Thursday, June 5th, 2014

Time	Event
11:30a	Elasticsearch Raises $70M as It Expands Enterprise Search and Analytics Play Real-time search and analytics platform provider Elasticsearch has raised a $70 million Series C funding round, led by New Enterprise Associates with additional participation from Benchmark Capital and Index Ventures. Elasticsearch was born as a search company but now offers a full solution to extract meaningful insights from big data. Its technology, used by the likes of Foursquare, SoundCloud and GitHub, is an open source search and analytics engine, and the company boasts 8 million total downloads for its three core products: Elasticsearch for Search, Logstash for centralized log data and Kibana for real-time analysis of streaming data. Kibana can be used to create time-based comparisons and to generally understand large volumes of data. Combined, the three products form what the company calls the ELK stack. The latest funding round is a big push behind its technology. The money will go towards marketing the solutions, continued product development and expansion of the startup’s sales organization. Harry Weller, general partner at NEA who has joined the Elasticsearch board of directors, NEA had been watching the startup for some time now. “It’s rare to see a company not only get developers excited by their technology but also sell successfully into the enterprise, and we see massive opportunities on an even broader scale. We hope to provide resources and guidance to help Elasticsearch continue down this path, delivering on the promise big data companies have been dancing around for years.” Elasticsearch has made a name for itself powering real-time search functionality on several popular social media and retail applications, but it is expanding quickly into the enterprise space for backend data analytics. The customer base has grown seven-fold, according to the vendor. Its recent big wins include Comcast, eBay, Facebook, Mayo Clinic and Tinder. The company’s talent base has also been expanding. It recently added former talent from Box, Citrix, Splunk and VMware to the team. Founded in 2012, Elasticsearch has raised a total of $104 million.“It’s incredible to receive this kind of support from our investors over such a short period of time,” said Steven Schuurman, the company’s co-founder and CEO. “This speaks to the importance of what we’re doing: businesses are generating more and more data — both user- and machine-generated — and it has become a strategic imperative for them to get value out of these assets, whether they are starting a new data-focused project or trying to leverage their current Hadoop or other big data investments.” Elasticsearch has an impressive track record for its search product, powering customers such as Fourquare (which indexes over 50 million venues), online music distribution platform SoundCloud, StumbleUpon, and  enterprise social network Xing, which has 14 million members. It also powers GitHub, which searches 20 terabytes of data and 1.3 billion files, and Loggly, which uses Elasticsearch as a key value store to index clusters of data for rapid analytics of log files. (Comment on this)
12:00p	Many Agencies Expected to Blow Federal Cloud Security Compliance Deadline Today is the deadline for U.S. federal government agencies and companies that provide cloud services to them to become compliant with a uniform set of security requirements called the Federal Risk and Authorization Management Program, or FedRAMP. The program is one of the biggest attempts by the government to streamline procurement of cloud services by its agencies. It has been about three years since former federal CIO Vivek Kundra rolled out his Cloud First initiative, establishing a policy that required agencies to consider cloud as a way to deploy an application before considering any other deployment options. Cloud First was part of Kundra’s efforts to optimize federal IT infrastructure to make it more efficient. But agencies have been slow to adopt cloud services over these past three years. Responding to a survey by IT services giant Accenture, about 70 percent of agency representatives said their cloud efforts were sluggish because they lacked the necessary staffing, and the rest attributed slow cloud take-up to lengthy procurement processes. FedRAMP was created to address these issues. Agencies can pick from a list of pre-screened providers and services and deploy applications in the cloud without having to figure out whether the services meet their security requirements. So far, a total of 16 service offerings by 11 providers have been certified as FedRAMP-compliant. They include Infrastructure-as-a-Service by the likes of Amazon Web Services, Microsoft, HP and IBM, three Software-as-a-Service offerings and two Platform-as-a-Service products (one by Microsoft and the other by Oracle). Non-compliance unavoidable for many Brian Burns, cloud division director at Agile Defense, a system integrator that helps federal agencies use cloud services, said that while most agencies were running applications within FedRAMP-compliant clouds, many would find themselves non-compliant as today’s deadline passed. Many of them host their own applications and have not gone through the FedRAMP authorization process, while others still have on-going long-term service contracts with non-compliant providers. FedRAMP has a provision that enables agencies using non-compliant hosting companies to seek waivers, but each such request will be reviewed by a special board, and there are no guarantees that the board will approve it. One of the worst consequences of non-compliance for an agency would be a requirement to move applications from a non-compliant service into a compliant one. “You’re going to have to tear that application down, move it to a new cloud and start all over,” Burns said. Amazon leading in government IaaS Providing cloud services to federal agencies is a sizable opportunity for companies. Total U.S. government cloud budget for 2015 is $3 billion, according to slides from a presentation Office of Management and Budget’s Scott Renda made at the FOSE conference in Washington, D.C., in May. That figure, by the way, represents less than five percent of the total government IT budget, according to Renda. About half of the cloud budget will be spent on Software-as-a-Service offerings; about 20 percent will be spent on Platform-as-a-Service, and about 30 percent on Infrastructure-as-a-Service. So far, Amazon Web Services gets most of the government IaaS action, Burns said. The company owes that success primarily to its legacy as a pioneer in providing raw cloud-based infrastructure. “Amazon has more [government] customers than any cloud service provider,” he said. “They’re like the Kleenex brand. Everyone knows the name.” VMware may threaten Amazon’s lead Amazon’s existing FedRAMP-certified IaaS rivals are AT&T, HP, IBM, Lockheed Martin and Microsoft. Burns does not think any of them pose a significant threat to the dominance of AWS. There is, however, a rival in the pipeline that may become a problem for Amazon, and its name is VMware. The Palo Alto, California, superstar of the enterprise IT space, has partnered with Carpathia Hosting in pursuing FedRAMP certification, which Burns expects them to receive this fall. “The big thing that’s going to cause some problems for Amazon is that VMware is a very trusted brand within the federal government,” he said. “Their technology works extremely well.” VMware’s vCloud Government Service promises agencies a completely painless integration of their existing in-house VMware environments with cloud infrastructure in Carpathia data centers. As Burns described it, the service can add VMware cloud as just another node on the agency’s network, extending its existing policies to the cloud environment seamlessly. Amazon simply cannot offer such an easy path to hybrid cloud, since its Citrix-based cloud is not compatible with VMware’s proprietary technology. The VMware threat is also not limited to VMware, the cloud service provider. Other providers, including Lockheed Martin and Verizon Terremark, also use VMware technology for their cloud services, Burns said. Lack of variety no longer an excuse Even with VMware’s government cloud still in the pipeline, agencies that have used lack of choice of providers as a reason for buying non-FedRAMP services no longer have that excuse. With 11 providers certified and offering a variety of solutions, justifying non-compliance will be a heavy burden for them, Burns said. “It really is on the agencies now to make the shift and move into one of those clouds.” (Comment on this)
12:30p	Human Error, Downtime and Design Rob McClary is the Senior Vice President & General Manager at FORTRUST. If one of the leading causes of data center outages is human error, why do we spend so much time focused on the tier classification or rating of a data center?  I would agree that the design plays an important part in reliability, but it’s a small part relative to people, process, operations, maintenance, lifecycle and risk mitigation strategies. Human error puts any data center design at risk for downtime, so why do we continue to over build and wait for the inevitable?  Have we arrived at the point in this relatively young industry, where we believe human error is unavoidable and that focusing on trying to fix the root causes of “human error” is just too hard? This thought process has caused gross over-provisioning and waste of resources, thus sparking a debate among industry thought leaders about what exactly the true value is behind a tier rating. The Tier System The tier classification system has traditionally been viewed as the industry benchmark for design standards and site reliability. A CIO searching for a data center for their organization will refer to the tier classification to predict a data centers expected reliability.  However, tier ratings can often be relied upon too much. Rather than a methodology that emphasizes the predictability of outcomes, it’s the actual results from the data center that matter. I’m not ready to give in to ignoring the “most likely” causes of data center outages in favor of trying to design around human error. I will argue that the management and operations of a data center matters far more than a tier rating. Tier rating or any design for that matter, is no guarantee of reliability. Until recently, data center reliability was always about excess capability and excess capacity, forcing clients to over-provision and to waste important resources in the process. Today’s world favors economy over excess, hence the noticeable shift in the industry’s mentality towards a more progressive approach to data center reliability: a fit for purpose design combined with world class operations and management. Going Beyond Design to Operations Despite the changes, too much importance is still placed on data center design, which is only one small part in creating high-availability. More time needs to be spent on the Uptime Institute’s expansion of their Management and Operations (M&O) Stamp of Approval Program. Why? Once the data center is designed and built, it will inevitably be operated by people. No design I know of addresses the complete removal of human error from data centers. This is a problem you can’t just throw resources at, or design around. This is a problem that can only be solved by creating an organizational structure that mitigates or eliminates human error.  It’s no small effort to do so, however. Fostering ownership, process discipline, procedural compliance, training, and a positive work environment will bestow an operational mindset on your team, which will in turn ensure your data center is reaching its maximum potential—no matter its tier rating. At the end of the day, the bottom-line metric of a data center’s success is simple: the years of continuous uptime that you deliver against the number of unplanned downtime events that you experience. With the focus on operational mindset and organizational strategy, continuous uptime can be achieved over a long period of time.  That’s not luck, that’s not just design, that’s strategy. 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)
12:30p	DCK Analyst View: As SEC’s IBM Cloud Probe Fails, It Is Time To Ask Why Break Cloud Revenue Out at All IBM recently avoided a potential regulatory penalty over the accounting treatment of its cloud services. The U.S. Securities and Exchange Commission’s investigation concluded there was no funny business. IBM disclosed the investigation and did change the way it described cloud revenue in press releases to make it clearer. However, the problem of reporting cloud services revenue isn’t only with IBM, but with all cloud providers. The SEC will have trouble tackling how providers report cloud revenue because all companies report cloud revenue differently, largely thanks to the differences in definitions. Many service providers, such as SoftLayer (which IBM now owns) and Rackspace, have arguably been providing cloud computing before it was broken off as a separate revenue line. Hosting on virtualized servers, before infrastructure cloud services evolved into what they are today, has been around for ages. Today, however, a wide variety of services that do very different things, many of them combining cloud with non-cloud services, get put into different categories, which makes delineating between cloud and non-cloud revenue a difficult task. It’s comparing apples to oranges. Investor pressure adds to confusion Let’s go back to the start of cloud. Once companies began breaking out cloud revenue,  investors were worried that cloud was cannibalizing core hosting businesses. For an analyst, it’s impossible to draw the lines in the sand properly with many companies, as the taxonomy for the hosting services industry is fluid, constantly changing. Rackspace, for instance, has been providing a lot of cloud computing for DevOps, which consists of managed services atop of cloud infrastructure. Do you count this as managed hosting revenue or cloud revenue? The answer is that it’s both; the problem only arises if it’s counted twice. The problem public companies like Rackspace or IBM face is that they are pressured to show growth in cloud investments, often before growth happens. Investors are an impatient bunch. Companies need to show growing cloud numbers without shrinking traditional hosting business lines. If you have massive cloud computing growth but your traditional business is shrinking, you’re potentially  guilty of cannibalizing your arguably higher-margin, more profitable core business. If your traditional business is growing nicely but cloud revenues are behind, you’re guilty of not evolving fast enough. It’s a double-edged sword. Different analyst firms count cloud revenue in different ways, which doesn’t help either. One firm  might only count utility-style compute and storage; another might count any sort of virtualized hosted services, including applications. Gigantic industry-wide projections also add to the pressure to boost cloud numbers and to count more traditional hosting lines as cloud. The futility of breaking out cloud revenue IBM wasn’t penalized because it didn’t double-count any revenue or participate in any sort of chicanery. It was looked at because it is a blue-chip stock that is pushing cloud perhaps the hardest out of the blue chips. But what’s the answer? Breaking out SoftLayer revenue as cloud without anything else? By now, SoftLayer is too deeply integrated into IBM. SoftLayer chief operations officer George Karidis illustrated the silliness in defining what counted as a server during the cloud craze in a 2012 blog post: “It’s easy to forget that the building blocks of a cloud infrastructure will usually come from vendors that provided traditional hosting resources. When a computing instance is abstracted from a hardware device, it opens up huge variations in usage. It’s possible to have dozens of public cloud instances using a single server’s multi-proc, multi-core resources at a given time. If a vendor prices a piece of software on a ‘per server’ basis, how do they define a ‘server’ when their users are in the cloud? It can be argued that a cloud computing instance with a single core of power is a server, and on the flip-side, it’s easy to define a server as the hardware object on which many cloud instances may run. I don’t know that there’s an easy way to answer that question, but what I do know is that applying ‘what used to work’ to ‘what’s happening now’ isn’t the right answer.” The National Institute of Standards and Technology (NIST) attempted to define cloud around 2012. Cloud was on-demand, self-service, broad network access, resource pooling, rapid elasticity and measured service. Straightforward, but again, the SoftLayer blog illustrated the ambiguity: “How quickly does an environment have to be provisioned for it to be considered ‘on-demand?’” wrote SoftLayer’s then CTO,  Duke Skarda. “If ‘broad network access’ could just mean ‘connected to the Internet,’ why include that as a characteristic? When it comes to ‘measured service,’ how granular does the resource monitoring and control need to be for something to be considered cloud computing? A year? A minute? These characteristics cast a broad net.” Convergence means cloud cannot be a separate entity As the industry converges, as cloud, colo, dedicated servers and managed services all become part of mixed deals, how do we count cloud revenue? As hybrid infrastructures take hold, does it make sense to break out what part of that is cloud? No. Service providers are being asked to break out something that can’t necessarily be broken out, unless it’s done with ridiculous granularity (e.g. managed cloud revenue that doesn’t include managed services revenue, cloud in colo/cloud hybrid minus the colo revenue). Then there’s private cloud, which could mean software on on-premise servers and is really maintenance fees. How do you break out “cloud” in a general outsourced service provider business? Cloud is helping to grow outsourcing services in general. Cloud revenue, in the big picture, doesn’t matter. What matters is whether or not a company is generating more revenue overall using cloud as a technology. Cloud is just one part of a larger paradigm that is outsourced services. Amazon Web Services, undeniably a cloud, has so far been immune to cloud revenue reporting scrutiny. Amazon reports cloud revenue under an “Other” category that includes revenue from advertising services and credit card agreements. Amazon isn’t required to break out the cloud business because it’s not 10 percent or more of the total. Is that “Other” category all cloud? No. But as the category grew like crazy, it was clear that cloud was the growth driver. Since Amazon is a retailer first and cloud services provider second, it doesn’t face the conundrum that traditional service providers evolving to cloud face. Analysts and investors will continue to pick at companies and try to determine what revenue is cloud revenue, while the best companies will boost their overall revenue by offering cloud as part of a wider portfolio. (Comment on this)
4:31p	Microsoft Azure Cloud’s Brazil South Region Goes Live Microsoft Azure cloud’s Brazil South region is now generally available. Public preview of the region was available several weeks ago, following roll-outs of new data center locations in China and Japan. Microsoft is after all high-growth emerging markets. The company says its compute capacity is doubling every six months and more than 8,000 customers are signing up weekly. Not only is Brazil a quickly growing market, it is moving forward with net neutrality laws that require data to be hosted within the country. This means service providers targeting the market need a local presence in order to succeed, which has attracted abundance of high-tech investment in the country. Microsoft already has a data center presence locally, having planned to add a data center in Brazil since at least 2009. A recent IDC report says IT investments in Brazil have reached $61.6 billion.  The country ranks seventh in terms of global IT investment, following the U.S., Japan, China, UK, Germany and France. Researchers expect Brazil to replace France next year in sixth place. A Frost & Sullivan report pegged Brazil’s cloud computing market at $328.8 million in 2013. A number of customers are already using the new region, including credit rating company Boa Vista Servicos, business management software provider Totvs, computing systems provider SiplanControl-M and e-commerce website platform Shop Delivery “In Brazil, and our other regions around the world, we’ve seen incredible demand for a global enterprise public cloud offering,” Takeshi Numoto, corporate vice president of cloud and enterprise marketing at Microsoft, wrote in a blog post. “Globally, more than 1,000 new customers join Azure every day, and revenue grew more than 150 percent in the last quarter. With Microsoft Azure now generally available in the new Brazil South region, even more customers can take advantage of the cloud.” Azure ain’t the only one Amazon Web Services first expanded its cloud to South America in 2011 with data center infrastructure in Sao Paolo. It added a second edge location for its cloud in in Rio de Janeiro early this year. Cisco is making a major investment in the country, and recently began producing its UCS servers there. Equinix  first entered the market with an investment in ALOG, providing it with three Brazil-based IBX data centers to offer managed hosting and colocation services for customers looking to expand in the region. Other providers that have expanded into the country include Dimension Data,  Verizon Terremark and cloud and managed services provider Latisys. A popular entry point is via strategic partnership. Latisys partnered with Ascenty, a rapidly emerging metro fiber and data center services provider in Brazil. (Comment on this)
7:32p	Fusion io Optimizes Server Flash from Application-Level Viewpoint Fusion io, which sells flash memory cards for servers, has unveiled its latest product series that doubles data storage capacity of NAND flash and features a slew of enhancements designed to accelerate application performance beyond the basic fast performance of flash as a storage medium. A new generation of NAND flash technology comes out about every 18 months, doubling capacity of the medium. Fusion io’s new Atomic series represents the most current leap. Gary Orenstein, the company’s senior vice president of marketing, said the new platform brought top mainstream NAND flash device capacity from 3.2 terabytes to 6.4 terabytes. Capacity cost does not paint full picture As the cost of flash inches down, it has become an increasingly popular alternative to spinning-disk hard drives in data centers, especially for storage used by applications that require fast delivery of data. Instead of focusing on cost per unit of storage capacity, however, Fusion io promotes other valuable aspects of flash. “We try not to measure it by capacity points,” Orenstein said. “Flash isn’t so much about storing data as it is about delivering data.” Looking at cost of storage capacity alone means to ignore the main value of flash, which is high performance. Many vendors today, Fusion io included, work to increase performance of flash beyond the sheer performance characteristics of the medium itself. They do it through technological innovation in management of the storage systems. One example of such a feature introduced in the Atomic series, for example, is flash-aware compression. While common in traditional storage systems, data compression is not usually employed when flash is involved, because it reduces performance and defeats the purpose of using flash. Flash-aware MySQL compression One of the applications Fusion io has optimized Atomic for is the open source database management system MySQL. When data is erased from NAND flash, it is erased from blocks of cells simultaneously instead of erasing from individual cells. To create those blocks, data needs to be moved around. MySQL also moves data around for its own compression purposes. These processes taking place simultaneously on the same storage area cause the degradation in performance, Orenstein explained. Flash-aware compression exposes the work happening at the NAND device level to MySQL operating at the application level and facilitates coordination between the two. With such intelligence sharing, the impact of compression on performance becomes insignificant, he said. Tight SQL Server 2014 integration Another example is work Fusion io has done to optimize Atomic for Microsoft’s latest SQL Server 2014, launched in April. The application has a feature called buffer pool expansion, which expands the amount of memory available to the database to process things in real time. Atomic integrates with this feature, enabling it to expand the buffer pool to flash memory. Such application-level integration demonstrates Fusion io’s efforts to look beyond working just with the hardware vendors that sell its products as part of their offerings. The company works closely with software vendors, such as Microsoft, VMware, Oracle and SAP, to improve performance of flash when used with the common enterprise applications. (Comment on this)

<< Previous Day

2014/06/05 [Calendar]

Next Day >>