So you have moved all your applications to Docker and have begun enjoying all the fruits of lightweight and fast-to-deploy containers.
That’s great, but once you have multiple containers spread across multiple nodes, you’ll need to find a way to track their health, storage, CPU, and memory usage, network load, etc.
To track these metrics, you need an efficient monitoring solution and some backend store to keep your container data for subsequent analysis and processing. Managing thousands of Docker containers in production made our team here at Qbox quickly realize that Docker container monitoring is a valuable addition to our cluster management process.
In a previous article, we discussed how to use Metricbeat to ship metrics from Kubernetes. Now, it’s time to share our experience of using Metricbeat to monitor bare Docker containers and shipping container data to Elasticsearch and Kibana. This knowledge may be useful for developers and administrators who manage Docker containers without orchestration. Let’s get started!