Availability testing

The Availability Test feature needs to be enabled for each account. If you want to use this feature but don't see it in the Synthetic Classic Portal, please contact your account manager.

What is an Availability test?

Availability testing tells you whether your web application is up or down. The simple protocol-based tests provide a measure of up-time for your application.

Currently, the HTTP test type is implemented. Other test types may be added in future releases.

An HTTP test makes an HTTP call to the server to show whether the target URL responds with a valid return code. Besides testing web pages, you can test simple single-step REST and web services.

The test is not intended to load the entire page and execute client-side JavaScript. It simply confirms that the page responds and whether an error occurred. Availability tests use cURL to request web pages; you can specify a user-agent if your web application is sensitive to the browser type.

Availability tests are available to all Synthetic Classic users, with no special activation required.

HTTP test features

Availability tests can run from any Backbone nodes. The tests don't require scripting or recording; they're defined through the Tests page in the Synthetic Classic Portal, or through the [Availability Test API]. You can use the API to create or manage multiple tests at the same time.

An Availability test supports a single request. In the test settings, you can add custom headers for user-agent, cookies, etc. You can also configure optional pass/fail criteria such as validation text and response time threshold.

Availability tests consume fewer XF points than Performance (Backbone, Last Mile, Private Last Mile, Mobile) tests. They can be scheduled to run as frequently as once every minute.

For the details of managing Availability tests, see Tests and Adding or editing Availability tests.

Limitations

  • Availability tests can't be used in interactive charts, operations dashboards, or Performance dashboards.
  • In this release, Availability tests can't be configured to run on Software Private Agents.

Viewing and analyzing Availability tests

You can view Availability test data in these Portal pages for data monitoring and analysis:

Availability test alerts are listed in the Alert logs, from which you can drill down from an Availability test alert to the Test overview page.

Error codes returned by Availability tests

The error codes listed below that are greater than 999 are specific to Availability tests.

  • 400 to 999 – HTTP status error codes.
  • 1000 to 1999 – Validation failed.
  • 1010 – Content validation failed. The content of a test did not meet a validation rule configured for the test.
  • 1011 – Time validation failed. A test took longer than the configured threshold, or longer than the built-in timeout of 30 seconds.
  • 1012 – Size validation failed. A test was larger (downloaded more bytes) than it should be.
  • 2000+ – Reviewing the error message may help resolve the issue; otherwise, contact Support and provide the error number in your help request.