26 March, 2020
Edge Testing takes aim at unnecessary overspend with new Performance Engineering service
Automated test capacity management service has saved customers millions
Pictured below: Dan Martland, our Head of Technical Testing.
We have launched a new Performance Engineering service today which combines two primary benefits to customers: finessing system design and capacity.
Fundamental aspects of system design can have a profound impact on the non-functional qualities of a system. For example, selecting an inefficient language or development platform can build-in overheads that will mean a business is paying too much for cloud services every day that a system is in use. By making the right choices from the very beginning of system specification and design you can reduce operational costs throughout the lifetime of a system. Our service advises on how to select the best technologies and configure them to design built-in efficiency and effectiveness.
Additionally, the new Performance Engineering service uses a unique and powerful combination of Machine Learning and pattern recognition to manage and model solution capacity. The algorithm behind the service is the result of years’ of research and practice, and provides the ability to accurately forecast capacity requirements in a wide range of environments and settings. This can yield significant resource savings in larger deployments, including test functions, and improve efficiency across the board.
Many enterprises have historically countered the obvious problem of under-provisioning hardware and cloud resources with the simple but inefficient tactic of over-provisioning. With our new Performance Engineering service this is no longer a risk that clients need to manage.
By creating actionable insights from existing provisioning and capacity data, the new service can predictably and automatically manage test capacity to much tighter tolerances, removing the guesswork from test capacity provisioning. The technique can be applied to all aspects of test capacity, from predicting cloud service requirements and scheduling them at the most cost-effective time, through to accurately specifying on-premise hardware requirements.
Performance Engineering offers a deeper view of the testing environment, delivering an early understanding of performance, making it a powerful complement to more traditional performance testing. Performance Engineering can enable earlier and more cost-effective interventions than traditional testing processes, although it should not be viewed as a replacement. Full solution, end-to-end load testing still brings value in terms of understanding the likely user experience. In combination with performance engineering the traditional performance test becomes more of a one-off acceptance test rather than an expensive late phase iterative activity. The resultant benefit of this combined approach being the reduction of the overall cost of quality.
Dan Martland, our Head of Technical Testing said: “We developed this solution for internal use originally, but the results were so startlingly strong that we began limited customer beta trials. In one case we reduced our client’s AWS annual hosting cost rate by £1 million, in another we saved a client £7.7 million by not over-specifying application servers. We believe that performance engineering will become a standard part of testing functions in the future.”
Our Performance Engineering service can reduce business risk and create operational efficiencies in a vast range of applications and requirements, including:
· Solution design for efficiency, including software, storage and network elements
· Effective scaling of test environments, especially in a high volume context.
Our team has offered advice on:
· Hyper-V and wider VM architectures
· AWS, Azure, Google Cloud Platform, Oracle and IBM hosting
· Lamba and Big Data Architecture
· Microservice Right-sizing, External Boundary Optimisation.
Back to News