Vello Blog

The Fall of the Optical Networking Oligarchy

by on March 11, 2014

No, the title of this blog is not trying to establish some sort of vaguely disturbing Internet resonance with the current troubles in Crimea. Instead, it is a direct reference to the fact that large, traditional metro optical vendors suddenly find that their complex – and expensive – proprietary data center solutions are now becalmed in an evolutionary cul-de-sac.

Why? Because a forward-looking faction of the previously stunted and slow-moving world of optical networking has gone to the gym and received a complete makeover for the modern world of next-gen data centers and cloud services.

First, six vendors – including Vello – and one innovative service and data center provider, Pacnet – joined forces as Founding Members of the new Open Source Optical Forum (announced at the March 2014 OFC event in San Francisco).

Chartered with promoting the adoption of standards-based, interchangeable, easy-to-use, and power-efficient optical networking technologies into next-generation data centers and cloud environments, OSO is providing open source device-ware code that is fully compatible with the OpenFlow 1.4 protocol from the Open Networking Foundation (ONF).

The ratified OpenFlow 1.4 release contains all the Vello-authored optical extensions previously lacking in OpenFlow. (Vello is a Founding Member of the ONF.) Now optical and Ethernet switches — and other devices, such as optical cross connects – running OSO software can be fully integrated in an OpenFlow environment and easily configured from a single interface. No more requirement for a resident optical high priest and six acolytes to install and maintain optical networking assets.

While the software disruption offered by OSO is a huge game changer in and of itself, there’s also a new hardware sheriff in town – a wholly new class of optical switch that completes the dismantling of the status quo. And these new switches are not a year out. More like 90 days or so from market thanks to a variety of OEMs.

Completely based on off-the-shelf merchant optics, these are the world’s first optical switches in an enterprise-friendly 1RU “pizza box” form factor, and they are coming to market with OSO software.

Think about that: 4x the capacity per network element of existing solutions with half the power consumption and saving 6x to 8x in rack space on top of that — basically three racks of today’s equipment shrink to less than half a rack.

With plug-and-play multi-vendor interoperability, automation and programmability, what’s not to like, particularly as they are half to a quarter of the cost of existing legacy chassis-based switches.

Oh, yes, the architecture of these new DWDM switches is fully extensible from 10G to 100G and all the way up to 1Tb – now that really is investment protection.

All of this comes at the perfect time to help solve the bandwidth crisis inherent in the growth of Cloud services where demand for optical connectivity is off the charts as data centers require increased performance, reliability, and geographic connectivity.

Vello’s innovative Connectivity Exchange software and applications can use white-box Ethernet switches and these new OSO-powered merchant-optical switches to provide optical router bypass, application-based routing and virtual cross connect capabilities.

Taken together, these developments genuinely promise to disrupt the business models of a different class of legacy equipment vendors. Sounds like a good topic for my next post.


Welcoming Big Switch to The-Right-Side-of-History™

by on September 16, 2013


Big Switch has it right. After years of fruitless wandering around the halls of the crowded virtual overlay world, Big Switch Networks has had the inevitable epiphany that the physical underlay is now where it’s at.

Ah, flattery.

We at Vello have always felt a fair bit of camaraderie with Big Switch. They are also charter members of the Open Networking Foundation, and, also like us, are active proponents of the value of OpenFlow. (Any friend of real open standards is a friend of ours.)

Still, while Big Switch followed the sirens’ lure of the overlay hype these past years, along with, well, pretty much everyone else in the ABC (Anybody But Cisco) crowd, Vello adopted the contrarian strategy of targeting the creation of open networking solutions starting at the physical underlay, using an SDN framework and OpenFlow as building blocks. Read More …


IBM and Us

by on September 5, 2013

It’s the ecosystem, stupid!

Vello has just completed successful OpenFlow-interoperability testing between the IBM 8264 64-port Ethernet data center switch and our industry-leading VellOS 7.0 open programmable application infrastructure software. IBM thus becomes the first major branded-hardware supplier to join Vello’s elite approved-vendor program and we welcome them to the fold. Read More …


Equinix and Us

by on July 14, 2013

Recognizing that “SDN” has become one of the most overused terms in all of high tech of late, we have remained focused on customer solutions and quietly won the race to build out a commercially available, end-to-end enterprise-class OpenFlow- and merchant-hardware-based merged optical and Ethernet infrastructure. We are thrilled to announce that this new networking paradigm, based on our new Linux-based VellOS 7.0 network operating system, is now up and running and showcased at Equinix’s NY5 International Business Exchange™ (IBX®) data center in New York. Read More …


Cisco/Insieme – We’re on a Road to Nowhere

by on June 27, 2013

In our last Talking Heads-inspired posting we took a look at the Cumulus announcement of its Linux-based support of the traditional Cisco data center architecture (buzzword-compliancy laws require us to disclose that this Cisco endorsement was all couched using today’s mandatory “software-defined” terminology in the Cumulus press release). Read More …


Cisco Killer, Qu’est-ce que c’est?

by on June 19, 2013

As one of Cisco’s first 100 employees my heart goes out to the good people at Cumulus who appear to have finally come out into the light with their L2/L3 switch software, only to be anointed as the latest in a long and questionable line of “Cisco Killers.” By my reckoning, they are the now the current baton holders in a seemingly endless relay race stretching back to Juniper in 2004 – one that has seen a new runner every 1.25 years since that time. (List available upon request.) Read More …


Fatal error: Uncaught exception 'Exception' with message 'String could not be parsed as XML' in /nfs/c10/h04/mnt/152111/domains/vellosystems.com/html/wp-content/plugins/pardot/includes/pardot-api-class.php:446 Stack trace: #0 /nfs/c10/h04/mnt/152111/domains/vellosystems.com/html/wp-content/plugins/pardot/includes/pardot-api-class.php(446): SimpleXMLElement->__construct('') #1 /nfs/c10/h04/mnt/152111/domains/vellosystems.com/html/wp-content/plugins/pardot/includes/pardot-api-class.php(128): Pardot_API->get_response('login', Array, 'api_key') #2 /nfs/c10/h04/mnt/152111/domains/vellosystems.com/html/wp-content/plugins/pardot/includes/pardot-api-class.php(457): Pardot_API->authenticate(Array) #3 /nfs/c10/h04/mnt/152111/domains/vellosystems.com/html/wp-content/plugins/pardot/includes/pardot-api-class.php(217): Pardot_API->get_response('account', Array, 'account') #4 /nfs/c10/h04/mnt/152111/domains/vellosystems.com/html/wp-content/plugins/pardot/includes/pardot-plugin-class.php(983): Pardot_API->get_account(Array) #5 /nfs/c10/h in /nfs/c10/h04/mnt/152111/domains/vellosystems.com/html/wp-content/plugins/pardot/includes/pardot-api-class.php on line 446