Netdata Solutions

Container monitoring with Netdata

Docker logo
Kubernetes logo
LXC logo

Challenges of container monitoring

Many businesses unintentionally solve one problem with containerization while creating a new one. They can deploy easily using an orchestration tool, and have nice features like automatic horizontal scaling, but this strategy makes their container monitoring look like a black hole. To alleviate this problem, many IT, Ops, DevOps, or SRE teams undertake time-consuming projects to customize their existing Docker images or create new containerized applications that properly expose metrics. If that doesn’t work, they onboard onto expensive SaaS products that charge per container and provide less granular metrics.

And in many cases, the collected metrics only reveal the resource utilization or network throughput of individual containers, leaving the team in the dark about the containerized application’s health and performance. Without per-application metrics, such as the request time for a containerized Nginx web server or the queued messages in RabbitMQ running on a Docker container, teams cannot make the fast, data-driven decisions they need to resolve problems and be proactive about maintaining their infrastructure’s uptime and performance.

Monitoring an Apache web server container with Netdata's custom dashboards

How Netdata helps you monitor containers across your infrastructure

The Netdata Agent automatically detects and maps the status of cgroups on a Linux system, the industry-standard for performant and secure containerized workloads. It collects metrics from any cgroups-compatible container, such as Docker, LXC, LXD, Libvirt, systemd-nspawn, and more, and the Agent itself runs in any environment, whether that’s bare metal, a virtual machine (VM) in the Cloud, or within orchestrated microservice deployments. It’s also infinitely scalable as metrics are stored on individual nodes, not centralized in a complex data lake.

Instantly, and with zero configuration, Netdata collects and visualizes resource utilization metrics for every container on that node. It also autodiscovers the specific application running inside of that container, and immediately applies one of the 300+ integrations to collect dozens of relevant per-second metrics. In minutes, a team can go from having zero visibility into their environment’s health and performance to having actionable, per-second metrics from containerized web servers, databases, message queues, custom applications, and much more. Netdata’s built-in health watchdog simultaneously analyzes every container and its workload to trigger preconfigured alarms to dozens of popular notification platforms.

Netdata Cloud brings that same instant value to the entire infrastructure of containerized environments. With unified multi-node views, composite charts, and custom dashboards, technical teams have every tool and visualization necessary to monitor the status of hundreds or thousands of individual containers, and the applications they run, from a single pane of glass. Metric Correlations simplifies the process of performing root cause analysis, and helps engineers focus on potential issues in the ways different containerized workloads interact with each other.

Key container performance metrics

  • Instantly view CPU, RAM, disk I/O, and network usage of Docker containers and Linux containers to discover bottlenecked or limited applications.
  • Collect per-second metrics from containerized services or applications, whether they’re web servers, databases, message brokers, and more, for full visibility into your container environment.
  • Monitor container agents like dockerd, Kubelet, and kube-proxy to monitor orchestrated Docker, Docker Swarm, Kubernetes infrastructure.

Wordpress container monitoring with Netdata

The impact of monitoring containers in real time

With every per-second application metric, from every container, from every node in the infrastructure at their disposal, technical teams have all the information they need to make data-driven decisions about their containerized infrastructure. There’s no more visibility gaps, and they also didn’t have to spend weeks configuring a different monitoring solution that provides only part of the picture in low resolution.

With Netdata, tech-oriented startups or SMEs can now feel comfortable taking advantage of containerized deployments without worrying over how they’ll monitor every component. Development teams can start to see how their deployments affect individual containers and how they interact across host nodes. DevOps teams can build per-second metrics, composite charts, and Metrics Correlation into their incident response methodologies to drive down MTTR and make proactive changes to containers to prevent similar incidents in the future.