Wednesday, February 11, 2015

Cyan Packs 800 Gbps in 1RU for Cloud Data Center Interconnect

Cyan unveiled its N-Series Open Hyperscale Transport Platform (OHTP), a new family of high-density, high-capacity transport products designed for the rapidly growing need of cloud and providers to deploy large numbers of 100G connections between data centers over campus or metro regional distances.

Cyan is using commercial off-the-shelf (COTS) silicon and photonic components combined with an open, carrier-grade Linux network operating system in a modular shelf.  The company says this approach will appeal to data center operators who need an easy-to-deploy platform supporting many 100Gs in limited rack space. Unlike traditional optical transport solutions that are vertically integrated, the modular and COTS-based N-Series architecture will enable the platform to fully leverage advances in best-in-breed optical networking technologies and standards as they emerge to
further increase performance and reduce costs. Cyan expects the market for 100G components will gain momentum. The slim form factor enables a "rack and stack" for scaling data center interconnect (DCI) capacity.

The carrier-grade Linux network operating system (NOS), called Cyan Linux, can leverage Linux applications and facilitate rapid integration of third party applications. Cyan believes cloud and content companies will find this familiar to managing their compute resources.

The first member of Cyan's N-Series family is the modular N11, which packs 800 Gbps of total line and client capacity in one rack unit (1RU), or up to 34 Tbps in a seven-foot rack. The N11 is focused on transponding and 100G client interfaces and also supports selectable 200G coherent line interfaces. It can also be paired with Cyan’s Z-Series 96-channel ROADM to drive up to 19.2Tbps on a single fiber and enables Cyan's SDN capabilities.

“Cloud services, video, Big Data and social media are all contributing to a massive surge in data center and cloud traffic,” said Mike Hatfield, president, Cyan. “The N-Series’ COTS-based and modular architecture addresses this need and will significantly reduce the cost of high-capacity optical transport when compared to traditional approaches while eliminating the vendor lock-in and complexity typically associated with proprietary, vertically integrated solutions. The N-Series will further expand Cyan’s addressable market within the cloud and content space and will fully interoperate with our Z-Series packet-optical platforms, which are optimized for metro and regional services and transport.”

Cisco Builds Momentum - Q2 Revenue at $11.9 Billion, up 7%

Cisco reported stronger than expected second quarter revenue of $11.9 billion, net income on a generally accepted accounting principles (GAAP) basis of $2.4 billion or $0.46 per share, and non-GAAP net income of $2.7 billion or $0.53 per share.

"Our Q2 results reflect continued progress as we transform Cisco to become the #1 IT company. In the quarter we grew revenues by 7%, with strong EPS growth, and saw the best balance of growth across all our geographies, products, and segments. We delivered this strong performance despite a volatile economic environment," stated Cisco chairman and CEO John Chambers. "Our strong momentum is the direct result of how well we have managed our company transformation over the last three plus years and our leadership position in the key technology transitions of cloud, mobility, big data, security, collaboration, and the Internet of Everything. Every nation, every company, everything is becoming digitized and the network is at the center of this transformation."

Highlights for quarter:

  • Sales in US grew 7% y/y…compared with 3% y/y in Q1
  • Sales in Latin America returned to double digit growth up 12% y/y, with sales in Mexico up 21% y/y
  • Sales in EMEA were up 7% y/y, with UK up 17% y/y, Germany up 12% y/y and Southern Europe up 20% y/y
  • Sales in India grew 11% y/y 
  • Sales in China dropped 19% y/y
  • Gross margins remain stable in switching.
  • Sales of Nexus 3K and 9K switches grew 350% y/y. The Nexus 9K data center switch passed the 1 million installed ports mark 
  • Cisco now has approximately 1,700 Nexus 9K and ACI customers, up from 970 in Q1.  APIC customers grew to over 300. Chambers said Cisco is pulling ahead of its competitors in SDN because market is "seeing benefits of ACI" especially powerpoints of aspirational competitors. Chambers says APIC and ACI will become cornerstone of next-gen architecture.
  • UCS sales have reached over $3B revenue run rate…over 41,000 customers incl. 85% of Fortune 500 
  • High end routing sales were up 5% y/y
  • Wireless sales grew 18% y/y, with Meraki’s cloud networking business up 100% y/y. Chambers said the Meraki group is on a stellar growth path with a annual run rate of $400 million.
  • Collaboration solution sales grew 10% y/y, with deferred Collaboration revenue up 26% y/y 
  • Cisco's Board of Directors declared a quarterly dividend of $0.21 per common share, a two-cent increase over the previous quarter's dividend.
  • Cash and cash equivalents and investments were $53.0 billion at the end of the second quarter.

Facebook Debuts "6-Pack" Open Hardware Modular Switch

Facebook is taking the wraps off of its "6-Pack" open modular switch platform designed for the flexibility, efficiency, and scale required in its massive data centers. The architecture allows Facebook to build different size switches using common line card and fabric card building blocks.

Last year, Facebook disclosed the specification of its top-of-rack network switch (code-named “Wedge”) and a Linux-based operating system for that switch (code-named “FBOSS”). It then described the modular network architecture it will use for scaling operations. The new 6-pack switch will serve as the core of this fabric.

In an engineering blog posting, Facebook said the 6-pack switch uses Wedge as its basic building block: it is a full mesh non-blocking two-stage switch that includes 12 independent switching elements. Each independent element can switch 1.28Tbps.

Facebook is currently building two configurations: One configuration exposes 16x40GE ports to the front and 640G (16x40GE) to the back, and the other is used for aggregation and exposes all 1.28T to the back. Each element runs its own operating system on the local server and is completely independent, from the switching aspects to the low-level board control and cooling system. Facebook said the advantage of this unique dual backplane design is the ability to modify any part of the system with no system-level impact, software or hardware.

In november 2014, Facebook's Alexey Andreyev outlined a new data center switching architecture deployed at the company's new facility in Altoona, Iowa.

In a blog posting, Andreyev writes that while its previous data centers have been built with a hierarchically oversubscribed system of clusters, Facebook set out to make its newest data center into a single, high-performance network for the whole building.

The new design is characterized by small, identical "server pods" that only require basic mid-size switches to aggregate the Top-of-rack (TOR) switches.  Each pod has 48 server racks, each with a 10G connection. This form factor is always the same for all pods.

Each pod is served by a set of four fabric switches.  Each TOR currently has 4 x 40G uplinks, providing 160G total bandwidth capacity for a rack of 10G-connected servers.  The smaller port density of the fabric switches makes their internal architecture very simple, modular, and robust.  The switches are available from multiple sources. Symmetrical bandwidth is provisioned to/from each pod. To implement building-wide connectivity, Facebook created four independent “planes” of spine switches, each scalable up to 48 independent devices within a plane.  The network is all Layer 3 – from TOR uplinks to the edge.  Standard BGP4 is the only routing protocol. Facebook is using its own centralized BGP controller that is able to override any routing paths on the fabric by pure software decisions.

The full posting is here:

In June 2014, Facebook introduced a new top-of-rack network switch, code-named “Wedge,” and a new Linux-based operating system for that switch, code-named “FBOSS.”  Both products are part of Facebook's efforts to "disaggregate" the network from traditional product categories so as to create more flexible, more scalable and more efficient data centers. 

The Wedge switch features a modular design that brings capabilities of a micro-server using a range of processors, including Intel, AMD and ARM.  On the software side, the "FBOSS" uses the same libraries that Facebook currently uses to manage its server fleet.  This will let Facebook program the switch with the same abstraction layer used for other software services.

In an engineering blog posting, Facebook describes its capabilities as a hybrid of distributed and centralized  control models.

Facebook is currently testing Wedge and FBOSS and plans to contribute key elements to the Open Compute Project (OCP).

Oracle Debuts Network Service Orchestration Solution

Oracle introduced its new Communications Network Service Orchestration Solution to enable communications service providers (CSPs) to more quickly create new network services to respond to market demand.

The solution is part of the Oracle Communications intelligent orchestration framework, which gathers and analyzes network performance information, evaluates the network’s operation against predefined policies and business rules, and can trigger the appropriate action to automatically modify the behavior of services, network functions, and virtualized infrastructure.

Specifically, the Oracle Communications Network Service Orchestration Solution fulfills and extends the role of the NFV Orchestrator as defined by the ETSI NFV Industry Standards Group.  Because it has a semantic understanding of each network service, the Oracle Communications solution can calculate and apply configuration changes, including updates to the network topology or service context data—critical for orchestrating network services through their full lifecycle.

Oracle said its solution open standards including RESTful (web service) application programming interfaces (APIs) for integration with other management and orchestration components, such as the previously released Oracle Communications Application Orchestrator or third-party VNF managers and virtualized infrastructure managers.

“CSPs are taking the first steps toward network virtualization—something they know they must do to compete for the long-term. This is a journey, not just a moment in time, which is why we have outlined an end-to-end intelligent orchestration framework. Our approach will provide CSPs with the business and service agility required to make progress toward full NFV. With our new Oracle Communications Network Service Orchestration solution, we are continuing to deliver on our vision for NFV,” said Barry Hill, global head of NFV, Oracle Communications.

“NFV is complex, and CSPs have made it clear that to be successful, they need comprehensive orchestration. The new Oracle Communications Network Service Orchestration solution is a direct response to this need. It allows for dynamic network operation, enabling the transformation of the business—not just the network. With these developments and an expanding portfolio of virtualized network functions, endorsement of Open Stack, leading OSS solutions, and strong IT heritage, Oracle Communications is enabling service providers to fully embrace the NFV opportunity,” said Liam Maxwell, vice president of products, Oracle Communications.

Facebook Launches ThreatExchange for Cyber Defense

Facebook is launching ThreatExchange - a platform where partners can securely exchange data on malware and phishing attacks.  Early partners include Bitly, Dropbox, Facebook, Pinterest, Tumblr, Twitter, and Yahoo.

The ThreatExchange operates using an API approach that builds on Facebook's internal ThreatData system to create a social platform designed for sharing indicators like bad URLs and domains. Various privacy measures are built into the platform to guard critical information that Exchange participants may not wish to be shared with larger communities.

Telecity and Interxion to Merge their European Data Center Operations

TelecityGroup plc and Interxion, both leading operators of data centers across Europe, have agreed to a merger.

Under the deal, Interxion shareholders would receive 2.3386 new TelecityGroup shares per Interxion share. As a result, Interxion shareholders would own approximately 45%, and TelecityGroup shareholders approximately 55%, of the combined group. The primary listing for the combined group would be in London with a New York Stock Exchange listing for TelecityGroup’s existing ADR programme

TelecityGroup, headquartered in the United Kingdom, operates 39 data centers in key European cities. It has annual turnover of £349 million.

Interxion, which is based in Amsterdam, operates 39 data centres across 11 countries.  It has annual turnover of £274 million.

The companies cited significant synergy potential. Incremental EBITDA from cost synergies and enhanced growth opportunities are estimated by TelecityGroup to be approximately £40m per year and capital expenditure synergies are estimated by TelecityGroup to have a net present value of approximately £300m. In total, this equates to a net present value of total synergies of approximately £600m.

John Hughes would be Chairman of the combined group, with John Baker as Deputy Chairman. David Ruberg would be appointed Chief Executive Officer of the combined group for a period of 12 months following completion of the transaction. He would lead the new, combined group and launch this exciting new phase for both TelecityGroup and Interxion. Eric Hageman would be appointed Chief Financial Officer. The board of the combined group would comprise a balance of independent non-executive directors from both TelecityGroup and Interxion.

Interxion Chairman John Baker said: “I believe that the combination of InterXion and Telecity represents an attractive value creation opportunity for our shareholders, with improved access to capital markets, reduced cost of capital and a strong balance sheet.”

Red Hat Enterprise Virtualization Adds Scalability, OpenStack Integration

A new version of Red Hat Enterprise Virtualization (release 3.5) adds new capabilities in the areas of scalability, management, and integration with OpenStack.

The new version now offer support for four terabytes (4 TB) of memory per host, 4 TB of vRAM, and 160 vCPUs per virtual machine.

Notable new features in Red Hat Enterprise Virtualization 3.5 include:

  • Lifecycle management and provisioning of bare-metal hosts via integration with Red Hat Satellite.
  • Compute resource optimization through advanced real-time analytics with oVirt Optimizer integration. This enables users to identify the balance of resource allocation that best meets their needs while provisioning new virtual machines.
  • Workload performance and scalability provided through non-uniform memory access (NUMA) support, which is extended to Host NUMA, Guest Pinning and Virtual NUMA. This enables customers to deploy highly scalable workloads with improved performance and minimizes resource overload related to physical memory access times.
  • Enhanced disaster recovery via improved storage domain handling, providing support for migrating storage domains between different datacenters supported by Red Hat Enterprise Virtualization, enabling partner technologies to deliver site recovery capabilities.

Red Hat said its new release can serve as the foundation for both traditional virtualization and highly flexible cloud-enabled workloads built on OpenStack. Features for supporting cloud-enabled workloads include:

  • Integration and shared common services with OpenStack Image Service (Glance) and OpenStack Networking (Neutron), available as a Tech Preview, enabling administrators to break down silos and to deploy resources once across the infrastructure.
  • Instance types, unifying the process of provisioning virtual machines for both virtual and cloud-enabled workloads.

ONF to Support Open vSwitch, Appoints Principal System Architect

The Open Networking Foundation (ONF) announced the appointment of Saurav Das as principal system architect, and the establishment of a new project to build upon the OpenFlow Configuration and Management Protocol (OF-CONFIG) to support Open vSwitch (OVS).

As Principal System Architect, Dr. Saurav brings deep SDN and networking software expertise to support new ONF development projects, working with ONF’s CTO and Software Leadership Councils. Saurav’s background in SDN can be traced back to his doctoral research at Stanford University. Under the direction of Professor Nick McKeown (a member of the ONF Board of Directors), he was part of the research group at Stanford that gave birth to SDN as we know it. His work developed a converged IP/MPLS/Optical WAN architecturally founded on SDN and OpenFlow. Post-graduation, he worked as part of the engineering team at Big Switch Networks, investigating SDN platform scalability in large data center networks.

“Open-source resources provide a tangible foundation from which the industry can quickly implement SDN, rapidly increasing the speed at which open SDN can be leveraged by end users,” said Das. “With this understanding, ONF is placing increased emphasis on open source. Stay tuned – much more is coming from the organization in 2015.”

OpenDaylight Project Adds Members

The OpenDaylight Project, which is a collaborative open source project that aims to accelerate adoption of SDN and NFV, announced that CA Technologies, IIX, Megaport and Spirent have joined the project to support the development of one common and open SDN and NFV platform that can be leveraged by all.

“The growth of the community is really quite astonishing ‒ already, over 300 developers globally are working together on OpenDaylight to solve real world problems for end users and deliver the network agility and automation our industry needs,” said Neela Jacques, executive director, OpenDaylight.

Tabula Shutting Down

Tabula, a start-up that was developing high-performance packet processing solutions, is shutting down its operations, according to Silicon Valley Business Journal, which cited a state business filing.

Tabula focused on silicon for 100G systems that leveraged innovations four key areas: a programmable 3D architecture, a RTL compiler, leading-edge process technology using Intel's 22nm Tri-Gate technology, and 3PLD devices.  Tabula was based in Santa Clara, California.

  • In 2011, Tabula announced $108 million in Series D funding for its 3PLD ABAX programmable logic products.