Wednesday, May 7, 2014

Video interview with Cisco: What is OpFlex?

In this video, Tom Edsall, CTO of the Insieme business unit at Cisco, introduces OpFlex, a new policy protocol designed to support physical and virtual switching infrastructure.  OpFlex
provides an abstraction that Cisco believes is better suited than OpenFlow is scaling out policy.

Cisco has published an OpFlex draft in the IETF and will be releasing an OVS implementation in the public domain as well as a controller in the Open Daylight consortium.  The company hopes that open APIs will enable more equipment to be brought in under the OpFlex policy umbrella.  OpFlex fits in with Cisco's Application Centric Infrastructure vision.  Edsall describes differences between the imperative control plane model of "traditional SDN" and the declarative control plane of its ACI model, saying that some things should be centralized while others are best to be distributed.

See 3-minute video:  http://youtu.be/Q4riPf2fgaQ



http://www.cisco.com


In April, Cisco introduced OpFlex - a new networking protocol designed to open up its vision of Application Centric Infrastructure (ACI) in the data center for automated applications and interoperability with other software-defined networking (SDN) elements.


OpFlex is a southbound protocol that is co-authored by Citrix, IBM, Microsoft, and Sungard Availability Services. It provides a mechanism that enables a network controller to transfer abstract policy to a set of “smart” devices capable of directly rendering rich network policy on the device.  OpFlex will enable leading hypervisors, switches and network services (layer 4-layer 7) to self-configure driven by application policy.

Cisco is submitting to the IETF for standardization. It is also an open source Contribution that Cisco is making to OpenDaylight in partnership with IBM, Plexxi and Midokura.  Other companies that are supporting OpFlex include Microsoft, RedHat, F5, Citrix, Canonical, and Embrane.  Hypervisor and software vendors will support OpFlex-enabled virtual switches and extend the Cisco ACI policy framework in their virtual environments. Network services vendors like Avi Networks, Citrix, Embrane, and F5 Networks will be shipping an OpFlex agent with their appliances.

In addition, Cisco is working with OpenDaylight to create a 100 percent open source, ACI-compatible policy model and OpFlex reference architecture.

Compared to the current SDN model, Cisco said its Application Centric Infrastructure avoids the scalability/resiliency challenge of having a single SDN controller managing the state of the network. Its ACI approach is to distribute complexity to the edges and operate disconnected from a central policy manager.  It also would not require application developers to describe their requirements with low level constructs.

Cisco is planning to support the OpFlex Protocol on the following Cisco products:
  • Cisco Application Centric Infrastructure, Nexus 9000 Series
  • Cisco Nexus 1000V
  • Cisco ASR 9000 Series
  • Cisco Nexus 7000 Series
  • Cisco ASA
  • Cisco SourceFire

Web Companies Call on FCC to Defend Open Internet

A coalition of leading Web companies published an open letter to the FCC asking commissioners to defend the principles of the Open Internet.  The letter comes in response to published reports that FCC Commissioner Tom Wheeler is circulating new rules concerning Net Neutrality.  Signatories of the Open Internet letter include:

Amazon
Cogent
Dropbox
Ebay
Etsy
Facebook
Foursquare
Google
Kickstarter
Level 3
LinkedIn
Lyft
Microsoft
Netflix
Reddit
Tumblr
Twitter
Vonage Holdings Corp.
Yahoo! Inc.
Zynga

http://cdn1.vox-cdn.com/assets/4422119/letter_to_FCC.pdf

Wind River Delivers Accelerated vSwitch Optimized for NFV

Wind River announced a new performance benchmark for its accelerated virtual switch (vSwitch) integrated within Wind River Carrier Grade Communications Server, which is designed for network functions virtualization (NFV).

Wind River said its accelerated vSwitch can deliver 12 million packets per second to guest virtual machines (VMs) using only two processor cores on an industry-standard server platform, in a real-world use case involving bidirectional traffic. This performance represents 20 times that of the standard Open vSwitch (OVS) software used in typical enterprise data centers. Providing unlimited scalability when instantiated on multiple cores, this industry-leading performance is achieved using up to 33% fewer CPU resources than other commercial solutions, with no requirement for specific hardware acceleration.

“As a key element in our Carrier Grade Communications Server, our accelerated vSwitch was designed from the ground-up to incorporate the Carrier Grade features that are critically important for telecom networks that must deliver six 9s reliability,” said Mike Langlois, general manager of the communications business for Wind River. "For example, the accelerated vSwitch provides fast convergence during live migration of VMs, while minimizing the impact of dirty page updates. To allow for optimum resource allocation, it provides deterministic processing performance without the jitter of over 10% exhibited by the standard Open vSwitch. Finally, protocols such as LAG, VLAN tagging, and VXLAN provide the security features that are essential for telecom networks.”

http://www.windriver.com


Dell Charts a Cloud-agnostic and Open Approach

Dell reaffirmed its commitment to open, standards-based architectures on which to build public, private and hybrid clouds.

Dell highlighted an enterprise-ready OpenStack-based private cloud solutions that it has co-engineered with Red Hat. As part of the expanded relationship, Dell was the first company to OEM Red Hat Enterprise Linux OpenStack Platform. In addition, Dell announced support for Docker containers in Red Hat Enterprise Linux.

In addition, Dell has partnerships with public cloud infrastructure providers like Google, Microsoft and CenturyLink, among others.

“At Dell, our cloud solutions are based on open architectures with no proprietary lock in. Customers get choice, flexibility and the maximum benefit from their investment,” said Michael Dell, Chairman and CEO of Dell. “Our partnerships with companies like Red Hat further demonstrate that Dell is the only truly open cloud vendor that’s helping customers design, build and manage across public, private and hybrid clouds.”

“Open source is the backbone of the cloud, and the cloud is inherently hybrid,” said Jim Whitehurst, president and CEO, Red Hat. “Our continued collaboration with Dell is about bringing the open hybrid cloud to the enterprise. Industry feedback on our collaboration has been outstanding, and I’m excited to continue working together to bring the value and power of OpenStack – and now OpenShift – to even more enterprises around the world.”

http://www.dell.com

Open Network Install Environment Lab Opens

The University of Texas at San Antonio (UTSA) opened the first ONIE certification lab.

ONIE is an industry standard network boot loader for installing software on network switches.

“With each new platform and chipset, there is a significant amount of development work that is involved to ensure compatibility. ONIE certification and compliance leverages best practices to validate this process in the most expedient way possible,” said Carlos Cardenas, Associate Director, Cloud and Big Data Lab, UTSA. “We are pleased to launch the certification lab as the demand for standardization and reliability across the entire data center ecosystem – from servers to switches and now networking – becomes standard protocol.”

http://www.onie.org/

Alcatel-Lucent Upgrades Transatlantic Apollo Cable to 100G

Alcatel-Lucent completed an upgrade of the Apollo undersea cable system which links the UK and France to the United States for 100G channels and over 25 Tbps of total capacity.

The Apollo system consists of two of the most advanced fiber optic cables: Apollo North connects the UK to the US, and Apollo South connects France directly to the US, collectively spanning a total of 13,000 km.

Alcatel-Lucent deployed its 1620 Light Manager (LM) submarine line terminal equipment using coherent technology at 100G to provide the most efficient use of the available optical spectrum. The 1620 LM leverages Alcatel-Lucent’s advanced coherent technology, which also incorporates the latest Soft Decision Forward Error Correction (SDFEC) technology to provide highest ultimate capacity at the same time as lowest cost per bit.

Richard Elliott, Managing Director of Apollo, said: “Wholesale providers expect higher speeds, faster provisioning and reliability.  The deployment of Alcatel-Lucent’s 100G solution enables us to further meet these expectations on the largest international route in the world.  At Apollo we only provide long haul high capacity services and so getting this right is our entire focus. We are proud to work with Alcatel-Lucent and delighted with the performance of this step change in our technology.”

http://www.alcatel-lucent.com

CALIENT Offers 320x320 Port Optical Switches for Data Centers

CALIENT Technologies unveiled two new optical circuit switch subsystems aimed at data center, telecom or media networking applications.

The new subsystems are based on CALIENT’s 320-port S320 Optical Circuit Switch that uses the company’s patented 3D MEMS optical switching technology. The new subsystems include the MSM-320, a 320x320-port pure optical switch with complete protocol independence and bandwidth scalability to 400 Gbps and up. The low-power module includes the 3D MEMS switching core, mirror driver circuitry and a redundant Linux-based control processor. The control processor manages all system functions including switching and real-time insertion loss optimization. The second OEM subsystem is the OMM-320, a 320-port optical power-monitoring module that uses scanning mirror technology to monitor power levels on up to 320 optical fibers. The OMM provides feedback to the MSM control processor to dynamically adjust mirrors on all 320 connections to minimize optical insertion loss. Alternatively, the OMM-320 can be deployed in other optical networking applications where low-cost, high-port-count power measurements are required.

CALIENT cited several successful deployments of the S320 into cloud datacenter and service provider networks.

http://www.calient.net

Mexico's Alestra Implements 400G DWDM with ALU

Alestra has implemented Alcatel-Lucent 1830 Photonic Service Switch (PSS) platform in its network in Mexico to support channel speeds of 400 Gigabits per second.  The 400G capability has already been tested in Alestra's live network.

“Once again Alestra revolutionizes ICTs optimizing our network to provide the most innovative solutions of Telecommunications and Informatics to the Mexican market. With the deployment of Alestra network leveraging Alcatel-Lucent technology, once again reinforce our offer and our commitment to the industry, making us the first supplier on the continent able to provide businesses with the best value proposition in terms of bandwidth,” said Alejandro Irigoyen, Director of Operations and Systems Alestra.

“The DWDM solution based on the 1830 PSS provides Alestra greater efficiency, flexibility and the scalability required to continue to innovate and lead in corporate services in the Mexican market. With this new network, Alestra will deliver services of up to 400 Gbps between two points on the network, providing greater capabilities in its transport services at the same time that it ensures the successful evolution to cloud services,” said Freddy Turriaf, Commercial Director of Alcatel-Lucent in Mexico.

http://www.alestra.com.mx
http://www.alcatel-lucent.com

See also