Facebook and Equinix this week at the Open Compute Project 2016 Summit announced a partnership under the two companies plan to collaborate on the development of a new platform optimized to run containers.
Based on the Open Compute Project (OCP) platform initially created by Facebook and the open source Mesosphere Data Center Operating System, Equinix CTO Ihab Tarazi says the goal is to create a truly hyperconverged platform that eliminates compute, storage and networking silos. While some progress has been made in that regard via the emergence of hyperconverged appliances and rack systems, Tarazi says this new project seeks to integrate all those system services inside a single software-defined environment.
In the meantime, Tarazi says Equinix will be adding support for systems based on OCP within the data center that make up the Equinix International Business Exchange. Tarazi says that significant because it creates a mechanism through which those systems can be integrated with other platforms hosted in Equinix data centers. At the moment, Equinix says more than 1,100 network providers, as well as 500-plus cloud service providers, make use of those data centers.
In general, the development of a highly integrated operating system by Facebook and Equinix could have profound implications for IT operations teams. In most IT environments today compute, storage and networking are managed in isolation by IT staff with dedicated expertise in each area. As IT evolves, however, a more holistic approach to managing those resources is starting to emerge. The benefits of that approach, says Tarazi, will primarily manifest themselves in the form of increased IT agility. Instead of waiting weeks to provision storage resources for a container, IT operations teams will be able to instantly provision compute, storage and networking resource at the same time.
In addition, Tarzi notes that consolidating all those functions on standard industry servers will dramatically reduce the amount of energy being consumed in the data center.
Designed from the ground up to support containers, this new platform should also help IT operations close the gap in cadence between how applications are now being developed and the provisioning of IT resources required to support them. Increasingly, developers are making use of agile development methodologies and microservices architectures to not only develop application faster, but also deliver more updates to those applications in production environments. A silo approach to managing compute, storage and networking resources simply can’t keep pace with that rate of application development.
In most IT environments that means it’s only a matter of time before all those container applications exacerbate an already complex DevOps equation. In fact, the arrival of those container applications should finally force the DevOps issue in many traditional enterprise IT environments.
Naturally, it will take a while for Facebook and Equnix to build this new operating system. It’s also safe to assume that providers of other operating systems have similar ambitions. In the meantime, IT organizations would be well advised to start thinking about realigning their internal IT structures today in anticipation of where data center technologies will be evolving in the years ahead.