The open-source application performance monitoring system Coroot is now generally available. Announced as an alternative to DataDog and NewRelic, the observability tool leverages eBPF for insights into system performance and supports monitoring heterogeneous infrastructures.
Coroot utilizes eBPF to collect telemetry data, including metrics, logs, and traces, providing a Service Map of the system. Additionally, Coroot includes predefined inspections that, according to the authors, can identify root causes for over 80% of issues without the need for any configuration. Peter Zaitsev, co-founder at Coroot and open-source advocate, writes:
Coroot looks at your complete environment, covering every component, service, and application, so you do not have information gaps. Coroot supports components deployed on Kubernetes, traditional VMs, bare-metal deployments, as well as cloud services such as DBaaS. Second, Coroot is very simple to deploy. We leverage modern, powerful instrumentation features provided by the Linux kernel, such as eBPF, allowing us to gather the most important information with zero configuration.
Source: Coroot blog
According to the documentation, Coroot creates a digital model of each system and audits every application using a set of predefined inspections based on industry best practices.
Dubbed "an open-source observability platform built for simplicity" and positioned as an alternative to DataDog and NewRelic, Coroot initially supported monitoring applications running only in Kubernetes clusters. With the release of version 1.0, it now offers an agent for virtual machines or bare-metal servers in addition to Kubernetes. Nikolay Sivko, founder and CEO at Coroot, explains:
Many Coroot users are happy using it to monitor apps in Kubernetes clusters, but they find it difficult to deploy Coroot in non-Kubernetes environments (...) Now it's easier to add standalone nodes to a Coroot instance in Kubernetes.
The new agent can be deployed as a systemd service or Docker container, sending telemetry data directly to the Coroot endpoint and eliminating the need to configure Prometheus to discover new nodes.
Source: Coroot blog
Moreover, with the GA version, deploying the entire stack is now achievable using Docker Compose or Docker Swarm. Sivko adds:
Even if you're not using Kubernetes, deploying Coroot on a dedicated VM gives you full system visibility without needing to switch to modern tech stacks.
Anton Peretrukhin, software engineer and deputy team leader at Flant, reviewed an earlier release, concluding:
Coroot appears to be a rather robust solution for monitoring apps and infrastructure in a simple-to-configure yet comprehensive fashion. It may become an essential SRE part for small businesses and teams who do not require complex systems and want to set up monitoring with SLOs, notifications, and request tracing quickly.
Coroot is not the sole open-source lightweight tool leveraging eBPF to create a visual network map of the services within a cluster, with Pixie and Caretta among different options available on GitHub.
Coroot has a Slack community and a live demo, and it is licensed under Apache 2.0.