Slideshow The elements of a successful data container strategy

Published
  • March 27 2017, 5:40am EDT
8 Images Total

The elements of a successful data container strategy

Line-of-business managers may not recognize the benefits of data containers, especially since they represent an architecture and not a specific solution, explains Dale Kim, senior director of industry solutions at MapR. Kim offers seven tips on how data professionals can have a successful data container strategy.

Identify your key goals for your organization

“Moving forward with containers probably means justifying the effort, and that starts with the business impact," Kim says. "Better hardware utilization, scalability, fault tolerance, and other virtualization benefits should be emphasized as part of your container strategy.”

Content Continues Below


Fill the platform gaps in your architecture

“A container-based architecture is not always as simple as putting applications into containers,” Kim explains. “You need a platform that supports distributed data persistence to run stateful containers anywhere. It’s worth exploring the technology vendors that facilitate the deployment of containers.”

Get your IT operations team on board

“IT ops teams might reject containerized applications because of the perceived extra work to manage them, especially for stately applications,” Kim says. “If you’ve identified your key goals, they will often help all stakeholders to see the advantages of a container-based architecture.”

Deploy in phases

“Even if you have a long-term plan for containers, you don’t have to implement it all at once,” Kim says. “Look for the early gains like running existing applications in containers to get isolation and load balancing flexibility. Even simply starting with a dev/test environment that promotes agility can be a step that later evolves into bigger gains.”

Content Continues Below


Incorporate new technologies into your architecture

“As you continue your journey with containers, you will likely build a new set of applications that ideally should run on newer technologies like NoSQL and event streaming engines,” Kim explains. “These technologies are especially useful for modern data architectures that solve big data challenges. Streaming engines are especially useful as a lightweight means of transporting data and messages from one container application to another.”

Consider microservices

“Containers are ideal for packaging applications as part of a microservices architecture,” Kim says. “Microservices architectures are useful because they promote focused, easier-to-maintain modules that work together to enable scale, parallelism, reusability, and fault tolerance. Once you understand the advantages of microservices and implement them, containers become a natural way to package and deploy them.”

Think hybrid cloud and inter-cloud

“Discussions on the cloud should no longer be about on-premises versus cloud,” Kim stresses. “Rather, the discussion should be about how to get value out of a multi-site architecture. The portability of containers makes them ideal for deploying anywhere, and containers will simplify that option for situations where it makes sense.”