Thursday, April 14, 2016

Docker Releases Runtime Compliant with Open Container Initiative

Docker announced its first release built on runC and containerd.

Docker said its Engine 1.11 demonstrates real progress with the Open Container Initiative (OCI), which was launched last year with a mission to build a vendor-neutral, portable and open specification and runtime for container-based solutions. Docker donated its container format and runtime to the OCI under the Linux Foundation in June of last year.

In December 2015, Docker introduced containerd, a daemon to control runC., which is the first compliant implementation of the Open Containers Runtime specification and the default executor bundled with Docker Engine.

Some highlights of Docker Engine 1.11:

  • DNS round robin load balancing between containers with Docker’s networking. 
  • VLAN support (experimental)
  • IPv6 service discovery: Engine’s DNS-based service discovery system can now return AAAA records.
  • Yubikey hardware image signing
  • Labels on networks and volumes
  • Better handling of low disk space with device mapper storage

Cumulus, Dell, Red Hat Collaborate on OpenStack

Cumulus Networks is working with Dell and Red Hat to simplify large-scale OpenStack deployments without the need for proprietary software-defined networking (SDN) fabric solutions. The goal is an all-Linux OpenStack pod that is easy to install and maintain, and incorporates the latest networking technologies.

The companies cited a joint project underway since February to see how advances in automation and networking could be applied to a production-scale OpenStack pod while removing scalability and maintenance challenges for end users. Some project highlights, “all the way down” the OpenStack pod:

  • Scalable 300+ node 100% Linux OpenStack pod
  • Installation and provisioning in six hours using standard open source DevOps tools
  • Simplified IP networking and controller-less VXLAN-based SDN

Using certified Dell hardware for Cumulus Linux and Red Hat OpenStack Platform the pod included more than 300 PowerEdge rack-mount servers installed in nine racks interconnected with 24 Open Networking switches. Dual attached 10Gb Ethernet server connections and 8.6TB cluster bandwidth provided ample capacity for the OpenStack storage and distributed application communications. Ansible by Red Hat, Git, and Red Hat OpenStack Platform formed the deployment tool chain and a virtual environment allowed for prototyping and testing of all components. DevOps tools and principles enabled the geographically distributed team to hit a six hour deployment window to full operation. The pod used an all IP VXLAN networking topology using Cumulus Linux and OpenStack Neutron spanning the leaf/spine switches as well as all OpenStack compute nodes.

“When we launched our Open Networking initiative more than two years ago, our intent was to innovate the network and transform the data center by delivering industry-standard hardware and providing choice of network operating systems and data center applications,” said Tom Burns, Vice President and General Manager, Dell Networking. “This project demonstrated the power of that innovation and how, when combined in a DevOps model, can greatly simplify business challenges for customers.”

“With the increase in production rollouts of OpenStack, customers want the ability to scale without compromising on performance and manageability,” said Radhesh Balakrishnan, General Manager, OpenStack, Red Hat. “In collaboration with Cumulus Networks and Dell we have demonstrated the ability to scale with Red Hat OpenStack Platform and we're looking forward to bringing this powerful solution to our customers. What makes this project even more unique is our focus on delivering a fully open source tool chain, including Ansible, to simplify deployment.”

“It’s powerful to witness the software-defined data center at work,” said JR Rivers, co-founder and CTO of Cumulus Networks. “This project tied together automation, virtual testing environments, tight deployment schedules, and ‘Linux all the way down’ to deliver a production-ready OpenStack pod.”

Digital Ocean Secures $130 Million for Cloud Buildout

DigitalOcean, a start-up based in New York City, secured a $130 million credit facility to continue its mission of building a next generation cloud for scaling startups and software defined businesses.

DigitalOcean offers SSD-enabled cloud servers with simplified pricing tiers.

The company said that over the past two years it has expanded its registered customer base from 253,000 to 708,000 users, who have launched over 13 million cloud servers. The credit facilities will be used to purchase equipment in order to continue its global expansion and support increasing demand.

"We are delighted with the outcome of our credit facility. It complements the $83 million Series B equity financing that we closed in June 2015 and our strong cash flows and balance sheet in pursuing long-term growth opportunities," commented Brian Cohen, Chief Financial Officer of DigitalOcean.

DigitalOcean Raises $83 Million for Cloud Hosting

DigitalOcean, a start-up based in New York City, announced $83 million in Series B funding for its cloud infrastructure for developers of websites and applications.

DigitalOcean offers simple cloud hosting with all servers built on powerful Hex Core machines with dedicated ECC Ram and RAID SSD storage. The company notes that more than 6 million Droplets (cloud servers) have been deployed on DigitalOcean by more than 500,000 developers to date.

The funding round was led by Access Industries, with participation from Andreessen Horowitz.

Telia Carrier Brings Wholesale Transit to Montreal, Launches New Brand

TeliaSonera International Carrier has shortened its name to Telia Carrier and launched new global branding.

Telia Carrier also announced the establishment of a new presence in Montreal, offering wholesale IP transit, Ethernet and IPX services for carriers, content and cloud providers.  Telia Carrier enables Montreal customers to connect to its AS1299 fiber optic backbone, one of the largest in the world, at the highest speeds available in the region.

Telefónica Mexico Orders 1,000 Ericsson Radio Dots

Telefónica Mexico will deploy more than 1,000 Ericsson Radio Dot Systems will be deployed in shopping malls, airports and commercial buildings in Mexico City to reinforce cellular coverage. The units will support 4G/LTE and 3G/WCDMA indoor coverage.

Eduardo Ricotta, Vice President, Ericsson Latin America, says: "Mexico City is the perfect place for Telefónica's largest deployment of Ericsson Radio Dot System. The city is full of high-rise buildings and millions of people wanting excellent mobile broadband services, which Telefónica can now provide in a cost-efficient way."

Ixia Trims Q1 Revenue Outlook

Ixia updated its financial guidance, saying it now expects Q1 revenue to be between $108 million and $111 million, below the company’s previous guidance of $121 million to $126 million. The company currently expects a GAAP loss for the quarter of between $0.03 and $0.05 per diluted share, compared with previous guidance of breakeven to a loss of $0.04 per diluted share. The company currently expects non-GAAP earnings for the quarter to be between $0.05 and $0.08 per diluted share, compared with previous guidance of between $0.10 and $0.14 per diluted share.

"We are disappointed with our first quarter results. Our topline performance was primarily impacted by an unexpected and marked slowdown in network test spending from our North America network equipment manufacturer customers in March, and we continued to experience delays in closing network visibility deals," said Bethany Mayer, Ixia’s president and chief executive officer. "Despite these near-term headwinds, we remain confident in our strategy to grow our business and are committed to financial discipline.”

Emerson Looks to Spin Out Network Power Division by Q3

Emerson confirmed previously announced plans to spin off its Network Power business into an independent business to be called Vertiv.

Emerson intends to spin off 100 percent of the outstanding common stock of Vertiv through a tax-free distribution to shareholders in a process expected to be substantially complete by September 30, 2016.

Vertiv designs, manufacturer and services mission-critical infrastructure technologies for vital applications in data centers, communication networks and commercial/industrial environments. The company offers a broad range of products in both power and thermal management and provides lifecycle management services and solutions for deploying, maintaining and optimizing these products. In addition, Vertiv provides infrastructure management, monitoring, controls and software solutions for its customers’ critical applications.

Scott Barbour, who has been Executive Vice President of Emerson and Business Leader for Emerson Network Power said, “Our years as part of Emerson have given us a tremendous foundation from which to launch and grow Vertiv as an independent company. Our team is looking forward to the opportunity to concentrate on our core competencies, expand our global customer base, strengthen our industry leading positions, and create value for our future shareholders.”

Seagate Misses in Q1, Signs of Enterprise Shift to Cloud

Seagate Technology updated its Q1 guidance, saying it now expects to report revenue of approximately $2.6 billion and non-GAAP gross margin of approximately 23% for the fiscal third quarter 2016. The Company expects to report HDD unit shipments of approximately 39 million, representing approximately 40% market share.

Seagate said its weaker than expected performance was driven "primarily by reduced demand for traditional mission critical HDD enterprise products, reduced demand for the company’s systems and silicon products, reduced demand for desktop client products primarily in China, and the company’s decision to not aggressively participate in the low capacity notebook market." A strong point was better than expected demand for 8TB nearline products, reflecting the company’s belief that an increasing level of enterprise applications are shifting to cloud environments.

“We are disappointed that we did not anticipate the weaker demand in the March quarter. There are many complex issues impacting the traditional go to market channels in our market, which are reducing our forecast visibility. Despite the disruption of the shifts in our traditional mission critical HDD business in the near term, we believe the long term benefit of cloud architectures for end users, and the related need for very high capacity drives, is a net positive for Seagate and the HDD industry," said Steve Luczo, Chairman and Chief Executive Officer.

AT&T Mobility Workers Endorse Contract

AT&T Mobility employees in the company's Southwest Region (CWA District 6) have voted to ratify a four year contract.

The contract covers more than 9,400 Mobility employees represented by the Communications Workers of America in Arkansas, Kansas, Missouri, Oklahoma and Texas.

Wednesday, April 13, 2016

Facebook's Terragraph Accelerates WiGig for City Access

Facebook took the wraps off of two terrestrial connectivity technologies for fast wireless access.

Terragraph is a 60 GHz, multi-node wireless system for dense urban areas and that uses radios based on the WiGig standard. Facebook said Terragraph will deliver gigabits of data capacity. IPv6-only Terragraph nodes will be placed at 200m intervals. Terragraph will incorporate commercial off-the-shelf components and aim for high-volume, low-cost production. Facebook noted that up to 7 GHz of bandwidth is available in the unlicensed 60 GHz band in many countries.  U.S. regulators are considering expanding this to a total of 14 GHz.

Facebook Terragraph will also leverage an SDN-like cloud compute controller and a new modular routing protocol that Facebook optimized for fast route convergence and failure detection. The architecture also tweaks the MAC layer to solve shortcomings of TCP/IP over a wireless link. The company says the TDMA-TDD MAC layers delivers up to 6x improvement in network efficiency while being more predictable than the existing Wi-Fi/WiGig standard.

Terragraph is already in operation at the Facebook campus in Menlo Park, California, where it delivers 1.05 Gbps bidirectional (2.1 Gbps total throughput per distribution node) in P2P mode, up to 250 meters away. A wider trial is planned for San Jose, California.

Facebook also released details on Project ARIES, a transmission technology that is a) spectrally efficient and allows for higher throughput in even the smallest bandwidths, and b) energy efficient, allowing for extended coverage range.

The proof-of-concept system features a base station with 96 antennas and can support 24 streams simultaneously over the same radio spectrum. Researchers have demonstrated 71 bps/Hz of spectral efficiency. The target is aiming for an unprecedented 100+ bps/Hz of spectral efficiency. Facebook said ARIES is an embodiment of Massive MIMO — by using “spatial multiplexing,” the antenna array at the base station can serve a multiplicity of autonomous user terminals on the same time-frequency resource. This opens up possibilities such as resource sharing as an alternative not only to the need for spectrum licensing.

Facebook already has an ARIES testbed delivering 10x spectral and energy efficiency gains in point to multi-point deployments.

MEF Sets Ethernet Interconnect Point Agreement for Carriers

The MEF published the Ethernet Interconnect Point (EIP) ENNI Implementation Agreement – MEF 54, which streamlines the interconnection of operator networks to support end-to-end Carrier Ethernet (CE) services.

The IA provides practical guidance so operators can evolve their networks to meet full CE 2.0 E-Access interconnection capabilities. The IA gives operators the steps involved to create these new interconnections either all at once or in a series of steps. The  IA represents the completion of Phase 1 of the MEF EIP project, which includes the most basic interconnection possible: an Ethernet Private Line created using two E-Access EPL services from different operators with an ENNI interconnect between them.  Phase 2 – which currently is in progress – will add test cases with greater complexity and new use cases focused on EVPL and E-Transit (EPL and EVPL). Additional considerations include color awareness, Class of Service, end-to-end service OAM, and LSO (Lifecycle Service Orchestration).

“The purpose of the EIP project is to produce a series of implementation agreements/ guidelines to create a more efficient marketplace of interconnected operators offering standardized CE 2.0 services that interoperate effectively over two or more networks,” said Nan Chen, President of the MEF. “EIP progress is key to future industry growth and innovation because widely deployed and interconnected CE 2.0 networking fabrics will form the foundation upon which to build agile, assured, and orchestrated Third Network services. “More than a dozen MEF member companies participated in Phase 1 of the EIP project that included Rapid Prototype testing at the University of New Hampshire’s Interoperability Lab,” said Chen. “We want to thank these companies for contributing to move the industry forward: AT&T, Verizon, Windstream, CenturyLink, Frontier, TelePacific, Accedian, Alcatel-Lucent (Nokia), Canoga Perkins, Ciena, Cisco, Juniper, RAD, and Veryx Technologies.”

NTT: Petahertz Frequency Observed in Gallium Nitride

Researchers at NTT and Tokyo University of Science  observed electronic oscillation (dipole oscillation) with attosecond (as: 10-18 of a second) periodicity using gallium nitride (GaN) wide-bandgap semiconductor. The experiment marks the first time the petahertz frequency barrier has been exceeded with a semiconductor. NTT said this shows the potential of future petahertz signal processing technology based on ordinary wide-bandgap semiconductor devices.

Previously, the maximum operational frequency was in the terahertz (1012 Hz) range, which was limited by the response time of band energy modulation with RF electric fields. To observe the ultrafast electron motion with petahertz frequency, extremely high temporal resolution was required.

CWA: Verizon Strike is On

Nearly 40,000 Verizon workers from Maine to Virginia, represented by the Communications Workers of America (CWA) and the International Brotherhood of Electrical Workers (IBEW), went on strike as of Wednesday morning at 6am. The parties have been unable to resolve differences since the last contract expired on August 1st, 2015.

The union cited the following unresolved issues:

  • Offshoring Jobs – Verizon has already contracted out work to more than 5,000 employees in the Philippines, Mexico, the Dominican Republican and other overseas locations. These offshore workers handle customer service calls originating in the Mid-Atlantic and Northeastern states. 
  • Outsourcing Work to Low-Wage Contractors – the union is concerned about the amount of outside line work, particularly vital work installing and maintaining telephone poles, that is being assigned to non-union workers.
  • Contract for Wireless Workers – namedly, a fair first contract for Verizon Wireless retail workers who formed a union in 2014, and for the 100 wireless technicians who maintain the network in downstate New York.
  • Call-Center Closings – hundreds of Verizon workers are at risk of losing their jobs or being forced to commute as much as three hours more each day because of the company's plan to close and consolidate call centers. 
  • Out of State Assignments – long assignments of up to two months at a time

Verizon Comments on Strike

In a press statement, Verizon said it remains committed to its negotiation objectives and seeks to bring negotiation to a closure after ten months at the bargaining table.

The company said 36,000 employees covered under these contracts currently have a wage and benefit package that averages more than $130,000 a year.  Over 99 percent of these employees support the wireline business which in 2015, contributed about 29 percent of Verizon's revenue but less than 7 percent of the company's operating income.

Verizon also said its wireline proposal offers a 6.5 percent wage increase over the term of the contract.

"We have trained thousands of non-union Verizon employees to carry out virtually every job function handled by our represented workforce – from making repairs on poles to responding to inquiries in our call centers.  We know the unions' strike order will be a hardship and pose challenges for our employees, but as a 24x7 customer service company, our contingency plans are in place and our company will continue to serve those who rely on us," stated Bob Mudge, president of Verizon's wireline network operations.

Windstream Connects Data Centers with Infinera

Windstream has deployed the Infinera Cloud Xpress to provide pre-provisioned high speed connections between data centers in Chicago. The initial deployment connects two data center locations within the Lakeside Technology Center at 350 East Cermak.

Infinera said its Cloud Xpress enables Windstream to leverage its unique photonic integrated circuit technology to support up to one terabit per second (Tbps) of input and output capacity in just two rack units. The Cloud Xpress delivers a full 500 Gbps super-channel of wavelength division multiplexing (WDM) line-side bandwidth over 150 kilometers (km) without additional multiplexers or amplifiers, and up to 600 km with an amplified line system. Cloud Xpress also enables efficient scaling with simple provisioning and open interfaces to plug into existing cloud provisioning systems using open software defined networking (SDN) application programming interfaces (APIs). Instant Bandwidth allows the activation of WDM bandwidth in 100 Gbps increments with a few clicks of a mouse enabling rapid scale to meet continued customer demand.

“We selected the Infinera Cloud Xpress to provide data center interconnect services to our customers because it is an extremely simple and highly scalable plug-and-play solution,” said Art Nichols, Vice President of Engineering Architecture & Planning at Windstream. “Windstream is working actively with Infinera and several other key technology partners to employ SDN for creation and management of optical services across its metro and long-haul transport network in 2016.”

Everspin Samples Highest Density MRAM Non-Volatile Memory

Everspin Technologies has begun sampling a 256Mb ST-MRAM Storage Class Memory (SCM) - the highest density commercial MRAM currently on the market. The expects to sample a 1Gb product based on its proprietary perpendicular magnetic tunnel junction (pMTJ) spin torque technology (ST-MRAM) later this year.

Everspin cited the following advantages for it ST-MRAM:

  • Supports 100,000 times faster write speeds than NAND flash
  • Operates on the DDR3/DDR4 DRAM memory interface
  • Requires no wear leveling
  • Provides the highest endurance of currently available non-volatile memories
  • Delivers low power operation
  • Enables instant on/off functionality

“We continue to bring the fastest non-volatile products to our customers, expanding our offering with a high density 1Gb DDR4 pMTJ ST-MRAM. This will provide new and innovative approaches to the way non-volatile memory solutions can be architected,” said Everspin CEO Phill LoPresti.

Tuesday, April 12, 2016

Facebook Delivers 60 Billion Messages per Day

Mark Zuckerberg kicked off Facebook's annual F8 develop conference in San Francisco by saying he is worried by the growing number of people and countries that are turning inward, wishing to build walls instead of bridges, and seeking to limit global communications. Facebook stands for connecting all people in a global community, he said, and it takes courage to choose hope over fear. A big theme of this year's event is bringing people together with new platforms and services.

Some highlights from Day 1 of F8:

  • Zuckerberg presented Facebook's roadmap for the next ten years.
  • Artificial intelligence will power new services with better than human perception.  
  • The next big platforms are Virtual and Augmented Reality.
  • The main Facebook website connects over a billion people.
  • Messenger and WhatsApp together are now handling 60 billion messages a day - about 3X the volume of SMS worldwide
  • 900 million people are using Messenger each month
  • Launching M, an AI-powered chat-bots on Messenger - for businesses, this lets customer interact with sales tools and support systems in an more automated, user friendly format.  Beta launch is underway with APIs being released for the Facebook Bot engine.
  • Launching Facebook Surround 360 - a professional-grade, 17-camera system for 360-degree VR viewing at 8K 60fps.  The system minimizes required post-processing by implementing custom software and codecs. Eight SSDs are used to capture content at 32 GB/s.  With RAID0 there is up to 2 hours of continuous raw capture. Facebook will open source the project on Github this summer.
  • Released Live API - letting apps to stream live video from any device directly to Facebook.
  • Facebook continues to work on drones, satellites, laser communication systems. The Aquilla solar-powered, drone plane will fly at 60,000 feet for months at a time, delivering Internet access to those below.
  • The first Facebook-backed satellite will launch later this year, bringing Internet access over Africa
  • Aries and Terragraph are two new systems that Facebook has developed to improve connectivity on the ground in urban and rural areas.
  • Reducing data consumption is a priority -  FB continues to improve its Lite applications and is introducing an Augmented Traffic Control tool to help developers simulate data conditions in other countries
  • Telecom Infra Project - the idea is to make it cheaper for telcos to operate their network, in the hope that they pass these savings on to consumers in the form of cheaper data plans
  • Free Basics - this program subsidizes the cost of basic mobile data services. So far, it has connected 25 million people in 37 countries

The keynote video is online.

Perspica Emerges from Stealth with Incident Replay Time Machine for Apps

Perspica, a start-up based in San Jose, California, emerged from stealth with a solution that delivers application infrastructure intelligence across hybrid-cloud environments.

Perspica describes its Incident Replay subscription-based service as a time machine for performance data, logs and topology. It provides TechOps and DevOps teams with tools to visualize all layers and siloes of the application stack, enabling definitive post-mortem and real-time automated root cause analysis to improve mean time to repair. The system uses machine learning algorithms to distinguish between critical events and false positive alarms.  The idea is to provide leverage behavior profiling and predictive analytics to detect anomalous trends and prevent dangerous issues before they impact application performance.

"We are excited to bring to market the world's first 'application infrastructure time machine' that combines topology and performance data. It helps overburdened IT teams identify problems and mitigate risks by analyzing the continuous changes to the application infrastructure stack. We are able to track when an incident occurs and automatically identify the root cause to provide IT departments with the ability to gain real answers immediately. Similar to instant replay during the big game, Perspica provides Incident Replay for big data," stated Dan Maloney, CEO, Perspica.

Perspica is funded by March Capital, The Fabric and The Hive.

Datera Launches Elastic Block Storage Software

Datera, a start-up based in Mountain View, California, emerged from stealth and unveiled an AWS-like elastic block storage solution for enterprises and Service Providers.

Datera Elastic Data Fabric is scale-out storage software that turns standard, commodity hardware into a RESTful API-driven, policy-based storage fabric for large-scale clouds. The company said its mission is to bring the operational efficiencies enjoyed by AWS and Google to the open market.  Datera Elastic Data Fabric natively integrates through iSCSI with OpenStack, CloudStack, VMware vSphere and container orchestration platforms such as Docker, Kubernetes and Mesos.

Key features include:

  • API-first operations provides web-scale automation with full infrastructure programmability.
  • Policy based configuration leaves the details to Datera Elastic Data Fabric so customers can optimize price/performance for each application without hand-crafting every LUN.
  • Grow-as-you-go model so customers only buy what they need when they need it. The scale-out architecture of Datera Elastic Data Fabric means customers do not need to compromise on performance or capacity.
  • Flash-first design delivers high efficiency and low latency (< 1ms) across distributed, diverse storage media types. This maximizes application performance and access density.
  • Multi-tenancy and quality of service for cloud-native and traditional workloadsoptimizes asset utilization and simplifies operations across containers, VMs and bare-metal.
  • Heterogeneous component support makes it possible for IT and DevOps to seamlessly scale across many configurations and generations of industry standard x86 servers with multiple storage media types.

Datera also announced $40 million in funding from Khosla Ventures, Samsung Ventures and Silicon Valley luminaries Andy Bechtolsheim and Pradeep Sindhu:

"In the cloud era, things scale up and down constantly. It's always in flux. At scale, you can't operate this kind of environment manually. It must be automated," said Marc Fleischmann, Co-Founder and CEO of Datera. "In this world, customers want their storage fast, at scale, with a self-service delivery model. They can't afford to wait for someone to hand-craft a LUN whenever a developer needs some storage."

Australia's nbn Signs Telstra for HFC Broadband

nbn, the company building Australia’s new broadband network, signed a contract with Telstra to manage the design and construction of fast broadband to more than three million homes currently in the footprint of Telstra’s pay-TV cable network (Hybrid-Fibre Coaxial or HFC cable). Once the build is completed, it will be handed to nbn for operation. Telstra valued the contract at approximately A$1.6 billion.

The deal streamlines the existing arrangements nbn has with Telstra, creating a faster rollout schedule and access to all nbn retailers.

nbn said it is on track to make the initial 10,000 HFC premises available for orders by the end June 2016.

nbn CEO Bill Morrow said:  “The nbn™ network is now available to two million homes and businesses, with close to one million already connected.  This deal will extend the nbn™ network to millions more, bringing new opportunity in education, health and online services to more families.

  • In 2011, nbn and Telstra signed the initial Definitive Agreements for the use of existing Telstra infrastructure and the migration of former wholesale customers from Telstra’s legacy network to the nbn network, with payments progressively scheduled as the new services were activated.
  • In 2014, the nbn and Definitive Agreements were amended to allow nbn to access and progressively acquire existing Telstra network infrastructure to enable the deployment of the multi-technology mix as required.