Software-as-a-Service (SaaS)-based monitoring provider LogicMonitor implemented new capabilities into its platform to help DevOps professionals peer into microservices and containerized applications.
LogicMonitor’s self-named performance monitoring platform is based on a SaaS architecture and relies on automation to collect performance data from a variety of environments. This includes on-premises, cloud, and hybrid architectures such as Amazon Web Services (AWS), Microsoft Azure, servers, storage, networks, virtualization, applications, websites — and now Kubernetes, microservices, and containerized applications.
As Kubernetes and other containerized environments grow increasingly popular, monitoring these environments has created new challenges for enterprises. These architectures are more difficult to monitor than more traditional environments because of the increase in application data, and they are more complex than what traditional monitoring tools are equipped to handle.
Prior to today's launch, LogicMonitor offered Docker container monitoring. "With increasing adoption of Kubernetes to orchestrate these containers, we saw a need to provide a more comprehensive container monitoring solution that integrates with orchestration," said Sarah Terry, manager of product management at LogicMonitor.
LogicMonitor’s new Kubernetes monitoring tool provides event-based Kubernetes monitoring. According to the company, the tool will monitor as an enterprise splits a monolithic service into microservices orchestrated with Kubernetes. The tool does this by automatically adding and removing cluster resources from the monitoring platform.
The Kubernetes monitoring also eliminates the need to have an agent on every node. According to Terry, this is because the tool relies on two applications running in the cluster: one as a pod for monitoring and the other as a pod for discovery. The discovery application reads the Kubernetes event stream and leverages LogicMonitor's existing API to keep everything up to date. Data is collected from Kubernetes nodes, pods, services, and containers using the Kubernetes API.
Other features of the tool include full performance and health metrics at both the cluster and application level and provides visibility into underutilized resources for increased optimization of resource usage. This includes providing insight on CPU and memory.
The second tool that LogicMonitor added to its monitoring platform provides service-oriented monitoring. What this means is that enterprises can now group resources that support a common application, service, or cluster into one group. This allows for better alerting and monitoring for the overall application or service, regardless of how the underlying resources change and evolve.
The need for this kind of tool was a result of the increasing complexity in IT systems. LogicMonitor "saw a need to allow better monitoring and alerting for services and applications backed by many resources. Especially when those supporting resources are ephemeral and dynamic," said Terry.
The service-oriented monitoring tool also generates service topology automatically to compare against the other monitoring and alerting data in a company’s environment. Currently, the grouping of applications, services, and clusters is not automated — customers provide criteria for how they want them grouped and then LogicMonitor finds the resources with that label and groups them into logical services.