Docker is gaining popularity in DevOps circles for its portability and ease of deployment as well as for its scalability. More and more critical applications and services are running in Docker containers, necessitating the monitoring of Docker performance to ensure that these applications and services meet usability and performance expectations. Effective Docker monitoring reveals resource consumption by each Docker container. Docker monitoring metrics include CPU consumption, memory used, network I/O, and response time for each request. Depending on the use case, Docker containers can start up and shut down very quickly. Docker container monitoring tools need to be engineered to monitor this short lifecycle of Docker containers.
Dynatrace is a market leader in performance management solutions for application monitoring, user experience management, infrastructure monitoring, and network monitoring. It provides monitoring support for newer technologies like Docker container monitoring, NGINX monitoring, and Node.js monitoring, and it also offers load testing tools for web applications. Dynatrace performs Docker monitoring using a lightweight agent that is installed with each Docker container. The agent collects CPU, memory, and network statistics from Docker containers. Dynatrace collates and presents this data in the Dynatrace Docker monitoring dashboard which shows software performance metrics for all containers being monitored and the containers’ host health in a single pane of glass. Using Dynatrace to monitor Docker containers facilitates the quick detection of performance hotspots in Docker containers thereby minimizing impact on user experience.
In addition to monitoring individual Docker containers, Dynatrace also monitors Docker container clusters and their hosts. Docker container clusters increase the availability and redundancy of applications running in Docker containers. Container cluster managers such as Kubernetes enable Docker clusters to scale up or down on demand and roll out updates without service interruption. Dynatrace Docker monitoring enables the visualization of the health and performance dynamics of Docker clusters. Cluster metrics are presented visually in the Dynatrace Docker cluster monitoring dashboard. This dashboard depicts the distribution of containers per host in the cluster, the response times of transactions at any point in time, and the distribution of slow and fast transactions.
In the case of multi-tier clusters with front-end and back-end tiers, Dynatrace Docker monitoring visualizes the balancing of requests by load balancers between the containers in the front-end and back-end tiers. Variations of thread counts and resource utilization between containers are also shown. With these metrics on the dashboard, unusual spikes in response time and CPU and memory utilization can be easily seen and correlated to determine where problems may exist.
Use Dynatrace Docker monitoring to gain comprehensive insight into Docker containers and Docker clusters to solve Docker performance issues. Supplement it with other monitoring solutions from Dynatrace such as SQL server monitoring and network monitoring to obtain a full view of application and infrastructure performance in the data center.