As the footprint and complexity of managed infrastructure and services grows, automating tests that measure performance and capacity levels of critical infrastructure and application services is the natural next step in taming cloud.
Wiley has adopted Kubernetes-as-Service platform based on Rancher to centralize certain concerns, like Kubernetes provisioning, user management, security policies, secrets management and monitoring, across the enterprise. At the same time individual teams have the freedom to manage their downstream clusters, users’ roles and responsibilities, and applications. Automated performance, load and stress testing that integrates easily with CI pipelines is crucial to sustaining the growth of our Cloud infrastructure and improving the velocity of continuous delivery for our applications.
Interestingly, the effort to develop and automate performance tests was started by SREs to measure and track the effect of changes on performance of critical platform services like Cortex, a scalable monitoring service based on Prometheus, and HashiCorp‘s Vault, a service for managing secrets.
A combination of Taurus, some of the Perf engines supported by it, and BlazeMeter we could easily execute such tests on demand or by a CI pipeline and review the results in real time in BlazeMeter.
On the heels of the success of this initial effort, now we are reshaping the performance testing of applications using Kubernetes, Gatling, a popular Scala based tests framework, Cortex/Prometheus, and Grafana.
Solution Architect with over 20 years of leadership and engineering experience in designing and implementing innovative solutions for complex business problems involving big data, real-time analytics and reporting solutions. Broad knowledge and leadership experience envisioning, designing, and executing technical solutions to business challenges. Solution domains include application development, enterprise integration, cloud infrastructure, on-premises virtualization, devops practices, and big data systems.