Advanced Computing in the Age of AI | Friday, April 19, 2024

Deployment Hurdles Emerge as Containers Go Mainstream 

Application container deployment is accelerating as security and persistent storage concerns are addressed and enterprises seek to capitalize on the potential efficiencies, flexibility and long-term savings offered by distributed applications.

The latest measure of container deployment released this week by enterprise infrastructure vendor NetEnrich, San Jose, Calif., finds that 70 percent of IT professionals it surveyed are currently using containers while an additional 21 percent said they planned to soon.

The transition has not been without growing pains, with more than half (55 percent) of respondents citing problems integrating containers with existing infrastructure. Despite the growing availability of certified image registries and other security initiatives, container security also remains a concern for the majority of those polled (53 percent).

Other hurdles to container deployment identified in the NetEnrich survey included avoiding infrastructure "sprawl" and cost overruns along with a lack of experience with container technology.

Among the early challenges identified in the container survey was the difficulty of using leading container technologies such as Docker and the Kubernetes and Mesos cluster orchestrators. Two-thirds of those polled called these technologies "moderately challenging" to use while 21 percent characterized their use as "highly challenging."

Hence, more than half of respondents (58 percent) said they are relying on third-party management tools and services to get their container infrastructure up and running.

Along with Google-backed Kubernetes and Docker containers, the survey found that one-third of respondents are using containers tools from San Francisco-based upstart CoreOS, which offers an open-source operating system, container engine, key value store and a container platform integrated with Kubernetes called Tectonic. That platform designed to delivered Google-like services to enterprises. Meanwhile, 11 percent of those surveyed said they are using container tools from cloud specialist Joyent Inc.

Docker, which remains the dominant container technology, is being used with a combination of orchestration and other tools. More than half of those polled (53 percent) said they are using Apache and/or Docker Swarm, the native clustering tool. Other popular tools used along with Docker include Mesosphere's Datacenter Operating System and Apache Mesos, on which the Mesosphere is based.

Underlying the shift to distributing applications via containers are so-called micro-services, a lightweight approach for delivering individual applications. As container technology enters production environments, micro-services were found to be outpacing public cloud services by a widening margin. The NetEnrich survey found that 76 percent of respondents are using micro-services while 69 percent favored public cloud platforms. Meanwhile, more than half of respondents said they have embraced a DevOps approach to deploying applications via containers.

Container adoption is on the rise, noted Chris Joseph, NetEnrich's vice president of product management and marketing, adding in a statement: "Companies are still grappling with issues such as integration, security and management…."

Container specialists such as Joyent have stressed persistent storage as perhaps a bigger challenge in deploying containers than security. "The challenge isn't so much about how to persist state, but how to do so without compromising the agility and automation that we love about containerization in the first place," Joyent's Casey Bisson noted in a recent blog post.

About the author: George Leopold

George Leopold has written about science and technology for more than 30 years, focusing on electronics and aerospace technology. He previously served as executive editor of Electronic Engineering Times. Leopold is the author of "Calculated Risk: The Supersonic Life and Times of Gus Grissom" (Purdue University Press, 2016).

EnterpriseAI