It sounds as if Brocade wants to bill itself as the company that can generate network resources on the fly, only when and where they’re needed. It’s also investing heavily in software, a move that could bring it into competition with giants such as VMware.

Among the new products are a couple intended for network-function virtualization. Brocade’s Vyatta vRouter can virtually set up and configure networks on the fly. It’s available through Amazon Web Services’ marketplace. The Virtual ADX is intended for fast application delivery and control of application management.

Brocade also has decided to make its VCS fabric for connections across hardware available as a plug-in for OpenStack, so users can scale out their networks across multiple clouds.

While software is in the spotlight, Brocade also has a new four-port, 40 GbE card for its MLXe router. The card’s ports support both OpenFlow and traditional routing protocols.

I expect more vendors could follow suit as hardware vendors continue to push their software lineups; hype surrounding software-defined networking and network virtualization continues, and companies wonder what they should try. That means we’ll see a lot of new products hit the market, even as customers try to hire to figure out how to wade through all the FUD.

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

• The promise of SDNs in the enterprise
• The promise of software-defined networking
• Cloud computing infrastructure: 2012 and beyond

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

VMware originally pulled intelligence & dollars from compute layer. Nicira acquisition aims to do this to networking; now Virsto for storage—
Stuart Miniman (@stu) February 11, 2013

In a blog post announcing the aquisition — terms of which were not disclosed — John Gilmartin, VMware’sVP of storage and availability, wrote:

“Our customers have told us that managing performance and data services for virtual machines can be challenging, especially in I/O-intensive environments like virtual desktops. Virsto has developed a VM-centric storage management model that accelerates I/O performance for any block-based storage system while providing efficient data services like VM snapshots and clones. These technologies have helped Virsto customers significantly improve the performance and utilization of their storage systems.”

Virsto handles I/O at the VM level

VMware will keep selling Virsto’s standalone appliance but will also loop Virsto’s data management services into upcoming VMware products, he said.

In another blog post about the deal, Wikibon analyst David Floyer said one rationale for this deal was Microsoft’s decision to include Storage Spaces storage-pooling capability in Windows Server 2012. VMware, he wrote, wants to provide “the same type of simplicity and software-led services that Storage Spaces provides to Hyper-V and Windows 2012. This is a move away from the complex APIs that have previously been provided to help integrate storage arrays with VMware.”

This acquisition is just the latest evidence of VMware’s attempt to push beyond its core server virtualization into software defined data centers where compute is just one piece of the puzzle. That push is also one reason VMware decided to spin off “non-core” technologies like Cloud Foundry, Cetas and vFabric offerings to the Pivotal Initiative late last year.

That VMware went to a third party for storage expertise is interesting since its parent company is storage leader EMC.

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

• How fourth-quarter 2012 will affect IT spending in 2013
• The promise of SDNs in the enterprise
• Cloud computing infrastructure: 2012 and beyond

Seattle-based F5 has been around since 1996 and sells a variety of appliance for everything from load balancing to application firewalls to cloud storage. LineRate is a Boulder, Colo.-based company that launched with $5.4 million in funding in 2012 and wants to help companies scale their network services using software so they can handle more web traffic. Here’s how GigaOM’s Stacey Higginbotham described LineRate’s technology when covering its launch:

“The LineRate software, which was developed at the University of Colorado, works on machines running both network-specific chips and x86-based silicon. This means someone could deploy LineRate’s software to deliver network services on top of Intel boxes or on top of gear using network processors. …

“LineRate’s operating system accelerates the way networking messages are passed through from one server to another, which is why it thinks it can speed up network services even on x86-based chips, which historically have been too slow to handle the real-time world of network services. The software is also massively multitenant, supporting up to 4,000 virtual machines located on one server. This serves cloud clients today, but also presages a future when servers get more cores and capabilities that will allow them to support more virtual machines.”

An F5 Big-IP appliance

Of course, as Stacey also pointed out, SDN startups like LineRate exist in part to disrupt the decades-old hardware-based model of network management that companies such as F5 exemplify. By acquiring LineRate, F5 gets to dip a toe into the future for its webscale customers (one of LineRate’s early customers, for example, is PhotoBucket) while still keeping its appliance on hand for the old guard. Over time, like so many other network vendors, it seems logical that F5 will transition into a company whose hardware is, buy and large, a commoditized delivery medium for its smart network-management software.

Both companies involved are stressing the importance of this acquisition at the application level, where F5 makes its money, as opposed to up the OSI stack where many other SDN products play. In a press release announcing the purchase, LineRate Co-founder and Chief Software Architect Manish Vachharajani is quoted as saying, “We recognize that the SDN fabric may be good at layer 2-4, but ultimately customers care most about the applications.”

Feature image courtesy of Shutterstock user macka.

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
• Cloud and data third-quarter 2012

He said as much on Wednesday at a Churchill Club forum entitled “Is Software-Defined Networking the Next Revolution?” at Ericsson’s office in San Jose, Calif.

It turns out Facebook wants to make networking more efficient, too. Najam Ahmad, director of network engineering at Facebook, said so while sharing the stage with Casado and others at the event. Ahmad is looking for potential solutions to problems such as getting applications to send packets onto servers’ network-interface cards and getting confirmation that that’s happened. No more lost packets that you don’t know are lost.

“We’re just starting out (with software-defined networking),” Ahmad said. “I wouldn’t say by any means that we’re using it wide-scale. We’re looking at use cases and developing that more prototype stage in that sense.”

But use cases are hard to come by, even though AT&T, eBay, Fidelity Investments, NTT and Rackspace have implemented Nicira’s Network Virtualization Platform, as my colleague Stacey Higginbotham reported. But software-defined networking and network virtualization hasn’t exactly gone mainstream. As Casado himself put it, the conversation needs to move toward actual use cases and how to change people’s lives with network virtualization, just as server virtualization has changed people’s lives — or at least IT.

For now, the hype around SDN continues. We’d like to see some network virtualization use cases, too.

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

“We’ve completed nearly two years of development and are moving to the next phase — going to market and supporting customers, we’re fortunate to have Lele,” PLUMgrid CEO Awais Nemat, another Cisco veteran, said in an interview Wednesday. He still isn’t talking — much — about PLUMgrid’s product, which Nemat said is already running at select, undisclosed beta customer sites. Very generally, he said PLUMgrid’s software runs on standard hardware moving packets around and will attack what he sees is a $40 billion market opportunity.

In an interview, Nardin said years spent working both with large customers and with service providers prepare him for this challenge. “I have lots of experience building projects and deploying large networks for such companies as AT&T, NTT and Goldman Sachs.What brought me here is the team.”

That team also includes Valentina Alaria, another Cisco vet who was most recently senior product manager at Nicira. Alaria is head of product management at PLUMgrid.

Companies like PLUMgrid, Nicira, Cisco spin-out Insieme, Embrane and others are all trying to do to networking gear what VMware did to data center servers — treat all the networking gear as a single pool of resources that can be managed and deployed via software.  The idea is getting traction: Last month, VMware bought Nicira for $1.26 billion. A week later Oracle bought Xsigo.

Sunnyvale, Calif.-based PLUMgrid, with 26 employees –20 of which are in engineering — was founded in 2011. Last year, it got $2 million in venture funding from US Venture Partners and reaped the additional $10.7 millon from USVP and Hummer Winblad.  With that new Series A round, Hummer Winblad  managing director Lars Leckie joins USVP’s Chris Rust on PLUMgrid’s board.

The new cash will go to strengthen the engineering team, add QA and customer support people, and fund R&D, Nemat said.

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

• The promise of software-defined networking
• Cloud computing infrastructure: 2012 and beyond
• A near-term outlook for big data

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

• Will Storage Go the Way of The Server?
• How fourth-quarter 2012 will affect IT spending in 2013
• IT spending update, third quarter 2012

Martin Casado, the CTO and co-founder of Nicira

Nicira, the networking startup that is not so stealthy but seriously hot, is ready to tell the world what it offers and who is buying its software. The list of customers is impressive. Nicira’s Network Virtualization Platform is used at eBay, NTT, AT&T, Fidelity Investments and Rackspace, and I expect more companies to announce their use of the networking virtualization solution in the coming months.

Nicira, the brainchild of Martin Casado, who gained fame in networking circles after his Stanford dissertation became the basis for the OpenFlow protocol, is backed with more than $50 million in funding from Andreessen Horowitz, Lightspeed Ventures and New Enterprise Associates. Diane Greene, the co-founder of VMware also contributed. Nicira’s founder, its funders and the technology space where it plays make it one of the hottest startups around.

The problem with virtualization

Nicira is one of several companies attempting to solve the problem that Greene helped create when she co-founded VMware to push hypervisors and virtualization. Once servers were virtualized, it created an easy way to separate computing from the physical infrastructure. The benefits of server virtualization were more-agile compute infrastructures — a developer would spin up a server in minutes as opposed to waiting days for approvals — as well as consolidating IT. Storage followed, but holding the whole virtualized infrastructure effort back was networking. Like a bird with its wings clipped, IT was tethered to the physical hardware by networking.

Server virtualization has made moving applications around on top of servers easy: Think of the transient nature of Amazon’s EC2 instances. But the challenges of maintaining the connection those servers must have to the underlying network has chained the cloud to physical infrastructures — to an IP address — and ensured that employees must manually reconfigure the network to make big changes or implement new networking policies associated with security or compliance. Solutions such as OpenFlow, a protocol that allows the suer to separate the intelligence inside switches from the packet routing and place that intelligence on commodity servers, has helped usher in the age of software-defined networks, which could help solve that problem.

Nicira’s secret sauce

Nicira’s products play into that, but they don’t require OpenFlow to work. Instead of requiring a special OpenFlow-enabled switch, its software works with a variety of protocols and hardware. The company starts by using the Open vSwitch network virtualization software inside the hypervisor. From its vantage point inside a VM thanks to Open vSwitch, Nicira understands data associated with that VM and communicates it to controllers running its software inside the data center.

These controllers are the data center equivalent of a post office box for each VM inside the data center. While the physical address of a virtual machine may change depending on demand, the Nicira controller knows where that VM is and can get messages to it. So, now, messages are sent to the controller, and, like mail sent to a P.O. box, the recipient gets it, no matter where she may physically live.

Agility, cloudbursting and follow-the-sun data centers, oh my!

The infrastructure is worth explaining, but the real excitement is found in how Nicira’s customers are using the product. NTT, which operates data centers around the world, uses Nicira’s software to move its desktop-as-a-service offering from data center to data center within Tokyo ahead of rolling brownouts in the wake of the Fukushima nuclear disaster. That’s right, we are talking about cloudbursting — or moving workloads on the fly from one data center to another. Of course, many caveats still apply.

Both data centers need Nicira controllers, as well as the data the application needs to use in order to run. Casado notes those data centers also need to be connected by pretty fat pipes with low latency to deliver something like NTT’s desktop-as-a-service application. However, customers could use it for cloudbursting if they have already prepopulated a second data center with the information the application needed.

E-commerce giant eBay is using it to deploy applications in minutes as opposed to days, since network engineers no longer have to manually configure networks to meet the needs of a developer. Rackspace is using it to change the products it can offer as a hosting company as well as add agility to its Infrastructure-as-a-Service offerings.

While Nicira’s coming-out party is a big deal for the new age of networking, it is by no means the only player seeking to free cloud computing from the confines of the physical network. Startups such as ConteXtream, Embrane, Big Switch and others are seeking to play a role in the networks of tomorrow. With customers having tested and deployed its platform into production environments, five-year-old Nicira seems further along than some of its fellow startups, but this is a big opportunity and there is still plenty of room.

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

• Cloud computing infrastructure: 2012 and beyond
• The promise of SDNs in the enterprise
• The promise of software-defined networking

In other words, VXLAN could help companies build true global clouds that are the sum of their parts rather than distinct sets of parts. Currently, virtual resources such as storage and compute are limited to the constraints of a single physical network within a single physical data center. Even when tools such as VMware’s vMotion or DRS move workloads or VMs from machine to machine, they can’t escape the data center’s four walls. Herrod describes the problem using a telephony analogy:

One of the fundamental challenges with today’s networking is that we use an IP address for two unrelated purposes, as an identity AND as a location. Tying these together restricts a (virtual) machine from moving around as easily as we would like. We had the same challenge with telephony before wireless came of age. . . our phone number rang for us at a specific destination rather than following us wherever we went!

VXLAN separates a VM’s network ID from its physical location using a Layer 2 abstraction. The result, writes Herrod, is that “VMs are completely unaware of the physical networks constraints and only see the virtual layer 2-adjacency [and] the fundamental properties of virtualization such as mobility and portability are extended across traditional network boundaries.”

Cisco already supports a similar capability with its Overlay Transport Virtualization technology, but the main difference appears to be that OTV is limited to Layer 2 networks (i.e., LANs) and Cisco Nexus gear. VXLAN can cross even Layer 3 boundaries to let users incorporate even cloud-based resources. Cisco’s Omar Sultan wrote on his blog that “VXLAN is the basis of a scalable cloud network where lots of logical networks (over 16M …) can be created instantly to meet the needs of the even the most complex and dynamic cloud.”

Also, VMware is pushing to make VXLAN a standard, and had broad network industry support, including from Cisco, Arista, Broadcom, Brocade, Emulex and Intel.

More information about the technology is available on Herrod’s blog, as well as on the group’s IETF submittal.

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

• The promise of software-defined networking
• What converged infrastructure means for the future of the data center staff
• Infrastructure Q1: IaaS Comes Down to Earth; Big Data Takes Flight

Intel’s decision to buy Fulcrum Microsystems is important and needs to be underscored. Why? Fulcrum makes silicon for Jayshree Ullal’s and Andy Bechtolsheim’s 10 Gigabit, high-performance switch company Arista Networks. The move is a forward-thinking one by the computer chip vendor, as virtualization continues changing the computing landscape. Fulcrum has been around for more than a decade and has raised at least $35 million in venture capital in its 11-year history. From Intel’s release on the acquisition:

“Intel is transforming from a leading server technology company to a comprehensive data center provider that offers computing, storage and networking building blocks,” said Kirk Skaugen, Intel vice president and general manager, Data Center Group. “Fulcrum Microsystems’ switch silicon, already recognized for high performance and low latency, complements Intel’s leading processors and Ethernet controllers, and will deliver our customers new levels of performance and energy efficiency while improving their economics of cloud service delivery.”

Intel’s moves here come amid a shift in the way applications think about architecting their services. Where once a server was an individual computer that ran applications, for cloud computing and large web scale applications such as Facebook, Google or Twitter, they are now components in a much larger system. Google first outlined this shift, but Facebook recently has built upon it with its Open Compute project that basically rethought the way the social network built its hardware and deployed it in a data center and then opened up that design for input and adaptations.

Open Compute left Intel’s (and AMD’s) chip business alone, while putting the margins of systems purveyors such as HP and Dell at risk. However, as other silicon companies such as ARM, Tilera advance, Intel must rethink its value proposition beyond the x86 CPU, much like its customers Dell and HP are now rethinking their value proposition beyond making servers. It’s also looking ahead to the changes that virtualization has wrought in the networking world, where efforts to separate the software that controls where packets go from the switches do the physical routing. This shift is occurring in part because as the data center become a computer, it needs a new way to communicate between nodes (servers). Thus, the rise of fabrics inside next generation servers and inside the data center.

And so back to Intel, which has seen this happening and wants to make sure it has the ability to provide the computing and networking brains inside the new version of the data center. For Intel, the acquisition of Fulcrum enables it to own one more component inside a rapidly commoditizing — but rapidly growing — market. So perhaps we’ll see Intel design a single system on a chip that offers what essentially is a data center on a chip.

Update: Looks like Arista’s Ullal agrees. She emailed GigaOM saying that deal was great for the industry because it validates the concept of merchant silicon as opposed to the specialized chips that Cisco and other networking vendors built. It also moves Intel into the networking sector helping it diversify. Now switch builders can choose from silicon from Intel, Broadcom or Marvell. She added, “This is great for Arista as Fulcrum now has backing of Intel (and truly address the enterprise to cloud networking migration in a mainstream manner). Arista has been one of Fulcrum’s top customers.”

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

• Cloud computing infrastructure: 2012 and beyond
• Infrastructure Q1: IaaS Comes Down to Earth; Big Data Takes Flight
• In Q3, Big Data Meant Big Dollars