Wednesday, March 5, 2014

ONS 2014 Google Keynote: Software Defined Network Virtualization at Scale

Google is a big believer in cloud networking because it promises an easier operational model and opens the doors to state-of-the-art infrastructure and services, said Amin Vahdat, Distinguished Engineer and Technical Lead for Networking at Google, in a keynote for Open Networking Summit 2014 in Santa Clara, California.

Google's Cloud Platform is enabled by its "Andromeda Network Virtualization" platform -- a software-defined networking (SDN) framework extending across the entire hardware/software stack.  Google takes a holistic approach to all layers in order to achieve QoS, latency parameters and fault tolerance.   Google is making huge investments in its cloud data centers -- at least $2.9 billion in additional planned data centers worldwide in support of cloud services. As reported in the press, managing the energy consumption of these huge facilities is as important as ongoing capital spending.  The company has an extensive CDN footprint to reduce delivery times to end users.

In talks at previous Open Networking Summits, Google first described its global Software-defined WAN, describing the massive backbone as a homegrown project that has enabled the company to treat WAN bandwidth as though it were LAN bandwidth.

Some notes from Amin Vahdat's presentation:

  • Google adopted a share infrastructure model pretty much since its inception.  Hundreds of application run across the same cloud platform.
  • Google is into cluster networking, where storage is disaggregated into pools for virtual machines
  • Google has found that a logically centralized/hierarchical control plane with a peer-to-peer data plane beats out decentralized models.
  • The Andromeda Network Virtualization stack runs on this centralized/hierachical control plane model.
  • Network virtualization bring challenges:  virtual networks must be isolated from each others, DDoS attacks are a threat, network functions virtualization is key to the operation of these networks.
  • Looking ahead a few years, based on Amdahl's Law, Google calculates that future clouds will require extremely high bandwidth and low latencies to interconnect compute resources and shared pools of Flash storage.
  • Virtual machine migration capabilities will be key to support high-availability infrastructure.
  • Vahdat believes that commodity servers running control protocols will be widely used for SDN.
  • Google wishes to eliminate 'one-off' protocols and siloed middle boxes in its network.
  • Google is looking at programmable packet processors to extend network functionality
  • NFV is really about virtualized SDN
  • Andromeda is capable of splitting NFV between soft switches and commodity packet processors.
  • Andromeda capabilities also include cloud load balancing, virtual machine migration, storage services, bandwidth provisioning, and network configuration tools (ACLs, firewall and isolation)
The full keynote has now been posted online by @OpenNetworking Summit.

http://youtu.be/n4gOZrUwWmc



ONS 2014 AT&T Keynote: SDN Innovations Drive Transformation

AT&T is on a mission to transform its networks, operations, business models and even culture by leveraging a new spirit of innovation, said John Donovan, Senior Executive Vice President for AT&T Technology and Network Operations, in a key note address at Open Networking Summit in Santa Clara, California.

A key theme driving this transformation is openness, said Donovan, and AT&T is moving aggressively to implement "the network of the future" - what it calls the User-Defined Network Cloud, a multi-service, multi-tenant platform where equipment is flexible.

AT&T's User-Defined Network Cloud is all about becoming fast, nimble and more customer-focused, said Donovan.  AT&T will take an evergreen approach, preferring to build a new network rather than overlays on the legacy infrastructure.  AT&T will look at new, non-traditional networking vendors, the academic community and the open source movement.

Some notes on Donovan's presentation:

  • AT&T has led the way through several technology transformation inflection points over the course of its long history.  This will be one of them.
  • Over time, all of AT&T's network capabilities will become cloud-centric workloads
  • Customers will have more control to adapt their applications in real time.
  • Applications will pivot from traditional networking to cloud models.
  • AT&T will pursue openness everywhere, including with its Domain 2.0 program, to encourage opinions and innovations.
  • AT&T has half a dozen "beach-head" projects underway in NFV.
  • At times, AT&T will develop its own solutions.  For instance, the control plane of the AT&T backbone is already a software-defined network that manages the 4 million routes of the footprint.
  • AT&T is actively recruiting people with skills in these areas.

One question after the keynote was "why are you really moving to SDN?"  Donovan's answer:  "Because no army can hold back an economic principle whose time has come"

The entire keynote has now been posted on YouTube.

http://www.youtube.com/watch?v=tLshR-BkIas&feature=share&list=UUHo2uqQqpmE_Cg5b4qiUpUg&index=8
http://www.attinvotationspace.com



Equinix Introduces Performance Hub

Equinix introduced its "Performance Hub" concept — a new solution that combines elements of data centers, networking infrastructure and connectivity, and cloud computing access.  The gial is to improve application performance and deliver a globally consistent quality of experience to end users.

Performance Hub deployments include:

  • Equinix IBX data center hosting - Cabinets, power, and cross connects
  • Network connectivity - Metro-Ethernet, long-haul (MPLS-VPN), ISP
  • Network equipment - Routing, switching, firewalls, SSL-VPN, load-balancers, WAN-acceleration
  • Cloud connectivity - Private, high-bandwidth connectivity to leading public clouds such as Amazon Web Services (AWS), Microsoft Windows Azure and many others

Equinix said its Performance Hub delivers significant improvement in performance across metro (coast-to-coast) scenarios including:
50 percent improvement in http-load tests for web-based applications

  • Greater than 40 percent improvement on file transfers for Microsoft Sharepoint®
  • 40 percent improvement on Virtual Desktop Infrastructure (VDI) launch times
  • Reduced jitter and latency for video streaming applications
  • For applications hosted on public cloud infrastructure, 30 percent improvement in page-load tests for web-based applications, if consumed via Equinix Performance Hub with direct connectivity to public cloud infrastructure vs. Internet connectivity; and 100x improvement in data transfers for data-replication applications such as Oracle Data Guard

The Equinix Performance Hub is being launched in Equinix's North American data centers and will be available globally later this year.

http://www.equinix.com


NTT Comm and 25 Partner Carriers Agree to Improve Global Network Service Quality

NTT Communications and 25 of its global telecommunications-carrier partners on a number of policies to further improve its Arcstar (TM) global network service quality during the Arcstar Carrier Forum 2014, which took place in Tokyo on March 5 and 6.

The Arcstar Carrier Forum is an international conference organized by NTT Com to consider strategies for strengthening Arcstar-branded managed network services for multinational corporations.

NTT Com Partner Carriers:
Australia:
Singtel Optus Pty Limited
China:
China Telecom Corporation Limited
France:
Orange Business Services
Hong Kong:
Hutchison Global Communications
Hong Kong:
Wharf T&T Limited
India:
Bharti Airtel Limited
India:
Tata Communications Limited
Indonesia:
PT Indosat tbk
Indonesia:
PT Telekomunikasi Indonesia International
Korea:
KT Corporation
Malaysia:
Telekom Malaysia Berhad
Philippines:
Philippine Long Distance Telephone Company
Singapore:
Singapore Telecommunications Limited
Singapore:
StarHub Ltd.
South Africa:
Internet Solutions
Taiwan:
Chunghwa Telecom Co., Ltd.
Thailand:
CAT Telecom Public Company Limited
UK:
Colt Technology Services Group Limited
UK:
Interoute Communications Ltd.
USA:
AT&T Inc.
USA:
Level 3 Communications, LLC
USA:
XO Communications
USA:
Verizon Communications Inc.
Vietnam:
Vietnam Datacommunication Company
Vietnam:
CMC Telecommunication Infrastructure Corporation
- Message by Guest Speaker who attended Arcstar Carrier Forum 2014

Corvil Develops Alliances for Streaming Data Capture

Corvil, which specializes in real-time operational performance monitoring, has kicked off its Industry Alliance Partner Program to drive advanced solutions and architectures for streaming data, storage and real-time analytics.  Alliance partners include APCON, Arista, Cisco, cPacket, Gigamon, Ixia/Anue and VSS.

Corvil and partners has developed a next generation architecture for time-stamped data capture over the network, lossless storage and real-time IT operational analytics.

http://www.corvil.com

Tuesday, March 4, 2014

Netronome Launches Data Plane Hardware/Software for SDN + NFV

 Netronome introduced its platform architecture to augment virtual switch implementations with hardware acceleration NICs in standard servers for SDN and network functions virtualization (NFV) designs.

The new products include a suite of FlowNIC PCIe Gen3 cards that scale up to 200 Gbps, along with a new FlowEnvironment software package that provides standards-based APIs and configuration protocols for virtual switch offload and acceleration.

The FlowNIC-6xxx family is a high-performance PCIe Gen3 adapter available in 2x40, 4x40, 1x100 and 2x100 gigabit Ethernet options. The cards feature up to four PCIe Gen3 x8 interfaces, delivering unmatched host bandwidth to standard single, dual and quad socket servers. The cards feature 216 programmable cores to keep pace with the rapid change in SDN protocols and standards. Additional hardware accelerators are provided for cryptography, nanosecond accuracy time-stamping, SR-IOV, VMDq, and RoCE. Massive on-chip and on-board memories deliver 24M flow table entries, and 128K wildcard rules to satisfy the requirements of the most demanding SDN and NFV applications.

Netronome said its new FlowEnvironment software delivers more than a 20X increase to virtual switching performance and significantly increases the number of virtual machine instances available per server. The FlowEnvironment includes standards-compliant support for Open vSwitch (OVS) 2.0, OpenFlow 1.4, Intel™ DPDK, and network virtualization protocols such as NV-GRE and VXLAN. The production-ready software provides standard APIs and is fully supported across Netronome’s FlowProcessors that scale up to 400 Gbps. The software is supported on both existing NFE-32xx acceleration cards and the new FlowNIC-6xxx cards.

"Netronome solves the scalability problem of virtual switch implementations in the intelligent network locations where the highest throughput and I/O densities are required, while maintaining the rapid evolution of a software-based edge,” said Niel Viljoen, founder and CEO, of Netronome. “With broad applicability to any virtualized server, the new products are optimized for use in servers running network and security applications, such as SDN middleboxes, SDN gateways and NFV appliances.”

Hyperscale data centers are leading a revolution that is migrating into service provider and enterprise network designs. A complex infrastructure is replaced with a simplified packet core built using merchant switch silicon, surrounded by a new, software-based, intelligent edge. The intelligent edge is comprised of standard servers that are responsible for hosting business applications, and also providing network and security services, while simultaneously directing traffic among virtual machines across a sophisticated overlay network. These network functions are implemented within a virtual switch and consume valuable processing resources. Consequently the complex and varying workloads stifle both application and networking performance.

“NTT Communications has been at the forefront of advancing SDN technologies with commercial use for our cloud computing, datacenter and network services for many years,” said Mr. Yukio Ito, director, member of the board, and senior vice president of service infrastructure at NTT Communications. “We’re evaluating Netronome’s solution of processors, software and NICs and expect to use them to extend our SDN offering into the Cloud-VPN Gateway with close collaboration with NTT Innovation Institute, Inc.”

“FirePower platforms have routinely set industry benchmarks for performance and security effectiveness in data center deployments,” said Tom Ashoff, vice president of engineering, Cisco Systems. “Netronome flow processors and software provide us the feature set and programmability to continue to keep pace with the rapid changes in these hyperscale networks.”

Netronome’s processors, software and FlowNICs are available to OEMs, ODMs and hyperscale network operators. Complete solutions can be purchased direct from Netronome, and through a premier set of partners and suppliers.

“We are pleased that so many customers and partners have chosen Netronome’s processors, software and NICs to help fulfill their vision for SDN and NFV,” said Jarrod Siket, senior vice president of marketing, at Netronome. “This complete solution benefits our customers by enabling their standard servers to reach line rate network performance while returning valuable, and previously wasted, compute resources to the applications and services that need them most.”





Infinera Announces Multi-Layer Automation for Intelligent Transport Networks

Infinera introduced a multi-layer automation solution that integrates the efficiency of its digital switching paradigm with the scale of optical switching to create the communication foundation of cloud-based services and the Internet.

Infinera's multi-layer automation includes the industry’s first super-channel FlexROADM (reconfigurable optical add drop multiplexer), the first standards based multi-layer control plane for spectrum switched optical networks (SSON) and the first 500 Gbps flexible-grid super-channels.

By leveraging these components, Infinera said its Intelligent Transport Network makes it easier for operators to automate the digital switching and the optical transport layers of their multi-terabit transport networks.

Some highlights:

  • Super-Channel FlexROADM for Scalable Optical Switching. Infinera’s converged switching capabilities have been enhanced to feature a new, purpose-built super-channel FlexROADM. The FlexROADM is colorless, directionless and contention-less (CDC) and is available in C, CD, and CDC versions up to nine degrees. When combined with the multi-terabit OTN switching of the DTN-X platform, service providers now have a converged multi-layer solution with digital grooming of client services and optical switching of efficiently filled super-channels. Together these enable service providers to respond quickly to changing service and traffic demands while simplifying network operations.
  • Standards Based Control Plane for Multi-Layer Transport Automation. Infinera's widely deployed standards based control plane now extends transport automation beyond super-channels to control the FlexROADM, which is compliant with the Internet Engineering Task Force’s (IETF) SSON framework.  For the first time, the same standards based multi-layer control plane controls the digital OTN switching for point and click service establishment as well as the switching of flexible grid super-channels at the optical layer. This combination uniquely enables service providers to automate everything in the Intelligent Transport Network. 
  • Flexible Grid Super-Channels for Efficient Long-Haul Transmission. Infinera’s third generation of 500 Gb/s super-channels takes full advantage of large scale photonic integrated circuit technology and the FlexCoherent Processor™ to deliver the industry’s first single-card long haul flexible grid super-channels. Infinera customers have deployed almost one petabit per second of Infinera’s first and second generation FlexCoherent super-channels since the introduction of the DTN-X platform in 2012. 

"At Infinera we work closely with our customers to understand when and where to best introduce technology innovations that drive the most value,” said Dr. David Welch, Infinera Co-Founder and President. “Through this collaboration, Infinera is the first to deliver multi-layer automation of transport networks to help our customers deliver services faster while simultaneously lowering operational costs.”

“At Telefonica I+D we continuously explore innovative technologies to help create the communication services of the future,” said Juan Fernandez-Palacios at Telefonica I+D lab.  “Our successful testing of the innovative FlexROADM helped shape the solution Infinera is announcing today. This is one example of how Telefonica I+D collaborates with technology innovators like Infinera to shape the next generation of optical networks.”

http://www.infinera.com/go/automate





Infinera and Telefónica Demo Network-as-a-Service Powered by SDN

Infinera and Telefónica are demonstrating Network-as-a-Service (NaaS) in a multi-layer environment using a software defined networking (SDN) approach. The collaboration highlighted the power of an open SDN approach and how it can unleash the innovation of a carrier like Telefónica.

As a first step in the collaboration, Telefónica developed its own multi-layer IETF-based Application Based Network Operations (ABNO) controller which included the path computation functions to control both the IP/MPLS layer as well as the Infinera transport layer. Telefónica was then able to rapidly integrate their ABNO controller with Infinera’s programmable SDN Open Transport Switch (OTS) running in conjunction with Infinera’s Intelligent Transport Network platforms leveraging both the OpenFlow protocol and Infinera’s RESTful application programming interface (API).

Telefónica also integrated the ABNO controller with a brand name router to support multi-layer integration. In aggregate, this SDN demonstration allowed end users to define MPLS services from a single screen, resulting in the dynamic allocation and configuration of associated networking resources at both the IP/MPLS layer and the transport layer.  With the speed and automation NaaS delivers, service providers can deploy services faster, more efficiently, and at lower cost.

Specifically this demonstration illustrated:

  • How SDN can allow service providers like Telefónica to unleash their creativity and innovation to develop new multi-layer services rapidly and free from the constraints of proprietary operating systems and traditional element management systems.
  • An SDN control framework based on IETF’s ABNO proposal and its ability to be implemented to support the orchestration and provisioning of services across a multi-vendor, multi-layer IP/MPLS and optical transport network.
  • A working prototype controller developed by Telefónica using multi-layer PCE for networking Layers 1 and 3, demonstrating how SDN technologies and a logically centralized control plane approach can help realize dynamic NaaS capabilities on carrier backbones.
  • Infinera’s Open Transport Switch, a lightweight Web 2.0 software construct that abstracts and virtualizes the underlying transport network resources so they can be easily programmed via a controller. The purpose-built design of OTS for SDN solutions is to enable rapid innovation, in contrast to other transport solutions that repurpose large and heavyweight Element Management Systems (EMS) in their SDN implementations.
  • NaaS and its ability to enable network operators to offer user-initiated connectivity services to multiple customers on top of a common physical infrastructure, empowering the customer to request bandwidth services as needed, without manual intervention. 

"This demonstration with Infinera is the type of research and development that puts us on the forefront of testing new technologies that continually reduce deployment time, the number of manual operations, and deliver more dynamic connections on our network," said Víctor López, responsible for SDN transport innovation activities in Telefónica I+D/GCTO.

"By working closely with the advanced research center at Telefónica I+D we are able to keep our finger on the pulse of network operator requirements for next-generation transport networks,” said Chris Liou, vice president, network strategy at Infinera. “These Network-as-a-Service demonstrations highlight an application for software defined networks that is gaining traction with global carriers."

http://www.infinera.com
http://www.telefonica.com

Vertical Systems: 2013 Global Provider Ethernet LEADERBOARD

Orange Business has regained the top position Vertical Systems Group's 2013 Global Provider Ethernet Leaderboard.. The current rankings (in order based on port share): Orange Business (France), BT Global Services (U.K.), Verizon (U.S.), Colt (U.K.), AT&T (U.S.), NTT (Japan) and Level 3 (U.S.).

Vertical's Global Provider Leaderboard is the industry's benchmark for measuring multinational Ethernet network market presence. Global Providers ranked on the Leaderboard hold four percent (4%) or more of retail business Ethernet installations based on billable ports at sites outside of their respective home countries.

"Orange has regained the number one position from BT based on year-end 2013 Global Provider Ethernet LEADERBOARD port share results," said Rick Malone, principal at Vertical Systems Group. "The share margin between the two is razor thin as both companies push Ethernet into their extensive bases of global IP VPN customers. Orange is now offering Ethernet in 120 countries."

Other Global Providers offering Ethernet services outside of their home countries have port shares that are below the Leaderboard threshold. These companies are segmented by share into two tiers: the Challenge tier and the Market Player tier.

The Challenge tier companies are the following (Note - in alphabetical order): Cogent (U.S.), Reliance Globalcom (India), SingTel (Singapore), Tata Communications (India), Telefonica Worldwide (Spain), T-Systems (Germany) and Vodafone (U.K.).

The Market Player tier encompasses all other providers selling retail Ethernet services outside their home countries. Providers in the Market Player tier include the following companies (Note - in alphabetical order): Bell Canada (Canada), Bezeq (Israel), CAT Telecom (Thailand), CenturyLink (U.S.), China Telecom (China), China Unicom (China), Chunghwa Telecom (Taiwan), Easynet Global Services (U.K.), Eircom (Ireland), Embratel (Brazil), euNetworks (U.K.), Exponential-e (U.K.), Globe (Phillipines), GlobeNet (Brazil), GTS (Poland), GTT (U.S.), Hutchinson Global (Hong Kong), Indosat (Indonesia), Interoute (U.K.), KDDI (Japan), Korea Telecom (Korea), KPN International (Netherlands), Masergy (U.S.), PCCW Global (Hong Kong), PLDT (Phillipines), Rogers (Canada), Rostelecom (Russia), Softbank Telecom (Japan), Sparkle (Italy), StarHub (Singapore), Swisscom (Switzerland), Symphony (Thailand), TDC (Denmark), Telecom Italia (Italy), Telecom New Zealand (New Zealand), Telekom Malaysia (Malaysia), Telenor (Norway), TeliaSonera (Sweden), Telkom Indonesia (Indonesia), Telkom South Africa (South Africa), TelMex (Mexico), Telstra (Australia), Vector (New Zealand), Virgin Media Business (U.K.), XO (U.S.), Zayo Group (U.S.), and other providers.

More information on global and U.S. Ethernet market share results are available from Vertical Systems Group.

http://www.verticalsystems.com/

Source Photonics Debuts 100G LR4 Transceiver for Data Centers

Source Photonics introduced a 100G LR4 optical transceiver in the compact QSFP28 form factor -- an industry first, according to the company, and likely to accelerate the use of 100G links between rack switches and the core of data center networks.

The new transceiver, which is based on the IEEE 100GBASE-LR4 standard, demonstrates a typical power consumption of 3.5W and delivers 100G capability with 40% lower power consumption compared to the industry best CFP4 form factor. The form factor provides 150% increased faceplate bandwidth density compared to 40G QSFP solutions, allowing for up to 48 modules per standard face plate, compared to only 12 with the currently available 100G CFP2 products.

http://sourcephotonics.com

CPLANE Releases its Network Virtualization Platform

CPLANE NETWORKS introduced its network virtualization platform for enterprises and cloud service providers.

CPLANE's software-only Dynamic Virtual Network (DVN), which is built for OpenStack and industry-standard protocols such as Virtual Extensible LAN (VXLAN), aims to transform static, physical networks into virtualized resource pools that can be allocated on demand.

“Our approach has been to unite the ‘divided house’ of the data-center, bringing together cloud application developers and network engineers to deliver agile, automated cloud networks,” said Brandon Williams, CEO of CPLANE NETWORKS. “Enabling application developers and network engineers to work together in a unified DevOps (development/operations) model, DVN and OpenStack significantly simplify and shorten the network services provisioning process.” Zero-touch configuration, reduction of unnecessary network traffic and elimination of external routing services ensure rapid response to service orchestration requests and significantly improve network performance.

http://www.cplanenetworks.com

NTT Comm Expands US Data Center Services

NTT Communications and its U.S. subsidiary, Verio, announced a range of Biz Hosting public cloud services previously available only from NTT Com's East Japan Datacenter.  The new Cloud services and features will support the growth of global businesses, including the expansion of EC-originated websites for U.S. customers, and offer increased reliability due to redundant U.S.-Japan system distribution.

Specific services will include Compute (FLAT Type), simple-, low-cost virtual servers connecting directly to the internet with global IP addresses; functions for increasing/decreasing virtual server resources automatically and in real time; Load Balancing Advanced (LBA) for balancing traffic among multiple virtual servers; Monitoring of virtual server components; and Auto Scaling of virtual servers according to set policies.

Load Balancing Advanced (LBA), Monitoring and Auto Scaling can be combined to maintain optimum resources at all times, ensuring efficient management of infrastructure for websites, content distribution and other services subject to sudden spikes in access.

http://www.ntt.co.jp

Google Compute Engine Boots 1000s VM in Seconds

Google claims its Infrastructure as a Service is now able to boot 1,000 VMs in seconds.  This provides the ability to activate very large workloads on Google's infrastructure with what Google is calling consistent cross-machine bandwidth over its software-defined fiber infrastructure.

Google Compute Engine service now provides native load-balancing to spread incoming network traffic across a pool of instances. Billing is on a minimum of 10 minute increments. Google is providing a number of configuration tools, including a RESTful API and Web console.

Qualcomm's New CEO Takes Over

Steve Mollenkopf has taken over as chief executive officer of Qualcomm and Dr. Paul E. Jacobs is now executive chairman of the Board of Directors.

As CEO, Mollenkopf will assume overall responsibility for Qualcomm, including all lines of business and all functional groups in the Company. In his role as executive chairman, Dr. Jacobs will help guide the development of new technology and Qualcomm’s long-term opportunities.

"I couldn't be more honored and energized to lead the Company through these exciting times," said Steve Mollenkopf, CEO of Qualcomm. "I see numerous opportunities in the coming years for both Qualcomm and the greater industry. Since our founding more than 28 years ago, Qualcomm has been an enabler of the mobile ecosystem and with the support and collaboration of our partners, Qualcomm will continue to push the boundaries of mobile technology."

Mollenkopf has been with Qualcomm for more than 20 years in a variety of leadership positions. Most recently, as the company's President and Chief Operating Officer, he oversaw a number of Qualcomm's investments in technologies that have propelled smartphones into the mainstream and made smartphones the indispensable tools they are today. During Mollenkopf ‘s tenure, Qualcomm has become a leader in a broad range of mobile technologies, including computing, graphics, multimedia chipsets, and 3G and 4G modems.

Prior to his role as President and COO, Mollenkopf led the Company's chipset business, overseeing the launch of 4G technology and making Qualcomm the world's largest mobile chipset supplier and a global leader in LTE technology. Over the years, Mollenkopf has guided the Company through the worldwide expansion of CDMA technology, the introduction of Wideband Code Division Multiple Access (W-CDMA) technology and the launch of 4G/LTE systems. Mollenkopf also spearheaded the Company's $3.1 billion acquisition of Atheros. This was the Company's largest acquisition to date, helping expand Qualcomm's technologies and platforms beyond smartphones into new platforms and segments.

Mollenkopf is a published IEEE author and holds seven patents in areas such as power estimation and measurement, multi-standard transmitters and wireless communication transceiver technology. He serves as chairman of the Global Semiconductor Alliance and a member of the Board of Directors for the Semiconductor Industry Association.

Monday, March 3, 2014

Open Networking Summit 2014 Gets Underway in Silicon Valley

Open Networking Summit 2014 got underway in Santa Clara, California.  This year's ONS features high-profile keynotes from AT&T, Google, Microsoft and Khosla Ventures.

The conference offers sessions on topics such as "Bare Metal Switching & Programming", "SDN for Cellular Networks" and "SDN Deployment Experiences."  It also features over 60 SDN exhibits; a Demo Track with SDN Idol Finalists, and an SDN Solutions Showcase.

Two keynotes are scheduled for Tuesday:

08:30A - 09:30A: "SDN AS A DISRUPTIVE INNOVATION," Vinod Khosla, Khosla Ventures

11:45A - 12:30P: "HOW AT&T PLANS TO TRANSFORM ITS INFRASTRUCTURE WITH SDN", John Donovan, Senior Executive Vice President—AT&T Technology and Network Operations, AT&T Inc.

http://www.opennetsummit.org/

Broadcom Announces OpenFlow Data Plane Abstraction Spec

Broadcom announced its OpenFlow Data Plane Abstraction (OF-DPA) v1.0 specification, software and API -- an openly published implementation of physical switch hardware pipeline abstraction for the Open Networking Foundation's (ONF) OpenFlow 1.3.1 Switch.

The OF-DPA v1.0 specification, software and API can be used to implement network virtualization, multi-tenant networks and traffic engineering with higher scale and performance.  The defines a pipeline that contains multiple tables, each table containing multiple flow entries. The OpenFlow pipeline processing defines how packets interact with these tables. The OF-DPA v1.0 physical switch hardware pipeline abstraction is an implementation of the OpenFlow 1.3.1 Switch optimized for Broadcom StrataXGS Ethernet Switch devices.

Broadcom said its OF-DPA v1.0 software and API enables OpenFlow 1.3.1 agents and controllers to access multiple tables implemented in Broadcom switch devices. The intent is to facilitate general availability of production-quality OpenFlow 1.3.1 switches from OEM and ODM vendors as well as provide a reference platform for use by end users and in academic and industrial research networks.

OF-DPA Version 1.0 Key Features

  • Provides an ONF OpenFlow 1.3.1 compliant switch pipeline and APIs for modifying and querying flow table (e.g., Layer 2 table, Layer 3 table, access control list table) and group table entries, as well as for configuring ports, queues, and VXLAN overlay logical ports. 
  • Includes OF-DPA v1.0 specification, API library, application development kit, and programmer's guide, all released under the Apache 2.0 license.
  • Supports SDN use cases including virtual tenant networks (VTNs), network virtualization using overlays, and traffic engineering.
  • Future OF-DPA versions are slated to support additional Broadcom switch features required in service provider and carrier class applications.  

"The openly published OF-DPA specification, software and API exposes OpenFlow compliant programming constructs over Broadcom's StrataXGS® Ethernet Switch Series," said Ram Velaga, Broadcom Senior Vice President & General Manager, Network Switch. "By mapping the OpenFlow 1.3.1 pipeline to high bandwidth and high density switch silicon like the StrataXGS Trident Series, we are enabling SDN applications to achieve high performance and scale."

http://www.broadcom.com/products/Switching/Software-Defined-Networking-Solutions/OF-DPA-Software

NEC Debuts OpenFlow-based Data center Interconnect Solution

NEC announced the commercial release of Version 5.1 of its ProgrammableFlow Networking Suite, which introduces a new OpenFlow-based, software-defined data center interconnect solution – the Unified Network Coordinator (UNC).

NEC says its new UNC significantly improves the scalability of the network controller by 10 times over previous ProgrammableFlow controllers. This enables construction and orchestration of virtual networks across multiple controllers within a data center as well as across interconnects between data centers. The UNC's capability of linking virtual networks across and between data centers gives customers the ability to link to specific policies within the UNC. In turn, users gain the ability to have an end-to-end policy across multiple controllers and multiple domains.

"We've found that our customers are looking for a policy-based approach to managing their data centers, which is now possible with NEC's UNC," said Don Clark, director of business development, NEC Corporation of America. "The ProgrammableFlow version 5.1 controller allows for better resource utilization and it simplifies data center management, both of which can increase service agility and reduce operating costs. Customers can deploy the UNC to move workloads across multiple data centers and control how traffic flows across WAN links. This allows the pooling of resources, which equates to better utilization of not only networks, but also storage and servers."

As an illustration of its 10x* scalability, each ProgrammableFlow UNC controls up to:

  • 10 controllers or 10 sites or 10 zones in a large data center
  • 2,000 switches (compared to 200 switches per controller in previous version)
  • 30,000 Virtual Tenant Networks (VTN)
  • 100,000 VLANs
  • 10 million flows


http://www.nec.com/en/press/201403/global_20140303_03.html

Infonetics: SDN Creating Hesitancy in Carrier Routing/Switching Market

SDN is leading the carrier router/switch market into a period of hesitation, according to new report from Infonetics.

“Major change is in the air as carriers of all sizes proceed cautiously with router and switch spending, in part because they are trialing SDN (software-defined networking) or just beginning to figure out how to proceed with SDN,” says Michael Howard, principal analyst for carrier networks and co-founder of Infonetics Research. “The fourth quarter is usually an indicator for the current year and sometimes a bellwether for the future. In the most recent quarter (4Q13), North America had a double-digit sequential decline in carrier router/switch revenue rather than the usual budget flush, with weakness coming mostly from Verizon and AT&T, among the global leaders of SDN activities.”

Some highlights of the report:

  • In 2013, worldwide service provider router and switch revenue—including IP edge and core routers and carrier Ethernet switches (CES)—was $14.5 billion, up just 2% from 2012
  • 4Q13 carrier router/switch revenue is down 4% from the same period a year ago
  • Prompted by the move to 100GE, the core router refresh is still mildly in motion: Core routers are up 3.3% for the full year 2013, but flat quarter-over-quarter (4Q13 from 3Q13)
  • There is intensifying focus on multiple content delivery networks (CDNs) and smart traffic management across routes, aimed at making routers and optical gear cooperate more closely
  • The top 4 manufacturers – Cisco, Huawei, Alcatel-Lucent, and Juniper stayed in dominant positions in 2013, together taking 83% of global carrier router and switch revenue
  • Over the past 2 years, Huawei has increased its market share the most of any vendor
  • With SDN hesitation slowing router and CES spending, Infonetics now expects worldwide service provider router and switch revenue to grow at a low single-digit CAGR from 2013 to 2018, when it will reach $17.2 billion

http://www.infonetics.com

See also