DockerCon, Flocker and the art of persistent storage

By Eric Carter | | Docker

DockerCon is bumping! This week we’ve had lots of conversations about real world workloads that attendees are running in containers. There is a good mix. Some only run apps that generate no data (or at least no data they care about) – BUT – others run things like NoSQL databases – and they DO care about the data. Of course my next question is – what are you doing for storage???

Enterprise customers who are building apps in containers want (actually NEED) a persistent storage solution. It remains a key challenge to production use of containers. Our friends at ClusterHQ released a 2016 version of Container Market Adoption survey done in conjunction with (Check out their blog on the topic to get the full report.) The adoption study takes on the issue of challenges in deployment. As you can see from the graph below, the biggest challenge cited around container deployment is delivering persistent storage for containers.

Now – it’s not all bad news. On the contrary. The same study shows that there was a 96% increase in container production usage over past year. That means real workloads driving real results. How do we keep that number growing? We have to address the persistence challenge.


Yesterday we announced our CloudScale Reference Architectures that included three blueprints, one of which focused on work we did with ClusterHQ to demonstrate how a highly available database, MongoDB in this instance, can be deployed in containers with Docker Datacenter using block storage volumes automatically provisioned care of Hedvig and Flocker. Beyond basic provisioning, the reference architecture showcases Flocker Storage Profiles that let you set gold, silver, or bronze differentiated service levels with Hedvig as the underlying storage platform for your containers. This is a simple, straightforward approach for putting the right app or microservice with the right volume type for best results. And of course, you can move your containers and not lose your data. Persistence problem solved!

We hope you find these reference architectures helpful in your journey with containers and software-defined storage. To download the paper, click on the button at the bottom of the page.