OpenDaylight's Hydrogen Poses Challenge to Proprietary Network

NEWS ANALYSIS: OpenDaylight's Hydrogen Software-Defined Networking platform is the latest challenge to the vendors of proprietary network infrastructure products.

The Open Compute initiative is designed to upset the proprietary data center model in place for about the last 50 years with common designs and commodity components. The Open Daylight plan intends to upset the proprietary networking infrastructure. And then there is OpenStack, CloudStack and essentially open anything to do with computing.

We are in the era when the idea of vendors building tightly controlled stacks of hardware and software that promised superior performance in return for your undying loyalty and IT dollars is falling before the concept of openness everywhere.

The most recent example is this week's release of the first software-defined network offering from the OpenDaylight Foundation, which is a collaborative project created under the aegis of the Linux Foundation. OpenDaylight was born in controversy just about a year ago when it was introduced and then berated for being too heavily weighted to the very networking and enterprise tech vendors that open software was supposed to challenge.

To its credit the Linux foundation appears to have successfully negotiated a path between the enterprise networking stalwarts like Cisco and new generation of network infrastructure startups with the development of its open Hydrogen SDN platform.

Hydrogen is a developer-oriented first release—more proof of concept than finished piece of software—aimed at providing a pathway to SDN and network function virtualization (NFV). While SDN and NFV have been compared to the server and storage open projects that are upending the compute and storage chunks of the data center, networking is different.

Servers were virtualized in recognition that a lot of server capacity was underutilized. Storage was virtualized in an attempt to bring some order to the storage chaos inherent in many data centers. Networks tend to run at maximum capacity and need the capability to scale instantly and assure security over a far-flung enterprise network of fixed and mobile devices.

The skills required to design, deploy and manage open data centers running open software and often tied to outside cloud computing resources are very different from operating in a locked down data center. The idea of securing a data center by building a digital moat surrounding the center has fallen to the idea of encrypting everything, everywhere and keeping the encryption keys in your pocket rather than handing them to a vendor.

The idea of allowing a vendor to populate your data center with proprietary servers and storage is falling to open compute specified designs and storage, which either adheres to one of several open models including solid state storage as part of the compute operations or replaces on-site storage with cloud-based storage. CIOs recognize that the skills needed in this open era may not exist in house and may not exist within the traditional consulting organizations used to upgrade hardware and software infrastructure.

I'm guessing the skill shortage will be the biggest impediment to enterprises hoping to gain the cost savings, flexibility and return to customer control that the many open projects promise. The techies with the required skill sets are working at Internet start-ups and learning about the new infrastructure models by studying how vendors including Facebook and Google built their computing empires by eschewing the old rules of data center design.

CIOs who are reading about all the open projects and feeling a bit overwhelmed by the thought of revamping their data centers and revamping their concepts of how software and hardware should be purchased, deployed and managed would be better off first thinking about how they can find, hire and retrain their existing technology staffs to embrace the unfolding open era of computing.

Eric Lundquist is a technology analyst at Ziff Brothers Investments, a private investment firm. Lundquist, who was editor-in-chief at eWEEK (previously PC WEEK) from 1996-2008 authored this article for eWEEK to share his thoughts on technology, products and services. No investment advice is offered in this article. All duties are disclaimed. Lundquist works separately for a private investment firm which may at any time invest in companies whose products are discussed in this article and no disclosure of securities transactions will be made.