Thursday, November 21st, 2013

Time	Event
12:49p	Big Data News: HP Expands Vertica Analytics Platform HP expands its big data analytics platform with Hadoop integration, SGI partners with Extreme Networks for Hadoop solutions, Zettaset secures Hadoop, and Concurrent delivers an all-access pass to business intelligence tools on Hadoop. HP expands Vertica Analytics platform. HP (HPQ) announced enhancements to the HP Vertica Analytics Platform, enabling organizations to accelerate business value from a vastly expanded variety of data. The new version 7, also called Vertica Crane, dramatically simplifies the exploration and analysis of semistructured and “dark” data, provides enhanced integration with Hadoop, and offers significant security and performance enhancements. The HP Vertica Analytics Platform also supports a variety of industry-standard business intelligence and visualization tools. It delivers open “SQL-on-Hadoop” capabilities. Unlike other SQL-on-Hadoop solutions, HP Vertica works with major Hadoop distributions, ensuring high-performance analytics across the broadest range of data types and sources. It also supports direct integration with HCatalog, Hadoop’s table and storage management layer. “Everyone is seeking to gain insight from new forms of data, but legacy products waste resources because of the need to wrestle the data into structured formats,” said Colin Mahony, vice president and general manager, Vertica, HP. “The new capabilities of the HP Vertica Analytics Platform deliver immediate value and rapidly ‘shine a light’ on formerly dark data—unlocking the value hidden in both traditional and emerging forms of big data.” SGI and Extreme Networks partner. Extreme Networks (EXTR) announced an agreement with SGI, to be its select networking partner in order to provide comprehensive Hadoop solutions globally. SGI will include Extreme Networks’ Summit X670 fixed Ethernet switches, with 10/40GbE capabilities, as part of its SGI Clusters for Hadoop. The integrated solution is designed to help customers harness valuable information and rapidly improves the speed at which data can be analyzed.  ”As businesses continue to tackle Big Data, they face the challenge of how to unlock value at greater speed, scale and efficiency,” said Bill Mannel, general manager of Compute for SGI. “We’re applying the expertise we learned in designing and building the world’s fastest high-performance supercomputers to our Big Data Analytics solutions, allowing customers to increase ROI on their Big Data investments as they scale.” Zettaset Encrypts Hadoop.  Zettaset announced that it has added data-at-rest encryption capabilities to Zettaset Orchestrator, its distribution-agnostic Hadoop cluster management platform. Organizations with strict corporate and/or industry compliance requirements, such as those in the financial, retail, and healthcare verticals can now confidently deploy Hadoop in environments where personal information data must be tightly safeguarded. Orchestrator uses the 256-bit Advanced Encryption Standard (AES-256) to ensure the highest levels of encryption and data security. “Encryption is a very specialized capability, and there are few viable options available today for Hadoop users,” said Jim Vogt, Zettaset CEO. “When it comes to risk management, Zettaset Orchestrator with data-at-rest encryption gives customers the upper hand, supporting compliance mandates such as HIPAA, BSA/AML and PCI-DSS, for example, and provides assurance that their Hadoop cluster data is protected against malicious attacks.” Concurrent BI Tools on Hadoop.  Enterprise big data platform company Concurrent announced the release of Cascading Lingual, an open source project that provides ANSI compatible SQL enabling fast and simple big data application development on Apache Hadoop.  With Cascading Lingual, enterprises that have invested millions of dollars in business intelligence (BI) tools and training can now access their data on Hadoop in a matter of hours, rather than weeks. By leveraging the power and broad platform support of the Cascading application framework, Cascading Lingual lowers the barrier to enterprise application development on Hadoop. “Concurrent is committed to our mission of simplifying enterprise application development and deployment on Hadoop,” said Gary Nakamura, Chief Executive Officer at Concurrent. ”Cascading Lingual abstracts the complexity of Hadoop into languages and interfaces that people already know, providing a pragmatic, yet powerful, tool for enterprises that are looking to make the most of their in-house talent. With Cascading Lingual, companies can now make the most of Hadoop without the costly investment in additional BI tools and training.” (Comment on this)
1:25p	Why The Hybrid Cloud is Gaining in Popularity What’s the driving force behind the hybrid cloud boom? Now that we all have our heads firmly in the cloud, it’s important to understand where this platform is going. The evolution of cloud computing has taken the data center and all of the required supporting technologies to a new level. Whether organizations are creating advanced disaster recovery solutions or are bringing their cloud to the edge with fog computing – more businesses are finding direct use-cases behind this model. When it comes to a high-level understanding of the cloud platform, there are generally 4 models to know: Private. This is where an organization elects to host their equipment and present a cloud model which is managed and controlled by the internal IT team. Resources, physical and virtual, all belong to the organization. Public. Here, and organization may choose to work with an outside cloud provider for their hosting needs. Applications, servers and workloads are all provisioned from resources owned by the third party provider. This is a good example of a pay-as-you-grow model. Community or cooperative clouds. Instead of just provisioning space in a public cloud, organizations can test and work on a cloud platform which is secure, “dedicated” and even compliant with certain regulations. A great example would be the need for a provider to host a specific application on a set of cloud-based servers. Instead of giving each organization their own server in the cloud for this app, the hosting company allows multiple customers connect into their environment and logically segment their sessions. Hybrid. These cloud models are being adopted by numerous organizations looking to leverage the direct benefits of both a private and public cloud environment. In a hybrid cloud, companies can still leverage third party cloud providers in either a full or partial manner. This increases the flexibility of computing. The hybrid cloud environment is also capable of providing on-demand, externally-provisioned scalability. Augmenting a traditional private cloud with the resources of a public cloud can be used to manage any unexpected surges in workload. This is where workflow automation can really help out. If an organization has peak usage times, they are able to offload their user base to cloud-based computers which are provisioned only on demand. This means that these resources are only being used as needed. So, for organizations still looking to keep a portion of their cloud environment private, but still use elements of the public cloud offering – moving to a hybrid cloud may be the right solution. Many organizations now see the direct benefits of working with some type of cloud model. As the dust settles and you create your cloud platform, you may quickly realize that almost all platforms have a piece of their cloud in the hybrid space. You’ll notice that you’re cross-connecting with more resources, more users and a lot more devices all over the world. In many cases, the workloads, data and applications your accessing isn’t even located on a corporate-owned data center. It’s all connected through an ever-expanding hybrid cloud infrastructure. Unless you’re completely cut off from the Internet, your organization is going to have some type of cloud presence. If you’re presenting applications to users via that cloud, you’re already adopting some type of cloud model. The reality now is that many organizations are utilizing resources both inside and outside of their data center. We’ve always been walking a fine cloud line when it came to defining the exact cloud model that we were using. Now, with so much cross-connection and outside integration, almost all cloud-based organizations look to resources located outside of their infrastructure. This can be a SaaS application, some type of file sharing solution, or even new initiatives around mobility. So what’s the driving force behind the hybrid cloud boom? What is causing the blurring of the cloud line? Software-defined technologies. A big reason for the hybrid cloud evolution has been the software-defined layer. SDN has helped bridge a lot of the cloud computing communication that has to happen on such a large layer. By better integrating complex routing and switching methodologies at the logical layer, software-defined networking allows administrators to create vast hybrid cloud networks capable of advanced inter-connectivity. Greater amounts of resources. Cloud computing is where it is today mainly because of the resources that support it. We have more bandwidth, new ways to connect, and greater amounts of infrastructure convergence. As the integration of storage, networking and computing capabilities have increased – so have the delivery methods of cloud computing. Hybrid cloud models are now able to cross-connect with distributed data centers and utilize vast amounts of bandwidth which has now been optimized by virtual WANOP appliances. All of this creates the platform and bridge for the hybrid cloud model to scale and grow. Logical and physical integration. There has been a logical and physical revolution. New APIs are forcing IT professionals to rethink how applications and resources integrate together. We are able to eliminate entire layers to allow for greater application and data communication. Take the Seagate Kinetic Open Storage platform, for example. This platform eliminates the storage server tier of traditional data center architectures by enabling applications to speak directly to the storage device, thereby reducing expenses associated with the acquisition, deployment, and support of hyperscale storage and cloud infrastructures. The way that we consume information and utilize resources has changed. The days of the PC, as we know it, are numbered and we are ushering in an era of mobility. Application-centric workloads are the rulers of the cloud world as users utilize apps and data in harmony. It’s no longer about the desktop and all about data-on-demand. All of this will continue to push cloud and its supporting technologies forward. Ultimately, there will be one general hybrid cloud model which will support a number of varying systems. For the business, the infrastructure will be simplified and easier to manage. For the user, they will have direct access into their resources, applications and data. For both, there will be direct operational and productivity benefits by utilizing the power of a hybrid cloud model. (Comment on this)
1:31p	Challenges in Using Outside Air Economizers and How to Overcome Them Robert F. Sty, PE, SCPM, LEED AP is Principal of SmithGroupJJR’s Technologies Studio in Phoenix, AZ and focuses on the Architecture and Engineering design of mission-critical facilities. Robert is on LinkedIn. ROBERT STY SmithGroupJRR’s Techologies Studio With the American Society of Heating, Refrigeration, and Air Conditioning Engineers (ASHRAE) expanding the window of operating temperature and humidity ranges inside the data center, many designers and operators clearly see the benefits in the use of outside air economizer strategies. Also known as “free cooling,” the economizer operation allows the facility to shut down the chillers or DX compressors, and cool the data center by introducing outside air at the proper conditions. Although many times the need for mechanical cooling is not completely eliminated from the data center, there are significant energy and cost savings associated with economizer strategies. The concept of outside air economizer seems simple. Turn the chillers off, open the outside air dampers to let the fresh air in to cool the cabinets, and let the air back out through some means of relief. Unfortunately, if not designed properly bringing in outside air can actually put the data center operation at risk. Listed below are four common issues with the design of economizers and the strategies used to minimize the risk to operation. Outside Elements Inside the Data Center It is obvious that with bringing in outside air into the data center to cool the racks a number of other particulates such as dust and particulate matter will also enter the facility. Dust can enter the data center and clog server fans, and under the right conditions can also become wet and therefore conductive (although this is a rare occurrence). ASHRAE published an update in 2011 to its original 2009 white paper “Gaseous and Particulate Contamination Guidelines for Data Centers.” This white paper goes on to recommend that room air, which is continuously recirculated be filtered with MERV 8 and that any outside air that is introduced into the data center be filtered with MERV 11 or 13 filters. Dust will always be present in the data center, but proper filtration of outside air can minimize a number of operational issues. Proper Mixing of Outside and Return Air Depending on the local climate, there could potentially be many times throughout the year when the outside temperature is below the ASHRAE recommended low temperature limit of 65°F. These times of the year the HVAC system would operate under what’s known as a “partial economizer.” During this mode a portion of the warm return air typically in the 90°-95°F range is mixed with the outside air to reach the desired supply temperature (typically in the 75°F range).  The balance of hot air is exhausted or relieved to the outside.  If this is done inside a central station air handling unit mixing plenum, it’s almost a guarantee that the return and outside air will mix properly. In cases where it is done outside the air handling unit, such as a large ducted mixing plenum or room, care must be taken that warm return air properly mix with cold outside air to ensure that cold/hot spots do not enter the air handling units.  Cold spots could potentially cause coil freeze and burst in cold weather climates, and warm spots will impact IT cabinet performance. Temperature and Humidity Changes Throughout the day the outside air temperature and humidity can fluctuate based on incoming weather patterns or during the “shoulder seasons” of spring and fall.  Control systems should include weather stations that will enable the BMS to react to the changing weather patterns. In certain climates such as Salt Lake City, UT, morning temperatures could fall within acceptable ranges, but by early afternoon can be much hotter than the allowable recommendations of ASHRAE TC 9.9.  As the data center changes modes from free cooling to chilled water cooling, there must be enough time allotted for the central plant to drop the water temperature to the appropriate set point before the outside air dampers are completely closed. Given the large volume of chilled water in some piping systems, it is beneficial to understand the time it takes to reach this desired set point so that a proper reaction time is established in the system control strategy.  In addition, the ASHRAE recommendations give a maximum rate of change for temperature and humidity inside the data center.  HVAC sequence of operations must be written to take these limits into consideration as well. Building Pressurization Introducing outside air to the data center requires exhaust from the return air plenum or the building can become over pressurized.  Some level of positive differential pressure in the data center is not necessarily a bad thing as it helps with moving the air into IT cabinets while using strategies such as hot aisle containment. The HVAC designer should take caution in just how much positive pressure is set as to avoid issues with opening/closing doors.  The means of economizer relief is an important consideration in the system design.  It is tempting to allow air from the hot aisle or return plenum to gravity relief without the use of relief fans. The elimination of fans reduces electrical consumption and lowers the PUE.  With gravity relief designs instances have occurred where return air plenums have been over pressurized to open a gravity relief damper to relieve air from the building.  This can create issues with “ballooning” of membrane roofing systems or “oil canning” (popping) of sheet metal roof decking.  Both of which are undesirable results.  Relief fans with variable speed drives set to maintain proper pressure differentials between hot and cold aisles are effective, provide finer pressure control, and do not add significant energy demands to the facility. Address Issues in the Design Phase Correctly designed outside air economizer systems can result in significant energy and economic savings for data centers. Without proper attention to the above items, the introduction of outside air can pose additional risks to data center operation and uptime. However, if one takes into consideration these issues, a properly designed HVAC system will yield positive results. 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)
2:00p	Online Tech Adds 34,000 Square Feet with Metro Detroit Data Center A view of Online Tech campus in the Detroit, MI region. This article originally appeared on TheWHIR. Web hosting provider Online Tech announced on Monday that it is renovating and building out a 34,000 square foot data center in metro Detroit. The company has made a $10M investment in infrastructure and will add 15 new jobs in the data center over the next five years to operate the facility. The network switching facility was formerly occupied by Sprint-Nextel, and will feature 18,000 square feet of raised floor space. Online Tech currently operates two data centers in Ann Arbor, and Flint, Mich., and the new facility will bring its total data center footprint to 100,000 square feet. It is the market leader in the Detroit Metro with the top market share in multi-tenant data center space in Michigan, according to 451 Research. Recently, Online Tech expanded its Ann Arbor headquarters to meet its rapid growth after doubling its employee count to nearly 50 over the past 18 months. “Our data centers deliver secure colocation and cloud hosting services to clients whose IT operations must comply with regulations like HIPAA, PCI and Sarbanes-Oxley. In the past, it has been a huge challenge for companies to find data centers with a core culture of security and regulatory compliance. Our data centers, including the new Metro Detroit Data Center, reflect our commitment to protecting our clients and their sensitive data,” Mike Klein, co-CEO of Online Tech said. The Metro Detroit data center will feature fiber connectivity to eight different telecommunications providers and will support 1.2 MW of IT load. Investments in healthcare startups are growing in Detroit and across the Midwest, and Online Tech’s infrastructure will be able to meet this demand with its security and regulatory compliance. Recently, Online Tech awarded $1,000 scholarships to students in information technology and healthcare IT at the University of Michigan. “This new facility is a major milestone for Online Tech because it is our fourth data center and it brings us to an overall total of 100,000 square feet of gross data center space. This facility will allow us to serve the large Detroit market, where we see strong demand for the secure, compliant cloud and hosting services that only Online Tech offers, but our growth won’t stop there,” Yan Ness, co-CEO of Online Tech said. “We see similar opportunities for us in other markets in the Great Lakes region and the Midwest, and we expect to continue our growth strategy by expanding our portfolio of data centers into other cities in the near future.” This article originally appeared on TheWHIR. (Comment on this)
2:30p	Chilldyne Debuts New Liquid Cooling Technology A server equipped with Chilldyne’s liquid cooling technology, which uses negative pressure to create a “leak proof” approach to liquid cooling for high-density servers. (Photo: Chilldyne) Supercomputing conferences often serve as a launchpad for liquid cooling technologies. This year’s SC13 conference marks the public debut for Chilldyne, a startup out of Carlsbad, Calif. that addresses one of the most common concerns about liquid cooling – the potential for equipment to be damaged by fluid leaks. Chilldyne is a spin-out of Flometrics, a company which specializes in fluid dynamics and thermodynamics engineering for the aerospace industry. Chilldyne uses direct-to-chip warm water cooling of CPUs using modified air-cooled heat sinks. Warm water cooling is a strategy for environments that focus the cooling as close as possible to the heat- generating components, and allows you to use your chiller less, or not at all. A distinctive feature of Chilldyne’s approach is its use of a low-flow rate and negative pressure to create a “leak-proof” system in which fluid retreats from the electronics rather than spilling. In this video, Chilldyne President and CEO Pete DeAngelis provides an overviw of the technology. This video runs about 2 minutes. Chilldyne’s approach includes a system of water piping that runs through the heat sink, and an exterior Cool-Flo unit that includes the pump and heat exchanger, which can connect to an exterior cooling tower. DeAngelis says Chilldyne’s revenue model will focus on technology licensing, OEM/ODM manufacturing and indirect channel sales for server-side components. (Comment on this)
3:00p	SC13 News: Micron Develops Parallel Processing Architecture High Performance Computing announcements this week from vendors at the Supercomputing Conference in Denver highlight a new parallel processing architecture from Micron, and customer wins from DataDirect Networks and SGI. Micron develops parallel processing architecture.  Micron announced the development of a fundamentally new computing architecture capable of performing high-speed, comprehensive search and analysis of complex, unstructured data streams. Its Automata Processor (AP) is an accelerator that leverages the intrinsic parallelism of memory and aims to dramatically advance computing capabilities in areas such as bioinformatics, video/image analytics, and network security which pose challenges for conventional processor architectures because of the amount of complex, unstructured data. Unlike a CPU, the AP is a computing fabric comprised of tens of thousands to millions of processing elements interconnected to create a task-specific processing engine capable of solving problems with unprecedented performance. ”The Automata Processor is a breakthrough technology that is designed to use advanced memory-based processing to solve complex computing challenges that existing solutions are not able to tackle effectively,” said Chirag Dekate, IDC Research Manager, HPC. “This technology has the potential to solve some of the world’s most complicated data-intensive challenges, including real-time security that could dramatically affect anti-terrorism efforts, or the highly efficient analysis of complex plant genomes that could allow scientists to rapidly advance their research agendas beyond what is possible today.” DataDirect Networks selected for multi-petabyte cloud. Data Direct Networks (DDN) announced that the Sunshine State Education & Research Computing Alliance (SSERCA) has selected DataDirect Networks Web Object Scaler (WOS) high-performance object storage to drive the development of a statewide research community. The multi Petabyte WOS object storage system will provide thousands of researchers across all disciplines seamless access to cloud-based scientific computing, communication and educational resources. This will enable SSERCA to consolidate traditional storage layers, including backup and archive, which will enable research teams to minimize storage management while maximizing productivity and efficiency. ”DDN’s ability to support multiple mechanisms for data transmission to and from third party high-performance parallel file systems, which may be tightly coupled with campus HPC assets, was a key differentiator in selecting DDN’s cost-effective solution,” said Joel Zysman, Director of HPC, University of Miami. ”This selection will enable SSERCA to share data seamlessly while leveraging our existing investment in DDN and/or other HPC storage equipment.” SGI selected by UK Atomic Weapons Establishment. SGI announced that the Atomic Weapons Establishment (AWE) in the United Kingdom has selected three SGI ICE X systems to advance sophisticated scientific and technological capabilities, and to enhance the simulation methods used to verify the safety and reliability of the Trident warhead. The systems will help AWE accelerate results by modeling more accurately and scaling up to perform more scientific models quickly – while also enabling them to easily add capacity for future simulation needs. ”We’re constantly looking for ways to enhance and develop programs to support the UK government’s nuclear weapons programme,” said Ken Atkinson, HPC Strategy and Procurement manager at AWE. “We turned to SGI again and chose their ICE X M-Cell systems to deliver a flexible and scalable infrastructure, which enables us to continue to underwrite the safety and effectiveness of the Trident warhead in the Comprehensive Test Ban Treaty era. It’s critical that advanced high-performance computing systems underpin our science program, and SGI’s supercomputing systems should enable us to quickly advance research and development. SGI met our thorough review process and was able to accommodate our tight schedule for manufacture, factory testing, delivery and commissioning.” (Comment on this)
8:25p	Submerged Supercomputer Named World’s Most Efficient System in Green 500 The Tsubame KFC system at Tokyo Institute of Technology immerses its servers and GPUs in cooling fluid. The system topped this year’s Green 500 list of the most efficient supercomputers. (Photo: NVIDIA) Green Revolution Cooling has touted immersion cooling as the most energy-efficient approach to high performance computing. Today it received a major affirmation when a Japanese system was named the world’s most efficient supercomputer, placing atop the Green 500 by using a combination of CPUs and graphics processing units (GPUs) immersed in GRC’s cooling fluid. The Tsubame-KFC system at the Tokyo Institute of Technology turned in a record performance to place first , with an efficiency of 4.5 gigaflops per watt, marking the first time a system had ever achieved 4 gigaflops per watt or more. That was about 25 percent more efficient than the list’s number-two, Cambridge University’s Wilkes, at 3.6 gigaflops per watt. The Green 500 was created to spotlight the world’s most energy-efficient supercomputers. Tsubame-KFC uses two Intel Xeon E5 (Ivy Bridge) processors and NVIDIA K20 Kepler GPUs, which were immersed in a single CarnoJet cooling enclosure from Green Revolution. Satoshi Matsuoka, a professor at Tokyo Tech, said the 4.5 gigaflops per watt was substantially improved from the to 3.1 gigaflops per watt for Tsubame 2.5 and less than 1.0 gigaflops per watt for Tsubame 2.0, which held world’s greenest title when it debuted in 2010. As a point of reference, Tianhe-2, the fastest supercomputer in the world according to the Top500 list, achieves an efficiency of 1.9 gigaflops per watt. The November 2013 list marks the first time that all of the top 10 systems on the Green500 are heterogeneous systems – using computational building blocks that consist of CPUs, GPUs and co-processors. All incorporate Intel CPUs combined with NVIDIA GPUs. The Top500 newcomer Piz Daint at the Swiss National Supercomputing Center came in at number 4, with 1.7MW of total power, and 3.185 gigaflops per watt in energy efficiency. “A decrease in the average measured power coupled with an overall increase in performance is an encouraging step along the trail to exascale,” noted Wu Feng of the Green500. Graph 500 A November 2013 Graph 500 list was also released – ranking the suitability of supercomputing systems for data intensive applications. The intent of the Graph 500 is to develop a compact application that has multiple analysis techniques (multiple kernels) accessing a single data structure representing a weighted, undirected graph. Backed by over 50 international HPC experts from academia, industry, and national laboratories, the Graph 500 will set large-scale benchmarks for big data applications. IBM took the top 3 spots on the Graph 500, going to Lawrence Livermore National Laboratory’s Sequoia, Argonne National Laboratory’s Mira and Forschungszentrum Juelich’s (FZJ) JUQUEEN, which all use IBM Blue Gene/Q systems.  China’s Milkyway-2, which ranked number 1 on the Top500, ranked number 6 on the Graph 500. (Comment on this)

<< Previous Day

2013/11/21 [Calendar]

Next Day >>