Load and Stress Tests

STRESS TESTS

Atentus has its own methodology, unique in the market, which consists of the installation of monitoring stations in the data centres of the main national and international internet service providers, generating users automatically, allowing us to objectively obtain the capacity of the website and its performance at different load levels.

What it consists of

This service consists of carrying out load or stress tests on your web platforms, determining the capacity of your site when faced with a high number of users. This is possible thanks to the Atentubots that generate an automatic and massive navigation, stressing and degrading the client's web application or platform, in order to know its real behaviour and capacity when faced with the demand of multiple simultaneous sessions.

 

This solution has an online help desk service during the execution of the service and complete reports with the results of each load test.

 

We stress Static Sites, Transactional Sites, Web Service SOAP & REST, Intranet Web Applications

NAVIGATION FLOW

1
1
Home: The initial URL of the website is loaded.
2
Login: login as a real user would.
3
Logout: Log out as a real user would after doing the full navigation flow predetermined by the client.

OBJECTIVES

Understand and mitigate risks related to poor performance of applications and websites, whether in development or production environments.

Proprietary Technology

We developed our own browser, which allows us to automate the navigation of a web application and interact with it in the same way as a real user would.

What are the steps:

We have a series of steps for the proper conduct of stress tests. These consist of:

  • Confirm technical feasibility
  • Running the load test
  • Running the stress test
  • Conduct an analysis of the results obtained
  • Deliver the results to the client for relevant decision making.
1

We deliver a set of metrics and results for each load test performed.

Availability

  • Successful flows
  • Failed flows
  • Successful steps
  • Failed steps
  • Error rate

Performance

  • Minimum times
  • Maximum times
  • Average times
  • Standard deviation
  • Variance
  • 95th percentile

Results

  • Performance

  • Availability

  • Screenshots of errors

  • Logs (Sources, Cookies, JS Errors)

Information Obtained

  • Point where response times degrade

  • Point where availability degrades

  • Number of simultaneous users supported

  • Critical sections of the navigation flow

  • Infrastructure performance