The goal with Mitaka is to help enable easier integration and management of all the projects in the OpenStack Big Tent model.
Mitaka, the first OpenStack cloud platform release of 2016, is now out after six months of development and the participation of a global community of 2,336 developers from 293 organizations.
OpenStack Mitaka is the 13th release from the open-source cloud effort, which Rackspace and NASA began in June 2010. The OpenStack community has grown significantly, and the technology is now used at some of the largest organizations, including Walmart, eBay, PayPal, AT&T, Comcast, IBM, Cisco, Dell, Intel and Hewlett Packard Enterprise.
Mitaka follows Liberty, which debuted
in October 2015 and was the first release to follow the so-called Big Tent model.
OpenStack is made up of a collection of projects that enable different capabilities. With the Big Tent, the number of projects considered to be part of OpenStack grew significantly.
The goal with Mitaka is to help enable easier integration and management of all the projects in the OpenStack Big Tent model by way of a project known as the OpenStack client. For most of OpenStack's history, each project had its own command line interface with which to manage a specific project's functions. There is also the OpenStack Horizon project that provides a unified dashboard, but that doesn't provide command line access.
"The OpenStack client is a command line client that unifies access across all the main projects," Jonathan Bryce, executive director of the OpenStack Foundation, told eWEEK
So if an administrator wants to create a user, a block storage device or a virtual server, or attach to a network, all those functions are now enabled in the single tool that is the OpenStack client. The OpenStack client provides a standardized set of commands, whereas previously, each project had its own command line client, Bryce said. He added that the OpenStack client can be run locally or in the cloud, and can be configured to control multiple OpenStack clouds.
Mark Collier, chief operating officer of the OpenStack Foundation, pointed to the benefits of OpenStack's being made up of many separate and disparate projects. For example, he said, the OpenStack Nova project specializes in compute, Keystone is about identity and access, and Neutron is about networking.
"There are all the different domain experts for different areas of cloud computing," Collier said. "But what has happened is that, from a cloud operator perspective, it's a disadvantage as it has been harder to configure OpenStack because there are so many options."
Part of the OpenStack client and the overall OpenStack Mitaka platform is an improved set-up and installation experience by defining default configurations. While an administrator can always choose from different configuration options, the goal with the default options is to provide guidance on the typical choices as made by the world's largest OpenStack cloud operators.
With the OpenStack client, there is now also what Collier referred to as common verbs. Instead of specific commands that might be unique to just one OpenStack project, with the OpenStack client, there is an effort to create common verbs that are common commands that enable capabilities across an OpenStack cloud.
Among the biggest OpenStack projects is the Nova compute effort, which has undergone a number of changes in the new release. One of those changes is the removal of Amazon Elastic Cloud Compute (EC2) API support, but that doesn't mean OpenStack is no longer compatible with Amazon's cloud.
"The idea is to move the Amazon functionality out of the core Nova project into a separate repository to be maintained, which is how we do it now for Amazon S3 and the OpenStack Swift storage project," Bryce explained.
As has been the case since OpenStack started, Amazon cloud API support will continue, he added.
Nova continues to benefit in the OpenStack Mitaka cycle from improvements in the Cells capability that was originally introduced in the OpenStack Grizzly release
in 2013. Cells enables the aggregation of multiple Nova compute modules to be managed by a single Nova API.
"One of the goals with Cells is to make Cells the standard way that a Nova deployment is set up," Bryce said. "So once you have an initial Nova deployment that needs to be scaled out, you can just create another cell and add it to your environment."
While Nova is one of the original OpenStack projects alongside the Swift storage project, with Mitaka, the Big Tent model brings along many other projects, including the Kuryr project for containers.
There will be a lot of discussion about Kuryr and containers at the upcoming OpenStack Summit in Austin, Texas, later this month, Collier said.
While new projects are always interesting, with Mitaka, the core of OpenStack has improved overall, Bryce said. "The Big Tent has released a flurry of innovation and has freed developers to take advantage of the collaborative process that the OpenStack Foundation has put together. One of the challenges that we have to keep addressing is that there is a very stable core for OpenStack, built around compute, storage and networking services, and it shouldn't get lost in the excitement around other features."
Bryce said: "OpenStack core services keep getting better and serve as the foundation for all of the other functionality."
Sean Michael Kerner is a senior editor at
InternetNews.com. Follow him on Twitter @TechJournalist