At the OpenStack Summit Tokyo conference today, Midokura announced that it has integrated its network virtualization software for OpenStack environments with the Kuryr project that enables Docker containers to be networked together.
In addition, Midokura announced that its Midokura Enterprise MidoNet (MEM) software now supports the latest Liberty edition of OpenStack alongside previous Kilo and Juno releases of OpenStack.
MEM provides an alternative to the Open vSwitch software that is considered the default network virtualization provided under the Neutron component of OpenStack. However, given some of the limitations associated with Open vSwitch, there are now multiple plug-ins that many organizations that are deploying OpenStack have opted to deploy instead of Open vSwitch. As one of those options, Midokura has now extended MEM in a way that enables Docker containers to invoke a Kuryr gateway to access Neutron networking services.
As two of the most significant emerging technologies in the enterprise, Susan Wu, director of technical marketing for Midokura, says IT organizations that are embracing both OpenStack and Docker containers are rallying around a Magnum project that integrates the Kubernetes orchestration framework for containers with OpenStack. The Kuryr project is the point at which integration between Docker containers and OpenStack will take place at the network level.
While OpenStack in general is still considered difficult to deploy, and after five years of development still has issues in terms of scaling, work on the open source cloud management framework continues to progress. That doesn’t mean that IT organizations are going to replace commercial network virtualization offerings from VMware or Microsoft tomorrow. But in those IT organizations that have a modest amount of engineering talent, OpenStack clearly represents a less expensive alternative that over time will only continue to get more robust.