CoScale
Performance Monitoring of Container Environments

The emergence of containers and microservices represents a new application architecture that facilitates the transition from monolithic legacy applications to smaller, discrete functions or processes. One of the key drivers is the need for an agile development method that promotes shorter release cycles, higher quality software, and improved user experience. In the wake of increasing container adoption, traditional APM tools have a hard time keeping up with these dynamic and scalable environments, built on new technologies such as Docker and Kubernetes. Adapted “container-native” tools are required to maintain visibility and control over the performance of these containerized environments. CoScale provides an integrated platform for full stack container monitoring. “CoScale is optimized for large and dynamic production environments, with lightweight non-intrusive container and application instrumentation,” explains Stijn Polfliet, CEO and co-founder of CoScale.

CoScale was conceptualized when Polfliet and his colleague Frederick Ryckbosch were completing their Ph.D. in Belgium. The duo was part of a group called Performance Lab and was researching on performance monitoring while focusing on innovative technology and new application architectures. “Perceiving the need for a solution that could effectively monitor software environments for container and microservices-based applications, we developed novel instrumentation techniques and machine learning algorithms to automate the analysis of performance data,” says Polfliet.

Utilizing their experience, the CoScale platform was designed to monitor the full stack of containerized applications. This full stack includes hosts, orchestrators, and services running in containers, along with its impact on user experience and key business metrics. Due to service and container-specific visualizations, performance or capacity problems can be detected before it affects end users. Using a lightweight agent, CoScale provides real-time insights across multiple hosts and containers while adding minimal overhead to the system, in contrast to traditional APM tools. It integrates with various container orchestrators including Kubernetes, Docker Swarm, and OpenShift and has plugins for specific applications that run inside the container, such as Nginx or Redis.

Along with the classic static and predictive alerting, CoScale leverages machine learning aimed at automatically detecting anomalies, sending alerts and correlating the abnormal behavior of metrics from services and containers. These alerts help teams to find performance issues and take pre-emptive measures to prevent any untoward incident from occurring. “We analyze what is normal behavior for a service based on historical data and identify any deviation from this standard baseline,” states Polfliet. While the solution is SaaS-based allowing fast and straightforward installation, CoScale also offers on-premise installation that is fully Dockerized.

Managing containers in production requires adapted tools. CoScale provides a container-native monitoring solution that goes beyond classical APM tools

Polfliet recounts how a client found CoScale to be the perfect partner for their performance monitoring needs. The client was deploying half a million containers per week, and the challenge was to provide visibility over the complex microservice applications and dynamic containers that were generating thousands of metrics per second. CoScale provided the capability to monitor the performance of individual microservices and application components inside containers, helping save time and resources in solving issues.

Over the past two years, CoScale has been building partnerships with many container companies including Docker, Red Hat, Microsoft, IBM, and HP. The company has been working closely with Red Hat to help boost the adoption of OpenShift and collaborating with Docker on multiple projects. As a member of the Microsoft Accelerator Program, CoScale was included in the official documentation of Azure for its container monitoring service. Moving forward, CoScale is expanding to make its presence felt across the globe and is working on an integration with OpenTracing for better visibility in microservices interactions. Deeply embedded in the container ecosystem, CoScale continues to maintain its competitive advantage with its automated platform for monitoring container environments.