Docker's push into cloud-based development took shape this week in the form of a Docker Build Cloud service that offloads portions of developers' workflow to AWS.
Docker Build was already a widely used feature of the Docker Engine and Docker Desktop products, primarily on local developer workstations. It's used to package application code and other artifacts into containers for testing and, ultimately, shipping to production.
The preview of Docker Build updates released in October allowed it to burst into public cloud infrastructure and added team-level caching of resources to accelerate its performance. Now, the updated Docker Build is generally available as a hosted service, Docker Build Cloud, for commercial customers. Docker Build Cloud also integrates with continuous integration (CI) pipelines to offload builds within customers' existing application delivery workflows.
This update is a natural byproduct of cloud-native development, where software engineers must test application code in environments that closely resemble production, said Devin Dickerson, an analyst at Forrester Research.
"In the past, it was relatively simple to simulate production-like environments on a local machine, but today it is far more complex," Dickerson said. "Today's applications are often dependent on infrastructure and application services provided by the cloud or third parties, [and] it's often more seamless … to test these builds in the cloud."
Docker touts major performance improvements in the build process with the new cloud service, claiming some customers reduced process time from minutes to seconds. Docker Build Cloud is also integrated with tools its customers already use. Both these features of the new service appealed to one enterprise user.
"Anything that increases developer productivity is of interest, particularly when it comes to improving the performance of workflows that enable engineers to build and ship code to production quickly," said Stewart Powell, engineering manager at JW Player, a video hosting company in New York. "Technologies that allow engineers to have as much choice about the type of hardware they develop on are also attractive."
Docker Build Cloud adds cloud infrastructure support and resource caching to boost the performance of an essential element of Docker's containerized development architecture.
Docker Build Cloud value-adds and fine print
Some enterprise platform engineering teams build their own shared build environments in public clouds using broader CI/CD pipeline tools. There's potential value in a prepackaged managed service built into tools developers already use, however, Powell said.
Managed services eliminate a certain degree of guesswork and allow development teams to engage directly with a service without burdening DevOps or release engineering teams. There's also a certain amount of value that comes with a 'single source of truth.'
Stewart PowellEngineering manager, JW Player
"Managed services eliminate a certain degree of guesswork and allow development teams to engage directly with a service without burdening DevOps or release engineering teams," he said. "There's also a certain amount of value that comes with a 'single source of truth' -- when teams know where to look for their builds or know where to look for their SBOMs or image vulnerabilities, it drives increased engagement, visibility and ideally does so in a way that doesn't require compromise on availability, performance or administrative overhead."
With this first release of Docker Cloud Build, there is some fine print for enterprise users to consider. The service is available in only the US-East region of AWS; support for multiregion builds is on the roadmap, according to the company's documentation. The initial release of the service doesn't support connecting to private container registries behind a VPN and is multi-tenant only, though Docker is considering a single-tenant version, according to company officials.
For Powell's company, single-tenancy and private registries aren't strictly necessary, as his company uses HashiCorp's Vault to inject sensitive data into containers at runtime.
"We generally decouple any sensitive information from containers before they're shipped and rely on other mechanisms to inject sensitive data at runtime. Having to guard the registry behind a VPN is more than we require at this time," he said. "Many of our services are multi-tenant, and I believe that multi-tenancy services are attractive because of their ability to help optimize costs."
Existing Docker commercial subscribers will receive Docker Build Cloud minutes based on their subscription tier among the Docker Personal, Pro, Team and Business tiers. Subscribers can purchase Docker Build Cloud plans, which add 200 minutes per seat per month starting at $5 per seat per month, and additional minutes beyond those start at $0.05 per minute.
Docker's cloud roadmap and strategy
Docker Build Cloud and another new product, Docker Scout, are part of Docker's plan to move beyond local developer workstations and appeal to software engineering teams interested in using hybrid cloud infrastructure. Docker Scout was made generally available in October for container vulnerability scanning.
The broader strategy grew out of the company's pivot from container orchestration in 2019, and a shift in Docker Desktop pricing that caused some controversy in 2021. Customers have several alternatives for container build automation and developer tooling among commercial and open source products for cloud-native development workflows.
"Docker has been pretty successful in its pivot and efforts to focus on developer tooling" overall, said Katie Norton, an analyst at IDC. But Docker AI features previewed alongside Docker Build updates and Docker Scout in October will be key to continued adoption growth for the vendor, Norton predicted.
In IDC's Generative AI Adoption and Attitudes Survey released in May 2023, responding developers perceived generative AI as the No. 1 contributor to developer happiness. It ranked higher than options like automation, collaboration and integrated security, Norton said.
"In the survey, respondents indicated generative AI contributed to developer happiness primarily because of increased productivity and the opportunity to spend more time on high-value development tasks," she said.
Beth Pariseau, senior news writer at TechTarget, is an award-winning veteran of IT journalism. She can be reached at [email protected] or on Twitter @PariseauTT.
