You’ve heard a lot from us about container growth in the last year and how Docker alone grew its user base 5x between 2014 and 2015. Why so popular? Containers enable companies to be more agile in moving through development, test, and production environments. In doing so, it reduces cost and improves time to revenue by allowing these features to be brought to market in an efficient and timely manner.
Built to be weightless, stateless tools just megabytes in size, containers enable quick application launch. The popularity of container managers such as Docker, Kubernetes, and Mesosphere and their easy-to-use tools have contributed to the growth in container deployment. Now companies deploy not just tens of containers to accelerate their development process but hundreds. In building hundreds of weightless, stateless containers, we help solve the need to quickly integrate new product features into production, but data management within these containers can become difficult, especially with the unpredictable lifecycles that often accompany containers. As reflected in the recent survey from Portworx, IT managers stated one of their biggest concerns with container storage was loss of data. So how do we address data which needs to persist?
Luckily, today’s container management applications offer a variety of solutions for this including: 1.) storage plug-ins, 2.) data volume containers, and 3.) local directory mounts.
Storage plug-ins are proving to be the most reliable and manageable option for persistent storage. NetApp has worked with Docker, Kubernetes, and Mesosphere to provide standard volume plug-ins to make the experience even easier to navigate with features built to allow you to create, delete, and mount persistent volumes all from the container manager itself.
To read more about the advantages and disadvantages of all storage management options and better understand the differences between containers and virtual machines, see our article in App Developer Magazine: Demystifying persistent storage: As containers die, data lives on.