DockerCon19 Takeaway: Operationalizing Container Environments is the #1 Problem

At DockerCon19, a number of companies, including Docker, announced new and upcoming products that simplify the deployment and cluster bringup of running containerized applications across public and private cloud environments. Their value propositions are simple:

“We’ll help you run your containerized apps anywhere.”

But what they really mean is:

“We’ll help get your Kubernetes clusters up and running, and help deploy your container images into the clusters, but then you’re on your own.”

All the DockerCon attendees I talked to agreed that the vendor community isn’t delivering on its core promise of letting companies run containerized apps anywhere. Customers believe that off-the-shelf solutions exist for “day-zero” tasks, such as container cluster bringup and deployment pipeline management. But everything else needed to operationalize clusters and manage them on an ongoing basis is their responsibility.

Customers are Building Internal Platforms

So in effect, customers are building internal “platforms” that developers can leverage to deploy and operate containerized apps across clusters. These platforms deliver the following:

  • Day-Zero Capabilities: Cluster bringup and container image distribution via CI/CD pipelines.

  • Day-One Capabilities: Cluster and application operations including ongoing Kubernetes administration across multiple clusters in multiple environments, persistent volume management, log and metrics aggregation, and secure distribution of secrets and configuration management.

  • Day-Two Capabilities: Application lifecycle management tasks such as cross-cluster image upgrades, application debugging, configuration changes, and state sharing.

Building platforms that automate day-one and day-two operations for container environments is a non-trivial task. Most companies that have undertaken this endeavor end up leveraging anywhere between 5 and 20 engineers who are experts in all things Kubernetes, infrastructure and networking. Even if you have access to such engineering talent and can afford to invest the time and resources into building, do you really want to have them build a platform that will require perpetual, expensive care and feeding? Wouldn’t it be better for you to consume all of these tools through a SaaS service, so your engineering teams can focus on building meaningful value for your business?

Automating Container Management with Rafay

We believe that every company out there should be able to deploy, operate and manage lifecycle for containerized applications in any environment by leveraging a platform like Rafay’s. You should be able to have your containerized apps up and running today in any cloud or on premises, and all aspects of the container management & operations workflow – from day-zero to day-one and beyond – should be delivered to you as a service. Feel free to sign up for a free evaluation to experience such a platform at your own pace.

Questions? Comments? Strenuous Objections? Please feel free to get in touch or email me directly.