The Client/Server Edition of Software Testing In The Real World gives you all the testing techniques in the standard class, plus the special techniques necessary for dealing with client/server systems, including Web systems. Software Testing In The Real World What works - and why What doesn't work - and why What can go wrong Testing Client/Server Applications Is Different Because: Both the data and the processing are distributed It's hard to simulate the complex interactions that occur New problems arise from timing - e.g. race conditions We need to predict performance and scalability by doing load testing on a small model The Basics and the Terminology Black Box vs. Glass Box Testing Why you can't test for everything How you can still test well anyway The difference between getting Code Coverage, Path Coverage, and other measures of completeness Risk Driven Testing How Risk Prioritization keeps testing on track Using Impact and Likelihood to determine Risk How to get the Risk numbers for your project and how to use them to prioritize testing How Bugs Get Missed Example: Lots of tests, but a major bug is still missed Test cases that would have found this bug, and why Getting traction on the problem: Systematic test case design techniques The Key: Designing the Right Test Cases What a good test case looks like Selecting outputs for testing, based on Risk What to use as inputs for the test case All Inputs Are Not Alike Do you have to test all combinations of inputs? The three types of inputs Which two types are trivial to test - and how to handle the type that counts Reducing the Number of Test Cases - New Techniques Making each test case count Reduction by using Equivalence Classes in an new way How to reduce the cases by exploiting dependencies How to spot redundant tests even though they look completely different Finding the Tough Bugs Using 'State of The World' variables to prevent surprises How to limit side effects by using Hidden Outputs Controlling unexpected 'Emergent Behavior' created by feedback loops What To Do When The Output is an Action The 'SEND msg TO userid' example How to tell whether the test worked right How many tests you need, and what they are How To Handle Complex Sets of Conditions Introduction to using Decision Tables in testing A better way to design Decision Tables How to subdivide Decision Tables to limit complexity What To Do When the System's Output is its Behavior Easier testing by using a States & Events model What the test cases are, with examples A new technique that keeps the number of States small and manageable How To Handle Interactions in Client/Server Systems How to interpret lack of response to a request Multiple processors: Who's in charge? How to test for partial updates and timing problems What causes 'emergent behavior' and how to coax it to appear during testing (the RASHR technique). What To Do When You Don't Have All The Pieces Using simple Stubs and Drivers to fill in for missing parts How to simulate network loading How to easily create other effects: locked records, timeouts, unavailability of servers, etc. Scalability - The Critical Issues What will happen with more users? How systems bog down - and why they crash from load The three levels of system loading, and meaning of each How to tell what kind of fix you need:       (1) more of the same (good, it scales up)       (2) architectural redesign (bad, it doesn't) How To Do Performance Testing Why performance prediction is so hard Ways to simulate system loading – which ones work The key: getting totally realistic simulation test data What tools and services are available Five pitfalls to avoid, with examples Mapping Tests to Requirements Breaking the requirements down for testing Requirements coverage based on Risk What to do if you don't have any (decent) requirements The Testing Process The realities: time pressure, getting along with developers We know it won't be perfect - but it can be very good Fatal Error: The Caboose How To Do Each Type of Testing Unit Testing Integration Testing - an innovative new approach System Testing A New Approach: CATS Beating the combination problem How constrained arrays minimize tests of many factors Results at Bell Labs on real systems How to do CATS testing yourself Testing Tools What's available How each type of tool works What each tool can - and can't - do for you (Includes coverage analyzers, automated capture / playback and script-driven regression testing tools, test management tools, and load / performance testing tools) Walkthroughs and Inspections What's the difference? How to perform them effectively The five pitfalls to avoid Metrics How much testing is enough? Metrics to tell how we're doing The one best metric for measuring testing effectiveness What To Put In The Client/Server Test Plan What is a Test Plan? The critical sections that must be in the Plan A complete client/server sample test plan with all the important sections filled out, and a template you can use for your own client/server test plan The Politics of Testing There's never enough time - how to get more Avoiding getting stuck with Mission Impossible How to handle technical people well and get their cooperation as part of the solution Resources Glossary Recommended Books Web resources Who Should Attend: This course is valuable for quality assurance and testing staff, programmers, systems analysts, and project managers.

 

Copyright © 1998 - 2005 by The Testing Center. All rights reserved. Images digitally watermarked.