Sascha Lüdemann | eHealth Manager

Can you present Sonic Healthcare?

Sascha Lüdemann: Sonic Healthcare is an international healthcare company focused on delivering quality, independent services in medical diagnostics and multidisciplinary medical practices.

Our Services include:

  • Laboratory medicine/pathology
  • Radiology
  • General practice
  • Occupational medical services
  • Clinical trials
  • Food and water testing

Our company is structured as a decentralized federation of medically-led practices, with our head office located in Sydney, Australia. We employ 32,000 people in Australia, New Zealand, the United Kingdom, Germany, Switzerland, Belgium, Ireland, and the USA. We are recognized for our dedication to our staff, our quality and services excellence and our financial strength.

Why were you looking for a web application load testing tool?

SL: The growing amount of more than 1,400 concurrent users of a specific eHealth application per server led to a higher risk for deployment of new major releases. The traditional testing did not provide the increasing reliability and performance requirements.

What were your challenges/unmet needs?

Simulate up to 5,000 concurrent real-world users for a load-balanced scaled server architecture for quality and reliability reasons and performance monitoring of new features and software refactoring issues. Additionally, we liked to benchmark our significant use cases for every single release and continuous integration.

How was web application load testing accomplished before NeoLoad?

SL: Homemade proprietary performance test drivers for server load testing. Only used for major releases.

Can you describe the application you tested?

SL: Our web application enables a medical referrer to communicate its orders to the laboratory online. Medical results are available inside our solution in that second when they are validated inside the laboratory.

This procedure saves time and almost eliminates errors and misunderstandings. The convenient and intuitive application means order management becomes easier, customized favorites can be defined and order transfer to the laboratory becomes easy and hassle-free.

Comprehensive information is received about the material required for analysis and essential sampling instructions. Required sample/collection materials are shown graphically.

Which protocols/technologies were tested?

SL: http, https, webservices, Gwt, atmosphere, rpc, websockets.

Why did you choose NeoLoad? Which alternatives were considered?

SL: Main criteria were compatibility to used technologies:

  • Gwt/Gxt
  • Atmosphere
  • Websockets

Additionally, criteria were the ability for cloud-based testing and local testing, available references and testimonials, pricing, and German-speaking contacts.

Can you briefly describe your load testing campaign?

SL: Compare all major use cases under load of 1,500 concurrent users for every single release as a benchmark against latest release. Compare specialized use cases for refactoring work provide sizing information for our data center partners.

How did NeoLoad contribute to improving your web application load and stress testing? What were the results of your load testing?

SL: While we were working on our benchmark tests, we recovered a couple of optimizations for our latest release to avoid hazards under full load. This was precious! We were able to increase the throughput of the application and realized up to 30% better average response time for the new release.

What is NeoLoad’s value to your business?

SL: Improved software quality, improved customer satisfaction, reduced risk, increased trust for new major releases (transparent benchmarks); Ability to provide reliable sizing information; Benchmark of alternative infrastructure, e.g., proxies, load balancers.

How will you use/deploy NeoLoad in the future?

SL: Add NeoLoad benchmarks into our continuous integration process. Automate/formalize benchmarking of releases.

Any additional comments?

SL: Great tool. Good support.

Sonic Healthcare

NeoLoad was precious! We were able to increase the throughput of the application and realized up to 30% better average response time for the new release