- Ashesh Badani, vice president and general manager, OpenShift
- Matt Hicks, senior director, Engineering
- Joe Fernandes, vice president, Management and Security
- Clayton Coleman, lead engineer, OpenShift
In front of a standing-room-only crowd, a roundtable of Red Hat cloud experts discussed the new features in OpenShift Enterprise 3 by Red Hat and laid out the roadmap for where it’s going in the future.
After opening with a slide of all the OpenShift sessions at Summit this week (around 10), Badani contextualized the theme of the talk: “Software disrupts business. Every company is a software company, but do they realize it yet? If not, they’re at risk of getting left behind,” he said. This harkens back to Jim Whitehurst’s opening keynote–referencing a PWC survey of CEOs, he revealed that 36% say they’re planning on entering a different industry, and 61% think their competitors will compete outside their own industry. Organizations have to adapt to thrive in the software-driven economy.
Off to the roundtable discussion!
OpenShift Enterprise 3 by Red Hat is “Probably the biggest release we’ve done since we launched 4 years ago,” said Fernandes. Most of the work on this release came upstream from the OpenShift Origin community–by working with Docker and Kubernetes, a lot of container and orchestration technology is included the version 3 release, in addition to Project Atomic, middleware, and others.
So why Docker? “We hear a lot about Docker, but people are surprised to learn that OpenShift has been built around open source Linux container technology, since day 1,” said Hicks.
Badani discussed the new benefits of OpenShift 3, which allow developers and IT operations engineers to:
- Deploy apps faster. This leads to a better developer experience.
- Lower infrastructure costs. “If we were paying Amazon for 2.5m VMs, that would be a rough monthly check to cut,” said Hicks.
- Increase operational efficiency. Fewer machines to secure and manage
“We’ve been asked a lot,” said Hicks, “Why has this really taken off in the last 12 months or so? Docker did a neat thing by adding a packaging format. On the infrastructure side, you want to have the portability of a packaging format where your code can move.”
With that packaging format and popularity came a really strong ecosystem. Docker’s packaging format has drawn “practically every technology known to man,” said Hicks. So now, you can take any Docker format and run it in OpenShift 3.
Shifting to standards, the roundtable discussed how Red Hat works with the open source community to drive standards for containerization–from isolation with Linux containers, container format with Docker, orchestration with Kubernetes, and container discovery via registry.
The roundtable traded off with demonstrations from Coleman, who illustrated the impressive speed of getting applications up and running from the Docker hub–all with built-in security, like SELinux, by default.
So why Kubernetes, asked Badani: “We like to think of ourselves as pure technologists, but this is also a fashionable industry, so is Kubernetes the new skinny jeans?”
Fernandes took this one: “We know applications don’t run in a single container. They run in multiple containers. So something needs to orchestrate and manage all of them, and you need more than Docker to do this. So, we started talking to Google and joined in on their open source Kubernetes project.”
Kubernetes provides orchestration of app services that span multiple containers, implements scheduling and placement of containers across multiple hosts, and offers management of container health to detect and automatically restart containers on failure. “We don’t want to have OpenShift-only solutions if we see a better one out there. Google said they were running more than 2B containers every single week,” said Fernandes.
So what’s next for OpenShift? The panel hinted at some of the new capabilities coming in the next year, including:
- Container management capability with Red Hat CloudForms.
- Expanded OpenStack integration.
- Additional JBoss Middleware services.