactly which Transaction Instance failed, why it failed, and what
else occurred at that point in the Agenda execution for that Load Generator.
To see more information about the actual event that triggered this error, double click on the
child transaction branch in the tree to expand its own sub tree, including its own Events sub-
branch. Double click on more entries under the Events sub-branch to display more information
about each selected item on the table at the right.
By working your way through the multiple levels of the Data Drilling reports you can achieve a
complete picture of the exact point at which your Agenda failed, along with a complete history
of the general system and events that led up to the failure. This level of detailed information
simplifies your work when testing and debugging your Web site. Table Compare Events display
graphically in the Table Compare Viewer. For more information on viewing the results of Table
Compare Events graphically see Table Compare Reports (on page 341).
Note: The Table Compare Viewer is only available with Internet Explorer 5.
The first level indicates that a minor error caused the table comparison to fail. For more
information, double click on the parent branch in the tree to enter the second report level,
expanding the Events branch under the selected transaction:
The second level tree lists the events that triggered the error. For an exact illustration of the
problem, click on the Table Compare branch:
The screen now displays the source of the error, showing both the expected and actual tables
with the mismatched table cells marked and the corresponding information highlighted in the
Description and Log windows.
Note: The Transaction Instance Tree is available only if User-Defined Transactions were
initiated in the Agenda for functional testing and WebLOAD is configured to Collect
Instances in the Tools | Default Options | Functional Testing tab.
Performance Reports are the product of your test Session. Performance Reports enable you to
combine several different measurements into a single report. You are able to view the data
accumulated in the NT and Unix Performance Monitors together with the performance data
from the Load Generators and Probing Clients. All of the data displayed in the Integrated
Reports and Statistics can be exported to HTML, Microsoft Excel and to Tab files.
In This Chapter
Report Export Options................................................................ 345
Exporting Selective WebLOAD Report Data............................. 347
Exporting All Statistical Data ..................................................... 353
Customizing the Export File ....................................................... 357
Creating a Summary Report ....................................................... 359
Category Description
The screen below displays the data for the Response Time in tabular format.
All of the examples in this section refer to these reports (EXAMPLE 1 and EXAMPLE 2)
-Or-
Notes:
If Excel is not installed, you receive the following error: Cannot Open Excel.
Exporting a large amount of data to Excel may take some time. Wait until the export to
Excel has completed before continuing work with the Console.
-Or-
The following screen shows an Integrated Report after it is exported to a Tab file and open in
Microsoft Word (using tab as a delimiter). You can change the presentation of this file using the
Export options dialog box.
-Or-
-Or-
Click on the Export All to Excel toolbar button.
Excel opens automatically and displays all of the Statistics gathered by WebLOAD.
The following screen shows the Statistics as they appear when exported to Excel.
-Or-
Click on the Export All to Tab File toolbar button.
The Windows Save As dialog box is displayed.
2. Enter the name of your new Tab file. The file has a *.Tab extension.
3. Click Save.
Once the file is saved, the data is displayed.
Configure the viewer and the view options in the Export options dialog box, see Customizing the
Export File (on page 357).
Change the presentation of this file using the Export options dialog box.
The left column contains the times (in seconds) that the measurement values were recorded.
The top row contains the name of each measurement in the following format: Agenda /
Measurement / Attribute.
See Customizing the Export File (on page 357), for information on configuring the sampling
frequency, the delimiters of your exported table, the viewer, and the view options.
-Or-
Click on the Export All to HTML toolbar button.
Change the presentation of this file using the Export options dialog box.
The left column contains the times (in seconds) that the measurement values were recorded.
The top row contains the name of each measurement in the following format: Agenda /
Measurement / Attribute.
See Customizing the Export File (on page 357), for information on configuring the sampling
frequency, the delimiters of your exported table, the viewer, and the view options.
Export Every Statistics The frequency of the number of samples exported. For
Reporting example, if “1” is entered for this value, WebLOAD
exports every sample taken. If “2” is entered for this value,
WebLOAD exports every second sample. When
configuring this option, you must consider the setting
selected for sampling. If for example, the Statistics
Reporting Resolution is set to 20 seconds and Export
Every Statistics Reporting is set to 2, the measurement
exported reflects a sampling every 40 seconds. The
measurements are the measurements taken at the intervals:
20, 60, 100, 140… This option is useful in cases where
there is a large amount of data due to a long test Session.
Open Viewer When Exporting Check to launch the Tab File and HTML Default Viewer
after the Tab or HTML file is exported. This is checked by
default
Tab File Delimiter Definition of the delimiter used when exporting the
statistical data to a tab file. All strings can be used as
delimiters by entering the string in the Tab File Delimiter
field.
To use tab characters as delimiters enter \t.
To use a new line as a delimiter, enter \r or \n.
To use a space as a delimiter, use \s.
Tab \ HTML File Default The default directory for saving Tab and HTML files.
Directory
Tab File Default Viewer The default viewer to open after the Tab file is exported.
1. Click OK.
-Or-
Begin your work in the Summary Report Wizard by customizing the report.
Report Destination
You may export the Summary Report to Microsoft Excel, HTML or a Tab file.
WebLOAD provides the Performance Measurements Manager (PMM) for monitoring the
performance of various server-side applications, databases, stream technology, system and Web
server measurements in real-time while your test is running.
In This Chapter
Using the Performance Measurements Manager ........................ 365
About the Performance Measurements Manager........................ 366
Opening the Performance Measurements Manager .................... 366
Using PMM Configuration Files ................................................ 373
Configuring the Data Sources, Hosts, and Measurements to Monitor 375
Application Server Monitors ...................................................... 385
Database Monitors...................................................................... 387
Web Server Monitors.................................................................. 391
Monitoring Generic SNMP Data Sources................................... 392
To open the PMM from within WebLOAD, select Session Control | Performance
Measurements Manager.
To open the PMM as a standalone tool select Start | All Programs | TestView | Tools |
Performance Measurements Manager.
PMM opens with no configured data sources listed.
Menu Bar
The PMM menu bar is located at the top of the window.
This section discusses each of the PMM menus, their options and their related toolbar buttons.
File Menu
The File menu includes the following functions for PMM configuration file management.
Note: Menu entries vary depending if PMM is run from within WebLOAD or as a standalone
tool.
Actions Menu
The Actions menu includes the following functions for creating a PMM configuration file and
controlling the Configuration pane.
Add Data Source Open the PMM Wizard for configuring new data sources, hosts
and measurements to monitor.
Delete Deletes the currently selected item and all sub-items from the
configuration.
Clear All Removes all elements from the current configuration.
(available only in
the PMM
standalone tool)
Expand Tree Expands the configuration tree.
Help Menu
Use the PMM Help menu to access online help, as well as copyright and version information
about PMM.
Configuration Pane
The Configuration pane displays the performance objects that are configured for monitoring,
including hosts, data sources and measurements.
Description Pane
The Description pane displays information about the item selected in the Configuration window.
As you click different items in the Configuration window, the information displayed in the
description pane changes appropriately.
The following figure shows the PMM configuration tree, configured to collect data from two
hosts, each with one data source and multiple measurements.
Data Source
Measurement
The following options can be accessed through the right mouse button:
Host Add data source Opens the PMM Wizard, enabling you to select
additional data sources to monitor on the selected
host.
Delete Removes the selected host from the configuration.
WebLOAD User's Guide 371
Opening the Performance Measurements Manager
Data Source Add measurements Opens the PMM Wizard to enabling you to select
additional measurement to monitor on the selected
data source.
Delete Removes the selected data source from the
configuration.
Measurement Delete Removes the selected measurement from the
configuration.
Action Result
Click Selects an item and displays a description for the selected item.
Double-click Displays or hides items contained by the selected item.
Right-click Displays the Actions shortcut menu for the selected item.
The following table lists the keystrokes you can use to move in the PMM main window.
Key Result
Key Result
Asterisk “*” on the Expands the entire tree below the selected node.
numeric keypad
To clear an existing configuration from the configuration window in order to define a new
configuration file:
2. Browse to the location of the PMM configuration files, select a file, and click Open.
If a configuration is already defined and displayed within the PMM configuration window, a
message box displays:
3. Select File | Close and update template, or click the Close & update toolbar button.
PMM closes and the configuration file is assigned to the current template.
Data sources
Hosts
Measurements
1. From the PMM activated in WebLOAD, select File | Export PMM, or click the Export
toolbar button.
The Save As dialog box opens.
2. Browse to the location you want to save your PMM file, enter a name for the file in the File
Name field, and click Save.
1. From the PMM standalone tool, select File | Save, or click the Save toolbar button.
The Save As dialog box opens.
2. Browse to the location you want to save your PMM file, enter a name for the file in the File
Name field, and click Save.
2. Click Next.
The Data Source screen displays.
3. Expand the tree to view all available data sources. PMM supports the following types of
data sources:
Application Servers:
BEA WebLogic
IBM WebSphere 5
Microsoft Active Server Pages
Microsoft ASP .Net
Databases:
Oracle
Microsoft SQL Server
Microsoft SQL Server using SNMP
System:
Windows platform
UNIX platforms Network data using SNMP
UNIX platforms using RSTATD
UNIX UC-Davis
Windows .Net Framework
Web Servers:
Microsoft Internet Information Server (IIS)
Sun One (former iplanet 6)
Stream Technologies:
Microsoft Media Server
Other:
General Performance Monitor
General Performance Monitor can be used just like the System > Windows Platform.
The difference is that it gives access to all Performance Monitor counters.
General SNMP
General SNMP can be used similarly to other SNMP-based PMs. The difference is that
is gives access to all MIBs served by the SNMP agent on that machine.
General UNIX
General UNIX can be used similarly to the System > UNIX platform using RSTATD.
The difference is that is gives access to all RSTATD counters.
6. Choose the host to monitor from the list and click Select.
The host is added to the Host Selection screen.
7. Press Next.
PMM attempts to connect to the host. If a connection is made, the Wizard progresses to the
Measurements to Monitor screen. If a connection cannot be established, the Wizard returns to
the Host Selection screen, enabling you to select an alternate host.
8. Select one or more measurements to monitor by clicking on the checkbox adjacent to the
item. PMM only collects the data for measurements that are checked. All other data is not
collected.
Explanations of each measurement are available by clicking on the measurement name. The
explanation appears at the bottom of the window.
The list of available measurements varies depending upon the type of data source. Each data
source is supplied with a default set of measurements to monitor.
Notes:
An empty circle next to a measurement indicates that the measurement contains sub-
components and that no default subcomponents are defined (the upper level component
cannot be selected). To select subcomponents, click on the + to expand the tree.
A circle with a checkmark next to a measurement indicates that the measurement
contains sub-components and that default subcomponents are defined. To view and
configure the configured subcomponents, click on the + to expand the tree.
9. Click Next.
The Wizard displays a summary of the host, data source, and measurements configured for
monitoring.
11. To add additional data sources, hosts and measurements to monitor, click the Add data
source toolbar button and repeat the process.
Deleting a Measurement
Ensure that the WebLogic SNMP agent is installed and activated on the WebLogic server
host.
Configure the WebLogic Server to be a proxy for the SNMP agent, described in Setting Up
the SNMP Agent and Configuring the WebLogic Server as a Proxy for the SNMP Agent (on
page 386).
Configure the WebLogic Server host to use port 1161 as the SNMP message port, described
in Configure the WebLogic Server Host to use Port 1161 (on page 386).
These configuration tasks are necessary because some UNIX/Linux machines have
authentication issues when working with the default SNMP port 160.
Attribute Value
Port 1161
Root .1.3.6.1.2.1 (be sure to include the leading period)
Community Public
Timeout 5000
Instructions for creating the proxy are available at:
http://e-docs.bea.com/wls/docs70/////ConsoleHelp/snmp.html (http://e-
docs.bea.com/wls/docs70/////consolehelp/snmp.html)
Look under Create an SNMP Proxy.
3. Restart the WebLogic Administration Server so that the changes take effect.
Database Monitors
PMM enables you to monitor database servers during runtime. This allows you to view
performance data from the database in real-time. Before running your test session, you must first
configure the PMM monitors.
WebLOAD supports monitoring for the following databases:
Oracle
Microsoft SQL Server
Microsoft SQL Server using SNMP
Note: The port (1161) must match the unused port specified in step 5.
start_peer Script
snmpd.config File
su root
./start_peer -a
This command starts all three processes.
3. Now use the ps command to verify that all three processes were started:
ps -aux |grep peer
ps -aux |grep snmpd
ps -ef | grep snmp
Setting up the SNMP Agent and Configuring the Sun One Server as a Proxy
for the SNMP Agent
To enable PMM to monitor performance data from the Sun One application server you must
first set up the Sun One SNMP agent. The Administration Interface will not start the OS master
agent on ports other than 161. The workaround is to use the Sun One SNMP master agent,
modify the native master agent to use a different port, and enable the Sun One SNMP subagent
to gather information about the server.
WebLOAD User's Guide 391
Monitoring Generic SNMP Data Sources
a. Make sure you know how to restart it and which MIB it supports.
b. Kill its process.
2. Edit the CONFIG file located in the plugins/snmp/magt server root
directory from:
…
snmp 161/udp
…
To:
…
snmp 1161/udp
…
3. Manually activate the Sun One master agent and subagent.
b. Restart the Sun One Server and the subagent from the Web server Administration.
SNMP relies on MIB (Management Information Base) files. These files are always readable,
and usually have a file extension of either .mib or .txt, though others may have proprietary file
extensions. These files provide a structure to the information that an SNMP agent can provide to
a program that monitors SNMP. There are publicly available specifications for SNMP MIBs,
and there are privately provided MIBs which are specific to a particular product.
All of the MIBs, both public and private, are organized into a tree structure that helps to identify
them. Each location in the tree is specified by both a name and a numeric value. The MIB tree
has ISO at its root, and then four main branches: directory, mgmt, experimental, and private.
The private branch is where enterprise level MIBs are placed, and their location in the tree can
be specified both by writing out the tree structure with points separating the names, and also
numerically, using the numbers associated with each branch in the tree.
For example, all private SNMP MIBs will have the following structure:
ISO has the value of 1
Internet has the value of .3.6.1
Private has the value of 4
Enterprise has the value of 1.
So any enterprise level MIB will have an address that starts with:
.iso.org.dod.internet.private.enterprises
The numeric equivalent is:
.1.3.6.1.4.1
Each enterprise which creates and supports its own SNMP MIBs, such as UC Davis, WebLogic,
or others will have an assigned place within the tree which provides it's name and numeric
name. Each MIB also has associated attributes, which describe the SNMP data that is returned.
There is often documentation of each of these MIBs and the base names and numeric numbers
associated with each in the products being used. Another useful way to get information on
installed MIBs is to use a MIB browser.
attr_range=1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21
,22,23,24
attr_names=oraDbSysConsistentChanges,oraDbSysConsistentGets,oraD
bSysDbBlockChanges,oraDbSysDbBlockGets,oraDbSysFreeBufferInspect
ed,oraDbSysFreeBufferRequested,oraDbSysParseCount,oraDbSysPhysRe
ads,oraDbSysPhysWrites,oraDbSysRedoEntries,
oraDbSysRedoLogSpaceRequests, oraDbSysRedoSyncWrites,
oraDbSysSortsDisk,oraDbSysSortsMemory,oraDbSysSortsRows,
oraDbSysTableFetchRowid,oraDbSysTableFetchContinuedRow,
oraDbSysTableScanBlocks,oraDbSysTableScanRows,
oraDbSysTableScansLong,oraDbSysTableScansShort,oraDbSysUserCalls
,oraDbSysUserCommits,oraDbSysUserRollbacks,oraDbSysWriteRequests
The structure of the UserMibFilter.ini file is as follows:
Parameter Description
[…] A comment that describes the MIB information that follows. This comment is
displayed by the Performance Measurements Manager (PMM) when a host
running the SNMP agent is added to a test.
base The name derived from the MIB tree. The first 6 elements describe the
common MIB identifier. The seventh element describes the object, in this
example Oracle, and the remaining elements describe the elements of the
private object. For more information, see Understanding MIB Files (on page
392).
numbase The numeric names, which correspond to the MIB tree location, listed in the
base. For more information, see Understanding MIB Files (on page 392).
instance Specifies a table or set of tables within the MIB file that contain the attributes
to monitor.
instdesc Provides a way for WebLOAD to display some of the MIB data in the PMM
window. This parameter is often either set to the generic term instance or
simply left blank.
prefix The prefix is provided to give the option of removing things from the names
provided by the attr_names parameter.
file Represents the number corresponding to the name of the MIB file from which
the SNMP data comes. This file is listed at the top of the UserMibFilter.ini
file.
attr_range The attributes or leaves of the MIB for which WebLOAD reports statistics.
attr_names The attributes or leaves of the MIB for which WebLOAD reports statistics.
a. Enter the base, numbase, and instance, determined from MIB file, to the WebLOAD
UserMibFilter.ini file.
b. Using a MIB browser, navigate through the MIB tree to the attributes you want
WebLOAD to report on.
c. Add the attributes to the attr_names in the WebLOAD UserMibFilter.ini file.
d. Add the corresponding number (which comes directly from the order in which the
attributes appear in the MIB file) to the attr_range parameter.
5. Save and close the updated UserMibFilter.ini file.
6. Restart WebLOAD and TestTalk.
7. Use Performance Measurements Manager to add the host providing the SNMP data to your
PMM configuration.
Note: WebLOAD monitors SNMP statistics that are numeric in nature. In the MIB file of the
attributes you want to monitor, locate the attribute by name and determine if it is defined as
numeric or string. If the attribute is a string attribute (such as a DisplayString) WebLOAD
cannot monitor it.
Example
The following example demonstrates adding information to the WebLOAD UserMibFilter.ini
file to run the UC Davis implementation of SNMP:
[UC Davis SNMP]
base=.iso.org.dod.internet.private.enterprises.ucdavis.dskTable.
dskEntry
numbase= .1.3.6.1.4.1.2021.9.1
instance= .serverTable.serverEntry
instdesc= instance
prefix=
file= 11
WebLOAD User's Guide 395
Monitoring Generic SNMP Data Sources
attr_range=1,2,3,4,5,6,7,8,9,10
attr_names=dskIndex,dskPath,dskDevice,dskMinimum,dskMinPercent,d
skTotal,dskAvail,dskUsed,dskPercent,dskPercentNode
This section describes the errors and log messages that can be generated by WebLOAD. These
log messages include messages generated by the JavaScript compiler, the Console, and any
human errors or user messages detected by WebLOAD at run-time. Log messages can also be
added to the Agenda by the user. The log messages are displayed in the Log Window.
In This Appendix
Message Severity ........................................................................ 397
WebLOAD Error Messages........................................................ 398
Message Severity
Messages fall into four categories:
User Note
Warning
Error
Severe
USER NOTE
User Notes are messages added to the Agenda by the user. User Notes can be used in order to
facilitate debugging. They do not stop WebLOAD from running.
WARNING
Warning errors do not stop WebLOAD from running. However, they do indicate suspicious
conditions. The source of the problem may be generated by the ABT, the connection to the
server, networking etc.
ERROR
Errors do not stop WebLOAD from running. However, they do stop the round and restart the
Agenda for the Virtual Client that encountered the error.
SEVERE
Severe errors cause the Load Generator, on which the error was generated to stop immediately.
A dialog box will be displayed.
Connection timeout Connection has timed out. Reason for Error /Severe
failure not specified. If this connection is
essential for WebLOAD to continue, this
error will be considered Severe and
WebLOAD will stop. If the rest of the
test Session is able to continue, this error
will be a Warning error only.
Data retrieval error Data retrieval error. Cannot retrieve data Severe
from the server.
Delete Cookie error- Parameters missing from DeleteCookie Error
Parameters function.
Delete Cookie error-not Error deleting cookie. Cookie not found. Error
found
Error while reading Error reading data from the server. Error
Forbidden, reason not The specified URL address or server Severe
specified request is forbidden.
Gateway error Invalid gateway response to WebLOAD Severe
request.
Gateway timeout Gateway did not respond within specified Severe
timeout period. Reason for failure not
specified.
Illegal command Illegal JavaScript command. Severe
Illegal request Illegal server request. Severe
Internal server error The server went down. Severe
Length required More length is required for the current Severe
request than is presently available.
Method not allowed Specified access or request method is not Severe
allowed.
No buffer space No buffer space available. Unable to Error
access the server.
Not Acceptable Specified configuration, object Severe
description, parameters, authentication
fields, or other submitted information is
not acceptable.
Not implemented Selected feature is not currently Severe
implemented
Other Failure due to any other reason not Severe
specified on this list.
WebLOAD User's Guide 399
WebLOAD Error Messages
B. Reference Materials
In This Appendix
WebLOAD Files ......................................................................... 403
WebLOAD Hot Key Commands ................................................ 404
WebLOAD Files
The following is an explanation of the files associated with a WebLOAD Load Session.
Index
Adding Host Computers to the Host Computers
. List • 173
.dat • 129 Adding Measurements to a Configured Data
.fta • 403 Source • 383
.fte • 403 Adding Performance Goals • 177
.fts • 403 Adding SNMP Agents • 394
.log • 271, 403 additional information
.ls • 129 application being tested • 128, 130
open • 132 created by • 128, 130
save • 129 custom • 128, 130
.mix • 139 load session • 130
save • 171 load template • 127
.pft • 403 load template fields • 128
.tpl • 125, 276 results description • 130
open • 127 session comments • 130
set • 128, 130
A
subject • 128, 130
About Agenda Options • 222, 224
template comments • 128, 130
About the Goal Definition Dialog Box • 175
test description • 128, 130
About the Performance Measurements Manager •
title • 128, 130
366
After Completing the WebLOAD Wizard • 159
About WebRM • 122
Agenda • 136
Acknowledgments from a Ping • 67
add to mix • 170
Actions menu
assign equal weight in mix • 170
PMM • 369
copy • 72
Actions Menu • 369
create • 36
Activating Throttle Control • 286
create new • 36
ActiveX
current session options • 222
counters • 320
default options • 222
timers • 317
delete from mix • 140, 170
Actual Repository Files
icon • 82
.fta • 403
mix • 165, 191
Adding a User-Agent • 233
options • 221, 222
Adding a User-Defined Timer • 303
record • 36
Adding an Agenda to a Mix • 170
run a mix of Agendas • 167
Adding Data Sources • 376
run in cruise control wizard • 167
Adding Data Sources to a Configured Host • 384
run mix • 192
Adding Host Computers • 142
WebLOAD User's Guide 407
Index
authentication for protocol scripts set load generator thread properties • 235
Agenda options • 240 set probing client thread properties • 235
session options • 240 set redirection limits • 235
authentication for protocol scripts tab • 240 set user agent • 232
authentication settings simulate DNS cache • 236
configure • 240 simulate SSL cache • 236
automatic data collection browser emulation settings
Agenda option • 248 configure • 231
counters for • 320 browser type
enable • 248 set • 232
HTTP level • 248
C
object level • 248
pages • 248 cache
session option • 248 clear • 36
timers for • 317 case inspections
Automatic State Management diagnostic syntax check • 244
messages Changing the Length of the Reporting Interval •
enable • 245 314
automatic transactions Changing the Reports Parameters • 302
data drilling • 248 Changing the Scale of a Report for a Selected
Measurement • 308
autosave file name
client
setting • 205
web service • 258
average • 323
Client Type Feature Summary • 26, 28
B Client Type Settings • 224, 225
BEA WebLogic client type tab
configuration tasks • 385 Agenda options • 225
configure • 385 full client • 225
configure hosts to port 1161 • 386 HTTP protocol mode • 225
data source • 376 session options • 225
set as proxy for SNMP agent • 386 command line
Browser Client • 26, 27 example • 278
browser emulation launch WebLOAD • 277
add user agent • 233 parameters • 277
Agenda options • 231 syntax • 278
delete user agent • 234 Commands Dialog Box • 104
enable keep-alive • 236 communication
enable persistent connection • 236 between Hosts • 62
options • 231 Communication Between Host Computers • 62
session options • 231 communications
set browser type • 232 errors • 278
set HTTP version • 237 establishing • 278
WebLOAD User's Guide 409
Index