Types of Tests

There are various standard classifications of tests which are appropriate for use when testing a AEM project. You should be familiar with these to decide which you will use:

NOTE
These are listed in their chronological order of application.

Units Tests - Tests (usually) made by the development team to ensure that the individual elements behave correctly - albeit in isolation.

Integration Tests - Tests modules when combined. These tests are made after Unit Testing, but before System Testing.

Smoke Tests - These are quick-and-dirty tests used to prove that the software is running and high-level functionality is available. The details are not tested.

Functional Tests - These are used to test the functionality of the software. A series of tests will be designed to cover all functional details, with both expected and unexpected and/or erroneous input.

Black-box tests are functional tests of a complete unit / component / module, performed without knowledge of the internal workings of the element in question.

System Tests - These test the entire system once it has been fully integrated and installed on a suitable platform.

They test the functionality on a black-box basis.

Performance Tests - Performance tests are crucial when testing AEM.

They are used to illustrate the performance under differing conditions:

  • Normal

    Conditions which the site will experience for say 90% of the time. For example, when only a proportion of the authors are using the system.

  • Peak

    Conditions which will be experienced for a proportionally short time due to special circumstances; for example, when all authors use the system concurrently or when new content is published and a increased number of visitors view your site.

  • Extreme

    Can be used to emulate the performance forecast when new, extremely interesting content is published on your website. Then an extreme peak may be seen - though this may not always be fully predictable.

    These circumstances are sometimes seen when tickets for specific events are made available, or a much-awaited website is published for the first time.

The results are then used to tune the application.

Stress Tests - Stress tests are made to confirm how a component or application behaves under extreme conditions. In particular these tests are used to show how behavior deteriorates, when the element will fail - and how.

Regression Tests - Regression tests are used to confirm that functionality already proven in a previous release of the software is still operating correctly.

Regression Tests are good candidates for automation (if possible) to ensure they can be repeated quickly and consistently.

Acceptance Tests - Acceptance Tests are a special category as they are used to indicate the customer’s acceptance of the project.

The list of acceptance tests may contain a combination of tests from the various categories above, and are selected to verify that the project fulfils the customer’s requirements

See Acceptance and Sign-Off for more details.

Getting Started

Before starting on your detailed Test Cases and Test Plan you can:

Define the goals - Define your high-level goals to act as a starting point for fine-tuning as testing proceeds. You will want to:

  • Test the functionality according to the Detailed Requirement Specification.
  • Test Performance according to the Target Metrics.

among others.

Collect traffic statistics from the existing website - This information can be extracted from the log files - see Performance Monitoring for more details.

These figures will give an indication of current traffic (volume and spread) on the existing website and can be used in forming a base point for the new website.

Collect traffic statistics from external websites - If possible you can try to collect traffic statistics from other websites for comparison, but these figures are not always published.

Confirm Target Metrics - Metrics are used to define quantitative measurements for the quality of the website, as they represent the performance goals to be achieved.

They should be defined at the start of the project, together with the customer. See Target Metrics for more information.

Experience Manager


Elevate and Empower Teams with Agentic AI for Exceptional Experiences

Online | Strategy Keynote | General Audience

Elevate and empower your CX teams with AI that transforms creativity, personalization, and productivity. Discover how Adobe is...

Tue, Mar 18, 1:00 PM PDT (8:00 PM UTC)

Register

How EY Enhanced Regulatory Knowledge Management with Adobe Solutions

Online | Session | General Audience

Delve into how EY leverages Adobe solutions to enhance regulatory knowledge management. EY leverages search analytics and personalization...

Wed, Mar 19, 4:00 PM PDT (11:00 PM UTC)

Register

Connect with Experience League at Summit!

Get front-row access to top sessions, hands-on activities, and networking—wherever you are!

Learn more