GCE customers are now able to do things like establish virtual private Layer 3 networks and assign static public IP addresses to instances, James said. Connecting networks will also become possible. And a load-balancing service is on the way “as part of the native fabric for Google Compute Engine,” James said.

Developers interested in trying out GCE load balancing can fill out a form to do so. Developers can also sign up for early access to all emerging Google Cloud Platform features.

The load balancing and routing services are the sorts of things that could help more businesses make the decision to try real projects on the newly publicly available Infrastructure as a Service (IaaS) piece of the Google Cloud Platform.

And the new capabilities move Google a few steps closer campaign to becoming a top, widely used IaaS provider — if not one day bigger than Amazon Web Services then at least No. 2. That position is already feasible for Google as it is.

Related research and analysis from GigaOM Pro:
Subscriber content. Sign up for a free trial.

• Takeaways from the second quarter in cloud and data
• The promise of SDNs in the enterprise
• A near-term outlook for big data

Juniper says its Contrail software will be widely adoptable across devices from many vendors because it supports common protocols such as BGP and XMPP, although not OpenFlow. The controller “doesn’t have a dependency on a particular protocol like OpenFlow,” said Brad Brooks, vice president of marketing and strategy at Juniper. So network admins can run the software on much of their existing Juniper or Cisco gear. Still, the architecture of the Contrail software is flexible and could support OpenFlow in the future, said Jennifer Lin, Juniper’s senior director of product management (pictured).

As it prepares to sell the controller as part of a new software-licensing model for the company emphasizing the ability to pay for it independent of hardware, Juniper is following through on a long-term strategic shift. The strategy could help Juniper stay relevant with software that can work with current and forthcoming hardware, preventing commoditization while enabling centralized configuration and programmability and applications higher up the stack.

Those objectives are increasingly important as data centers become more complex and fluid with virtualization, multitenancy and rapid scale-out. Juniper wants to run networks that allow those things to happen easily — and stay in the hardware business, too.

Related research and analysis from GigaOM Pro:
Subscriber content. Sign up for a free trial.

• WAN design for the cloud age
• The Structure 50: The Top 50 Cloud Innovators
• Infrastructure Q1: IaaS Comes Down to Earth; Big Data Takes Flight

According to a job description posted on the careers section of Facebook, the social-networking company looks like it wants to deploy software-defined networks at production scale. The right person for the open software-engineer job will head up efforts around “designing and implementing control plane systems for our network.” The job “may involve evaluating third party and open source software,” among other possible tasks. That could mean Facebook wants to explore existing proprietary controllers and other software components from vendors such as Big Switch, alongside the code that will emerge from the OpenDaylight Project vendor-led consortium.

The proposed Facebook SDN hire suggests that Facebook is now ready to go beyond the early-stage work that took place at the beginning of the year and perhaps earlier.

As for Amazon, the retailer and cloud infrastructure provider earlier this month posted openings for an “SDN Software Development Engineer” and a “Systems Engineer” focused on “EC2 High Performance Network Virtualization.”

Amazon has a couple of ideas up its sleeve, judging by the job descriptions. First, it wants to improve the performance of its virtualized networks, to make them perform as well as networks running on bare metal. Then, Amazon wants to make multiple SDN features available to AWS customers. And because AWS is so popular and continues to grow, that could have some neat implications for customers interested in making their bandwidth as scalable as their compute and storage resources.

Amazon and Facebook spokespeople were not able to provide any additional detail. But Jay Parikh, Facebook’s vice president of infrastructure engineering, and Werner Vogels, chief technology officer at Amazon, will talk at GigaOM’s Structure conference in San Francisco on June 19, and maybe then we’ll hear more information.

If the new Facebook and Amazon SDN hires move quickly, next year’s Open Networking Summit could be filled with some of the biggest use cases yet. Even so, wide-scale use of SDN among enterprises might come later next year.

Related research and analysis from GigaOM Pro:
Subscriber content. Sign up for a free trial.

• The promise of SDNs in the enterprise
• A near-term outlook for big data
• NewNet Q2: Google closes the quarter with a bang

As is the case with software-defined networks, the Jeda controller software will separate the control plane from from the data plane. It will run on a virtualized server and take charge of the intelligence that would otherwise reside on a switch. SDN can have a wide range of benefits, although it generally can permit more programmability and elasticity of networks. For Jeda, virtualizing storage-area networks could allow customers to generate and disable those networks in response to changes in demand.

Other companies other than Newport Beach, Calif.-based Jeda are likewise looking to make software-defined storage a reality, including Convergent.io, ScaleIO and SwiftStack. Approaches differ among those entities, but they share the goal of bringing storage up to the level of programmability of compute resources and, increasingly, networks.

Related research and analysis from GigaOM Pro:
Subscriber content. Sign up for a free trial.

• The promise of SDNs in the enterprise
• Takeaways from the second quarter in cloud and data
• A near-term outlook for big data

• One Convergence, based in Santa Clara, Calif., is playing the easy-to-use card with the OpenStack-based network-virtualization overlay software it was demonstrating at the conference. It will launch in the coming months. The One Convergence offering is similar in some ways to Nicira’s Network Virtualization Platform, said Roshan Gudapati (pictured), vice president of marketing and sales. “As traffic increases, you can bring up more and more (network) instances,” he said. It’s just as easy to hit the delete button on a network, as Gudapati demonstrated at his company’s booth. (Gudapati later said the One Convergence software is similar in some ways to the approach Nicira had taken in the past.)

• Saisei Networks, based in Sunnyvale, Calif., is working on software for traffic management — the act of setting and enforcing policies that can limit the traffic over a network. The idea is to deliver a product that eliminates the need for purpose-built legacy hardware and works at carrier grade, said Dave Newman, Saisei’s co-founder and chief operating officer.
• EstiNet, based in Hsinchu, Taiwan, has a network simulator to see what would happen if an OpenFlow controller were to send packets across multiple OpenFlow-enabled switches. The software visualizes the route of packets across switches and lets customers figure out the most efficient path and simulate the failure of switches in a network.
• Accelera Mobile Broadband Inc. is combining two favorite GigaOM topics, heterogeneous mobile networks (het-net) and network virtualization. The company, which was founded in 2009, is trying to use network virtualization to set policies and take action across the next generation of wireless networks that combine Wi-Fi, LTE, 3G technologies and sometimes even older networks.

As conference attendees heard this week, investors are still looking for SDN startups to fund. Perhaps these companies will help meet the investor demand. Perhaps still more startups need to start up.

This post was updated at 5:31 p.m. PT to add context to Gudapati’s statements at the conference.

Related research and analysis from GigaOM Pro:
Subscriber content. Sign up for a free trial.

• The promise of SDNs in the enterprise
• Takeaways from the second quarter in cloud and data
• A near-term outlook for big data

Bruce Davie, principal engineer at VMware, took the stage with a provocative message: SDN has promised many things, but most of those things are being delivered with network virtualization, and SDN isn’t necessary. Network virtualization, of course, is the preferred term for Nicira, which VMware bought last year for $1.26 billion and subsumed into its own product last month. Davie said he often sees people claiming that SDN can let network administrators do application-level programming of networks, easily provision and manage networks, improve performance and add bandwidth. But Davie has his doubts about that.

After Davie’s address, I ate lunch with a bunch of network engineers from a Fortune 100 company who are getting pressure from executives to lower capital and operational expenditures with SDN. But they’re just not sure what to do. They need to find something fast, but they also don’t want to bring any more risk into their data center. That’s why implementing open-source code on the way from the OpenDaylight Project on top of white-label network gear might not be as good a choice as sticking with expensive hardware with reliable support from a legacy vendor. These guys already hear different definitions from different vendors. When they hear that, no, SDN is not going to make networks programmable, they only become more uncertain of what to choose.

It’s a little easier to see why about half of more than 200 enterprise network administrators surveyed earlier this year couldn’t identify the correct definition of SDN. And so it seems we still are arguing over definitions, that we are still in the hype cycle, that we are still trudging through the FUD phase.

Related research and analysis from GigaOM Pro:
Subscriber content. Sign up for a free trial.

• The promise of SDNs in the enterprise
• Takeaways from the second quarter in cloud and data
• Microsoft Azure: What It Is, What It Costs and Who Should Care

It’s true that if useful vendor-agnostic code for many networking components is to come out of OpenDaylight, participants will have to clear several hurdles. Some of the 18 companies sponsoring OpenDaylight, such as Juniper and Cisco, compete with each other, and developers might end up having to maintain controller code that works best with certain kinds of networking appliances. While the tech press has brought up that possibility, it’s not completely unfounded; a Big Switch spokesperson has called into question how Cisco specifically will interact with everyone else when it comes to giving code the OpenDaylight stamp.

The lack of customer leadership in OpenDaylight — unlike, say, the Open Networking Foundation (ONF), which has board members from Yahoo, Goldman Sachs and other non-vendors — has been another area of contention. Although it might not have been for lack of trying. When executives from Cisco and IBM were organizing the OpenDaylight Project a few months ago, they reached out to Google and NTT, but neither company got on board, according to an engineer involved in the project. Perhaps companies other than network vendors could jump into the project in the coming months, though.

The role of the ONF, which nurtures the development of the OpenFlow networking protocol, is another open question. In public remarks at the Open Networking Summit on Tuesday, ONF Executive Director Dan Pitt said OpenFlow is a “substrate (that) allows you to build things like open-source software.” He said he didn’t think OpenDaylight would have been possible if there hadn’t been “something to build upon.” Asked if he or the ONF will get involved with OpenDaylight, Pitt said he had no information along those lines.

Even so, the OpenDaylight Project is “much more of a meritocracy” than the ONF, said Dave Husak, founder and CEO of Plexxi, which has paid five digits to be a silver OpenDaylight member. He views OpenDaylight as a vehicle for promoting Plexxi algorithms and application programming interfaces, which Plexxi will contirbute to the project. At the same time, OpenDaylight could surely benefit companies that seek to do more with their networks.

As much as I might want to predict the future and approximate the outcome of OpenDaylight, I’m afraid I can’t do that, and I haven’t found anyone here who can. They all say they’ll have to wait and see. And so will I.

Related research and analysis from GigaOM Pro:
Subscriber content. Sign up for a free trial.

• The promise of SDNs in the enterprise
• Takeaways from the second quarter in cloud and data
• A near-term outlook for big data

There was Shirish Sathaye, a general partner at Khosla Ventures, calling out at this year’s Open Networking Summit on Tuesday for more software-defined networking startups for him and others to fund. His firm saw the promise of Big Switch Networks a couple of years ago. It seems the hunger for SDN companies to back has not fallen away. It’s not particularly surprising in this golden age of enterprise IT, in which VCs are hunting for enterprise plays following the rise of SDN and other disruptions.

While the other side of the coin is that production-scale use cases from enterprises are still hard to find, managed-hosting service providers have been quicker to try out the benefits of SDN. After all, those companies can improve their bottom lines by rolling out new services and potentially lowering capital and operational expenditures. But the enterprises have looked more hesitant to dive in and see what’s possible.

That level of market penetration — where large data centers and Wall Street banks give SDN a shot — could happen in 2014 or 2015, said Paul Santinelli, a partner at North Bridge Venture Partners who has backed Embrane.

Rather than present software that can impress network administrators, the key for investors is whether SDN startups can solve real business problems, said Alex Benik, a principal at Battery Ventures. “I think it’s kind of incumbent on the community to work together to not overpromise and underliver on what SDN can deliver on a reasonable time scale,” Benik said. That delivery, he said, is not coming right away, but more like the end of this year and throughout 2014. With SDN code on the way from the OpenDaylight Project, founders that have been working on open controllers and other components in stealth mode could now be pivoting, so VCs might have to wait a while for the most promising SDN solutions.

Related research and analysis from GigaOM Pro:
Subscriber content. Sign up for a free trial.

• Social 2013: The enterprise strikes back
• The promise of SDNs in the enterprise
• Crowdfunding’s rapid growth and future opportunity

More and more IT people are coming to understand and express interest in implementing OpenFlow, which separates the control plane from the data plane and lets servers take charge of telling switches what to do with packets. As the trend takes hold, NoviFlow will surely have to put up with fierce competition, as more vendors move to make their switches OpenFlow-compatible and as OpenFlow-friendly code from the OpenDaylight Project hits the market.

The news comes a year after NoviFlow was founded and just a few months after NoviFlow promised switches that could deliver 100 Gbps. Clearly NoviFlow is serious about capturing marketshare as enterprises consider OpenFlow options and at least think about moving away from legacy vendors such as Cisco.

Cisco, of course, downplays the threat to its bread-and-butter business and is taking steps to protect its market-leading position in switches and routers and enviable profit margins.

“I don’t see this (software-defined networking and network-function virtualization) as a commoditization threat whatsoever to Cisco,” said David Ward, Cisco’s chief technology officer for engineering and chief architect, during a call with investors on Thursday. Startups like NoviFlow are hoping to prove Ward wrong.

This story was updated at 9:03 a.m. PT to clarify the potential speed of NoviFlow switches.

Related research and analysis from GigaOM Pro:
Subscriber content. Sign up for a free trial.

• The promise of SDNs in the enterprise
• Takeaways from the second quarter in cloud and data
• The promise of software-defined networking

Midokura, an SDN startup with overlay-based software for creating load balancers and other virutal devices in Infrastructure-as-a-Service (IaaS) clouds, is closely watching the vendor-led consortium and could join depending on what happens, said Midokura’s chief strategy officer, Ben Cherian. IBM is contributing to OpenDaylight a version of the architecture underlying its virtual overlay product for connecting virtual compute and storage resources, according to the OpenDaylight announcement.

That could pose a problem for Midokura if it’s approved for inclusion in OpenDaylight, particularly if OpenDaylight code is able to do what Midokura’s does but for free. “Are they (OpenDaylight) going to be an eventual competitor? Maybe. But it’s too early to tell until we see how some of these things map out,” Cherian said.

There are similar questions about competition around OpenDaylight’s expected controller code, which Cisco and Big Switch Networks are contributing. The Cyan SDN product runs applications on top of its Blue Planet controller. Rafael Francis, senior director of service provider solutions at Cyan, said he views OpenDaylight as less of a threat and more of a validation of an open, vendor-neutral approach. At the same time, Cyan could get involved with the consortium if customers demand it, he said. Beyond that, OpenDaylight controller code could turn out to work best with Cisco gear, said Mike Bushong, vice president of technical marketing at Plexxi. In that case, developers at multiple participating companies would have to work on maintaining code.

Relations among the many companies involved in OpenDaylight looked copacetic on Monday, with no parties quitting the consortium and with even the Open Networking Foundation feeling good about the OpenDaylight premiere. Dan Pitt, its executive director, said in a statement that the organization is glad to hear that OpenDaylight will support the OpenFlow protocol.

But the relative calm is only based on what’s happened before. Jason Edelman, a solutions architect at New York-based channel partner Presidio, pointed out that only Cisco and Citrix appear to have contributed code, while others seem to be planning to do so. Unity could fracture as companies start unveiling their proposals. For example, what will VMware put forth? The official announcement doesn’t say. When that information comes to light, executives at Cisco and other sponsors might change their minds about participating.

Meanwhile there are questions about the fairness of OpenDaylight’s current organizational structure. Jo Maitland, research director for cloud and infrastructure at GigaOM Research, noted in an email that OpenDaylight “will need an elected board, much like the OpenStack Foundation, to offset the competing interests of the more powerful vendors who will throw lots of money and people to make sure the ‘standard’ evolves to suit their position.”

Although the formation of OpenDaylight seems like a significant development on its own, it ought to seen as the beginning of a story. At this rate, the middle and end will be dramatic to witness.

Related research and analysis from GigaOM Pro:
Subscriber content. Sign up for a free trial.

• The promise of SDNs in the enterprise
• Takeaways from the second quarter in cloud and data
• A near-term outlook for big data