Monday, February 27th, 2017

Time	Event
1:00p	How to Get a Data Center Job at Facebook If you’re a mechanical or electrical engineer, know how servers work, or simply someone who enjoys building computers at home and lives near one of the sites where Facebook is building data centers, chances are the company wants to hear from you. The social network is continuing to grow quickly, and every new Facebook data center means lots of Facebook data center jobs. Typically, a few months before a data center comes online, “we put out a huge number of job requisitions, and we start looking to build the team,” Delfina Eberly, VP of infrastructure for site operations at Facebook, says. The roles that need to be filled can be anything from server technicians and infrastructure engineers to program managers. They also range from entry-level to highly skilled. Facebook often hires people out of local junior colleges or trade schools, people who like working with their hands and don’t necessarily have data center experience. “We want somebody who loves technology,” she says. See also: How to Get a Data Center Job at Google Doing More with Fewer People Still, if you’re applying for a server technician position, it helps if you understand how hardware and firmware works. It also helps if you understand automation. Like other hyper-scale data center operators, Facebook has put a lot of effort into reducing the amount of people it takes to service hardware. Servers-per-technician is one of the data center operations team’s key efficiency metrics. Today, they’re at one technician for every 25,000 systems in most locations and in some instances up to 40,000 systems. One of the key processes the company has automated is diagnosing and remediating server issues. Its home-baked FBAR system (stands for Facebook Auto-Remediation) takes a problem through three stages of remediation, and a ticket gets generated for a human tech only if the problem isn’t solved automatically. There’s also automation on the data center infrastructure management side, enabled by Facebook’s DCIM software. “We’re doing a lot of work to automate and detect issues in the environment,” Eberly says. Her team is taking a similar approach to automating mechanical and electrical infrastructure management as it’s done with FBAR on the IT side of the house, where the guiding principle is to have as little human intervention as possible. Hot aisles in a data hall at Facebook’s Altoona data center. (Photo: @2014 Jacob Sharp Photography) Sourcing Locally While local communities usually provide a labor pool to draw from for Facebook data center jobs, not every necessary skill set is easily found in every location. It depends heavily on the presence of other companies and education institutions in the area. The Luleå University of Technology in Luleå, Sweden, for example, is a reliable source of skilled workers for the company’s data center in town. In Forest City, North Carolina, being close to Apple, Disney, and Google data centers made some, but not all skill sets easy to source. See also: Inside the Facebook Data Center in Sweden (VIDEO) “We actually had a very hard time sourcing technicians, people that were more familiar with the internals of servers” in Forest City, Eberly said. To fix the problem, the company partnered with a local community college and created several courses to train server techs. It took people who signed up for the courses on as interns and then placed the interns that worked out onto its hardware-repair team. Be Ready for Constant Change When sorting through resumes submitted for Facebook data center jobs, Eberly pays close attention to introductions. “I always look at the introduction,” she says. “That says a lot about their area of interest, the kind of work that excites them.” She also looks for signals that a person is able to solve complex problems and work in a fast-changing environment. The weight of formal education on a decision about a candidate depends on the specific role they’re applying for. For a server technician, high-school education or previous work experience will usually be enough. Higher-level data center operations roles do require formal technical certification. Panoramic picture of Facebook’s Altoona data center building one. (Photo: @2014 Jacob Sharp Photography) Once a candidate’s resume has been selected, the candidate goes through a process the company calls the interview loop. The process is essentially set up to ensure people from multiple teams and multiple levels in the organization have had a chance to speak with the candidate and can make a decision collectively. As the candidate goes through the loop, a consistent ‘yes,’ ‘no,’ or ‘needs work’ signal typically emerges across the teams they interview with, Eberly says. If you’ve read anything about working at Facebook, you probably know that things move fast at the company. Facebook software engineers are famously encouraged to deploy new features during their first week on the job. The company’s data center team culture is similar. People who end up getting hired for Facebook data center jobs are usually flexible and open minded, comfortable in an environment with quickly shifting priorities. “We really are looking for people who are comfortable with moving fast,” Eberly says, “people who can pivot and make a change quickly.” Read more: What Facebook Learned from Regularly Shutting Down Entire Data Centers (Comment on this)
4:00p	Equinix Heats Up Data Center Alley’s Landgrab Rush Land near the main Equinix data center campus in Loudoun County, Virginia, core of the Northern Virginia data center market, is now more valuable than ever, as evidenced by a recent land deal in the area by the colocation and interconnection giant itself. Equinix has expanded its Ashburn land holdings by paying $34.5 million to acquire four land parcels totaling 34.5 acres, according to county records. This resets the high bar to $1 million per acre for land adjacent to the campus that housed the original MAE-East “Network Access Point.” MAE-East was a carrier-neutral peering point that interconnected long-haul fiber backbones to help create the internet as we today know it. The price is a record high for large, data center-scale parcels in Ashburn, Allen Tucker, managing director at the Tysons Corner office of the real-estate brokerage Jones Lang LaSalle, said in an interview with Data Center Knowledge. Tucker, a Mid-Atlantic region data center expert, said the deal reinforces Northern Virginia’s position “as the premier data center market given its excellent internet connectivity with more fiber than any other global market.” The property is adjacent to the main Equinix campus, at the corner of Waxpool Road and Beaumont Circle, near Loudoun County Parkway. (Source: Loudoun County records) By purchasing land suitable for a future “Equinix 3.0” data center campus, the company is placing a defensive pin in the map of the Northern Virginia data center market. Equinix is widening its moat by preventing a competitor with deep pockets from opening next door. The company is now adding a third property to its existing pair of Boardwalk and Park Place-like campuses, high-end hotels for data. Equinix’s newly purchased plot in relation to its existing campuses (Map by Allen Tucker, JLL) Changing the Competitive Landscape The $1 million-per-acre sale price equates to just under $23 per square foot for the land, or even higher if you consider the cost of carrying the land on the balance sheet for several years. This transaction could also have implications for data center design trends in Loudoun County, which is facing a supply shortage for land parcels big enough for data centers. A typical single-story data center footprint might cover just 35 percent of the land. The high price of land in Loudoun County may begin to accelerate a trend toward two-story or three-story data center designs, which would help new development compete, since higher floor area ratios, or FARs, help reduce cost per square foot. New developments by CyrusOne, Digital Realty Trust, Equinix, and others are using multi-story designs to help pencil out competitive deals while still generating the double-digit returns on invested capital that shareholders have come to expect. This land sale could help make land in neighboring Prince William County more attractive to large enterprise users and multi-tenant data center developers looking to expand in the Northern Virginia data center market. Growing Up in Ashburn There’s little surprise in Equinix’s move besides the price tag. This is one of the most important data center markets — if not the most important —  for the company. Its first building there was the foundation the company was built on. “The company has been adding a new data center in Northern Virginia every 18 to 24 months since it built its first facility there,” Howard Horowitz, global head of real estate at Equinix, told DCK in an interview in 2015. “And I don’t anticipate that changing.” Read more: Equinix Doubles Down in One of Internet’s Most Important Locations The company has traditionally been been expanding in Ashburn in big chunks. It followed the first building, the 40,000-square foot DC1, with the 150,000-square foot DC2. DC3 was a 100,000-square foot data center in a building leased from DuPont Fabros Technology. It appears with the secular tailwinds of cloud computing, SaaS, streaming content, the Internet of Things, and virtual reality, Equinix’s growth in the Northern Virginia data center market will not be changing any time soon. (Comment on this)
4:30p	Ericsson CEO Says Cloud Capability Key to Capturing 5G Market By Niclas Rolander and Caroline Hyde (Bloomberg) — Ericsson AB is betting on cloud technology to capture as much as possible of the $1.2 trillion market it expects fifth-generation mobile broadband services to spur, Chief Executive Officer Borje Ekholm said. “5G is really starting to happen,” Ekholm said in an interview with Bloomberg TV at the Mobile World Congress in Barcelona. “What we’re starting to see is operators positioning their 4G networks and starting to evolve into 5G.” 5G services aren’t expected to be commercially available until 2020. In the meantime, Ericsson has signed development agreements with more than 30 carriers, and some, like Verizon Communications Inc. and its closest rival AT&T Inc., are preparing for trials of the new technology. Verizon said last week it will work with equipment partners including Ericsson and Samsung Electronics Co. to test faster 5G mobile broadband service in 11 markets in the first half of this year. Ericsson, which has suffered from reduced spending by carriers as fourth-generation networks are largely completed in many of its largest markets, said 5G represents an “incredible opportunity” that will be worth a total of $1.2 trillion by 2026. After a troubled 2016, in which Ericsson ousted its CEO Hans Vestberg and blindsided investors with a massive profit warning, sentiment is turning more positive to the former wireless network-equipment market leader. “Our industry, and actually all industries, are experiencing massive disruption,” Ekholm said at a media briefing Monday. “We will need to change, and as we have said, we at Ericsson are working on our strategy.” Since Ekholm started as CEO six weeks ago, Ericsson shares have gained 11 percent, boosted by positive analyst comments, a successful bond sale and speculation about a possible Cisco Systems Inc. takeover. Investors are also looking to Ekholm to save enough to propel the company back to profit growth while still keeping step with investments to capitalize on 5G. The former Investor AB boss said cloud technologies will be crucial for the future of the Swedish company that has suffered a slump in investments by carriers and fierce competition from Chinese rivals Huawei Technologies Co. and ZTE Corp. Ekholm said that so-called operations and business support systems — software used by carriers to manage their networks — represent a key strength for Stockholm-based Ericsson. “We even believe that the network will be in the cloud in the future,” Ekholm said. “That’s why our strength in the virtualization of networks as well as OSS and BSS offerings are central to our future product portfolio.” Fifth-generation networks provide more bandwidth for data and allow easier connections between machines, from traffic lights and cars to thermostats and fridges. Cloud-based services let users access computing power and data on server farms far away, rather than on their own devices such as personal computers or smartphones. (Comment on this)
5:02p	Interxion Buys Amsterdam Data Center from Dutch Service Provider Vancis Interxion is expanding its data center capacity in Amsterdam, one of Europe’s most important network interconnection hubs, where the company is also based. Interxion announced Monday acquisition of the colocation business of the Dutch service provider Vancis Holding BV, including its data center in Amsterdam’s Science Park, a major interconnection point in the market, where Equinix, Interxion’s Silicon Valley-based competitor, has a large presence. With more than 100 carriers and internet service providers, the asset doubles its carrier density in Amsterdam, the company said. The data center is about 20,000 square feet and 75 percent occupied, at least partially by Vancis, which will continue providing cloud and managed services, now apparently as Interxion’s tenant in the facility. Amsterdam is the fourth European data center market where Interxion has announced expansion this month. Just last week, the company said it will build new data centers in London, Frankfurt, and Stockholm. Read more: Interxion to Build £30M London Data Center Despite Brexit Concerns The company is facing increased competitive pressure after Redwood City, California-based Equinix acquired its major competitor TelecityGroup, becoming Europe’s largest data center provider. In a deal that was connected to its Telecity purchase, Equinix also sold some key European assets to Digital Realty Trust, expanding European presence of the San Francisco-based giant with a recently added focus on retail colocation and interconnection – Interxion’s core business. Read more: Why Equinix Data Center Deal is a Huge Win for Digital Realty (Comment on this)
5:30p	Why It’s Time to Make the Leap to the Cloud Peter Sjoberg is Vice President and Chief Technology Officer of Hitachi Data Systems. No technology trend has more potential to be a game-changer for businesses than the cloud. Start-ups like Airbnb that might not otherwise have been launched due to the prohibitive cost of sophisticated IT systems rely on cloud-based pay-as-you-go services to get their starts and disrupt entire industries. If you are an established company with in-house IT systems, you are vulnerable to this kind of disruption. The stakes are high. Yet many companies hold back from moving to the cloud. There are three reasons: fears about security and privacy, IT inertia, and simply not knowing where to begin. Security and Privacy Many companies are just plain nervous about sending their data to an unknown location. According to a TechValidate study, 68 percent of surveyed IT organizations see concern about the security of cloud computing solutions as an obstacle to cloud adoption. However, many experts believe that cloud providers generally have more effective security than most businesses that rely on in-house IT. Since hosting data is their job, cloud providers have good reason to take security very seriously. IT Inertia Many businesses have invested large amounts of money in “old school” IT. This investment in equipment, software, real estate and people creates inertia that tends to perpetuate the traditional data center. Moving to the cloud would appear to mean giving up on a commitment, undergoing a possibly difficult transition and launching into the unknown. Regardless, legacy IT systems must be evaluated objectively and with the future in mind, especially since many cannot stand up to the comparison with far more economical cloud services. Organizations that cling to the past are likely to fall behind competitors that are unafraid to make the cloud transition. Not Knowing Where to Begin Benefits of cloud deployments include high availability, low costs, scalability and business continuity. Even though this knowledge is widespread, many IT leaders are unable to answer the simple question, “How do I get started?” Here are some answers: Begin with one application. It’s not practical to move everything to the cloud at once. Start small. Pick one application, such as email or business intelligence, and move it to the cloud. With a little early success, you will feel more comfortable about moving applications that are more business-critical. Build your next application in the cloud. Go ahead and stick with your legacy IT applications — but look to the future, too. If you are considering deploying a new application, why not start fresh in the cloud? If this new application requires legacy data, you can implement data refining technology to send the data that is needed to the cloud. Push old data to the cloud. Every company stores reams of data for legal or backup purposes. If you’ve been longing to get rid of tape, archiving that data in the cloud is a great way to do it. Go hybrid. Put your current IT systems in a private cloud that delivers services over the Internet, much like a public cloud. Then augment your private cloud with public cloud services as needed. This will enable you to choose between private and public clouds, so your applications are running in the best place for performance and cost. Even if yours is not one of the “cool” companies that was born in the cloud, you can get going now, and your company will reap significant benefits from the cloud that will pay dividends — for years to come. Opinions expressed in the article above do not necessarily reflect the opinions of Data Center Knowledge and Penton. Industry Perspectives is a content channel at Data Center Knowledge highlighting thought leadership in the data center arena. See our guidelines and submission process for information on participating. View previously published Industry Perspectives in our Knowledge Library. (Comment on this)
6:23p	Microsoft Joins Facebook’s Push to Disrupt Telco Infrastructure Microsoft is one of the latest companies to join the Telecom Infra Project, the group of companies Facebook put together last year to reimagine telecommunications infrastructure technology. Both Microsoft and Facebook, as well as other companies with hyper-scale internet platforms with global reach, have been investing heavily in telco infrastructure in recent years. As their bandwidth needs grow, these companies, which also include Google and Amazon, have been playing larger and larger roles in development of new intercontinental submarine cables and expanding their terrestrial fiber capacity. Read more: Facebook and Google to Build Transpacific Submarine Cable Amazon’s Cloud Arm Makes Its First Big Submarine Cable Investment Microsoft, Facebook Build Undersea Cable for Faster Internet Microsoft was one of several new members the organization announced Monday in conjunction with Mobile World Congress, the mobile tech industry’s largest trade show taking place this week in Barcelona. Other new members include Airtel, BT, Dish, Etisalat, and Nextel. Modeled on Facebook’s older open source data center and hardware design community Open Compute Project, the Telecom Infra Project, or TIP, is aiming to accelerate innovation in the telco space because “scaling traditional telecom infrastructure to meet … global data challenge is not moving as fast as people need it to,” Jay Parikh, Facebook’s global head of engineering and infrastructure, wrote in a blog post earlier. Microsoft is one of the biggest OCP members. It was the second hyper-scale data center operator to join after Facebook. Apple and Google are also members but joined years later. Related: Visual Guide to Facebook’s Open Source Data Center Hardware TIP’s focus is on three layers of the telco network: access, backhaul, and core and management. Like OCP, TIP operates around the principle of disaggregation, aiming to replace proprietary, vertically integrated solutions with interoperable components that can be unbundled from the stack and improved individually. One of the first pieces of technology to come out of TIP, the optical transponder Voyager, is being deployed for testing by Telia as part of its longhaul fiber network and by Orange over its optical transport network. Data center provider Equinix, one of TIP’s earliest members, began trials of Voyager in its data centers last year. As part of today’s announcement, BT said it will launch TIP Ecosystem Acceleration Centers in London and Adastral Park, the telco’s research campus in England’s Suffolk County. The first such center is in Seoul, funded by SK Telecom and announced last November. TIP also announced that a group of investors, incubators, universities, and private funds in London will invest $170 million in telecom infrastructure startups. (Comment on this)

<< Previous Day

2017/02/27 [Calendar]

Next Day >>