Information:

Environment

  • Dynatrace Benchmarks
  • Synthetic Classic Portal
  • Backbone

 

Symptoms

Asynchronous content (such as third-party monitoring pixels) is included in Synthetic Classic results, and is impacting the reported page speed.

Solution

The Dynatrace Benchmarks standard has defined asynchronous objects as a part of the page, and as such they are included in the measurement time. Page load time is determined by the methodology of the test and how it waits for the page to complete. In this case, the test waits for DOM Ready and Browser Complete. There is a slight pause (usually ~.025 seconds), and then the test agent waits for any network activity to cease.

Asynchronous content affects Synthetic Classic measurements because the object time and pauses involved in waiting for that content are included in the measurement. The browser does not see a distinction between one object request and another; all content requested for the page is considered a part of the page, thus the Synthetic Classic measurement counts any asynchronous content called by the page as a part of the response time.

We have found numerous situations across many customers where third-party monitoring pixels have dramatically impacted page load times despite the pixels being asynchronous content. We have seen situations where a bad pixel has taken an entire page down, replaced the page post loading with a white screen, or locked up the browser preventing a user from accessing the page. Based on that experience, we consider the behavior of asynchronous content relevant to the end-user experience our metrics are intended to represent. 


 

 

 

 

  • No labels
  1. Anonymous (login to see details)

    If there is any script in page and it is working asynchronously then whether Gomez will wait for that script to load or will pass on to next step?