Blade
Download as DOC, PDF•1 like•237 views
The document is about a seminar presentation on blade servers given by Sarath T.S. from the Department of Computer Engineering at MTI Thrissur. It includes acknowledgements thanking various people who helped with the seminar, an abstract providing an overview of blade servers, and an introduction section describing the history and features of blade servers. The introduction discusses how blade servers optimize space and power usage while maintaining server functionality, and how their compact design allows for higher server densities in data centers.
![Seminar 2013-14
BLADE SERVER
ABSTRACT
A blade
server is
a
stripped
down server
computer with
a modular
design optimized to minimize the use of physical space and energy. Whereas a
standard mount server can function with (at least) a power cord and network
cable, blade servers have many components removed to save space, minimize
power consumption and other considerations, while still having all the
functional components to be considered a computer. A blade enclosure, which
can hold multiple blade servers, provides services such as power, cooling,
networking, various interconnects and management. Together, blades and the
blade enclosure form the blade system. (Different blade providers have differing
principles regarding what to include in the blade itself, and in the blade system
altogether.)
In a standard server-rack configuration, 1U (one rack unit, 19" [48 cm] wide
and 1.75" [4.45 cm] tall) defines the minimum possible size of any equipment.
The principal benefit and justification of blade computing relates to lifting this
restriction so as to reduce size requirements. The most common computer
rack form-factor is 42U high, which limits the number of discrete computer
devices directly mountable in a rack to 42 components. Blades do not have this
limitation; a
s of 2009, densities of up to 128 discrete servers per rack are achievable with
blade systems.
Dept. of Computer Engg.
2
MTI Thrissur](https://image.slidesharecdn.com/blade-140307061753-phpapp01/85/Blade-2-320.jpg)
Blade
- 1. Seminar 2013-14 BLADE SERVER ACKNOWLEDGEMENT I would like to thank HOD Mrs. NJANAMBIKA .V.A Dept. of Computer Engineering for giving me such an opportunity to do this seminar . I would like tothank sincerely to Mrs. Thajbi, Mrs. Yamini who has been helpful and cooperative for giving me the necessary guidelines for doing my seminar .I also express my sincere gratitude to Ms. Navya , Mrs. Aiswarya&Mrs.Seena who has been helpful throughout the Seminar presentation. I would like to express my gracious gratitude to those other peoples who spent their valuable time with me to discuss about the topic help me to collect all the relevant information’s about the seminar topic and to give necessary advices . Over all these thanks to almighty for blessings . SARATH T.S Dept. of Computer Engg. 1 MTI Thrissur
- 2. Seminar 2013-14 BLADE SERVER ABSTRACT A blade server is a stripped down server computer with a modular design optimized to minimize the use of physical space and energy. Whereas a standard mount server can function with (at least) a power cord and network cable, blade servers have many components removed to save space, minimize power consumption and other considerations, while still having all the functional components to be considered a computer. A blade enclosure, which can hold multiple blade servers, provides services such as power, cooling, networking, various interconnects and management. Together, blades and the blade enclosure form the blade system. (Different blade providers have differing principles regarding what to include in the blade itself, and in the blade system altogether.) In a standard server-rack configuration, 1U (one rack unit, 19" [48 cm] wide and 1.75" [4.45 cm] tall) defines the minimum possible size of any equipment. The principal benefit and justification of blade computing relates to lifting this restriction so as to reduce size requirements. The most common computer rack form-factor is 42U high, which limits the number of discrete computer devices directly mountable in a rack to 42 components. Blades do not have this limitation; a s of 2009, densities of up to 128 discrete servers per rack are achievable with blade systems. Dept. of Computer Engg. 2 MTI Thrissur
- 3. Seminar 2013-14 BLADE SERVER CONTENTS 1. INTRODUCTION 2. HISTORY 3. FEATURES 4.TYPES 5. ADVANTAGES AND DISADVANTAGE 6. APPLICATIONS 8. FUTURE SCOPE 9. CONCLUSION 10. REFERENCE Dept. of Computer Engg. 3 MTI Thrissur
- 4. Seminar 2013-14 BLADE SERVER 1. INTRODUCTION With the high cost of data center floor space and current advances in technology, new installations with denser cabinets that require more power and cooling continues to be the trend. Besides the challenges that new installations present, equipment cabinet upgrades can also be a problem as the existing power and cooling currently provided may not support the new cabinet configuration. Surveys show that Information Technology equipment is typically replaced every 2 to 5 years depending on the individual organization and its needs. Surveys also show (See Chart 1) that when asked about their top 3 concerns; Heat/Power Density is the number one concern of Data Center Management. High density applications like cluster server configurations have in some cases pushed the kW power demands as high as 40 kW per cabinet. The required power depends on the equipment, how dense the cabinet is and whether redundancy is required. This has led to new and innovative solutions for providing cabinet level power utilizing CDU’s Dept. of Computer Engg. 4 MTI Thrissur
- 5. Seminar 2013-14 BLADE SERVER (Cabinet distribution Units).Distributed server architectures based upon "blades" are rapidly emerging in the data centers of corporations and Internet service providers. Historically, servers only assumed the form factor of a re-purposed desktop computer or rack-mountable appliance. These servers utilized high-quality components and leverage additional memory and hard disk capacity. Now server form factors are evolving into single PC cards that can be plugged into a chassis as a single module. Bladed servers stack numerous independent lower-end servers within a single chassis. Chassis can accept anywhere from eight to 24 blades. Each blade is an independent system with its own memory, processor and network connection. Due to their compact size, multiple blade servers can be placed in a single server rack or enclosure, allowing numerous systems to share electricity and HVAC resources. Gartner Dataquest predicts that worldwide server blade shipments will increase from 84,410 units in 2002 to more than one million in 2006. The IT research firm anticipates that revenue from server blades will reach at least $1.2 billion during this time period. Blades are thus becoming the one major segment of the server market that is experiencing escalating growth. The popularity and fast growth of the blade server can be attributed to cost-savings that the device accrues to its users, especially hosting firms and service providers. Because more than 250 blade servers can be effectively placed into a single rack, it is possible for hosting firms with data center operations to quadruple their hosting capacity with the devices. By comparison, most of today's low-end servers have only a single computer in one enclosure, allowing only one 42 systems to fit in an industry-standard rack. Because blade servers are small, consume less power and generate less heat than an average server, they are emerging as an ever-popular option for niche Web hosting services. With the cost of data center space at approximately $300 per square foot and with energy costs increasing throughout continental North America due to deregulation, Dept. of Computer Engg. 5 MTI Thrissur
- 6. Seminar 2013-14 BLADE SERVER blade servers have become the de facto standard for increasing data center profitability. The systems are typically used as Web servers and caching servers that deliver Web pages to Internet browsers, SSL servers for encrypted communication, and streaming servers for audio and video transmissions. Most hosting companies and service providers appreciate the devices because they are easy to install and employ dedicated software that improves their administration, performance and reliability. The devices are also increasingly being utilized as firewall devices and to increase capacity in corporate data centers. Blade severs are excellent devices for hosting companies with large existing data center deployments who want to capitalize upon existing PCI expansion space. The inclusion of a single blade allows a hosting company to double its computing power or hosting offering, utilizing the same amount of physical space. The main disadvantage concerning PCI-based blade servers is that the processors are usually not as robust as traditional high-end servers. This factor limits the use of the server to functions such as low-end Web hosting. While blade severs have the capacity to serve streaming video and other demanding applications, often the emerging technology is used for back up or storage purposes. Lower-end models usually depend on server or operating system virtualization, causing the server to run much slower than traditional equipment. It is thus incumbent upon a reseller to determine whether a hosting company is utilizing blade servers for mission-critical deployments and whether those servers can accommodate demanding applications. Resellers who operate their own equipment might want to consider utilizing blade servers as an effective technology to add firewall or other security specific capacities to their collocated equipment. Dept. of Computer Engg. 6 MTI Thrissur
- 7. Seminar 2013-14 BLADE SERVER 2. HISTORY Developers placed complete microcomputers on cards and packaged them in standard 19-inch racks in the 1970s soon after the introduction of 8-bit microprocessors. This architecture operated in the industrial process control industry as an alternative to minicomputer control-systems. Early models stored programs in EPROM and were limited to a single function with a small real-time executive. The VME bus architecture (ca. 1981) defined a computer interface which included implementation of a board-level computer installed in a chassis backplane with multiple slots for pluggable boards to provide I/O, memory, or additional computing. The PCI Industrial Computer Manufacturers Group PICMG developed a chassis/blade structure for the then emerging Peripheral Component Interconnect bus PCI which is called Compact PCI. Common among these chassis based computers was the fact that the entire chassis was a single system. While a chassis might include multiple computing elements to provide the desired level of performance and redundancy, there was always one board in charge, one master board coordinating the operation of the entire system. PICMG expanded the Compact PCI specification with the use of standard Ethernet connectivity between boards across the backplane. The PICMG 2.16 Compact PCI Packet Switching Backplane specification was adopted in Sept 2001 (PICMG specifications). This provided the first open architecture for a multi-server chassis. PICMG followed with the larger and more feature-rich Advanced TCA specification targeting the telecom industry's need for a high availability and dense computing platform with extended product life (10+ years). While Advanced TCA system and Dept. of Computer Engg. 7 MTI Thrissur
- 8. Seminar 2013-14 BLADE SERVER boards typically sell for higher prices than blade servers, Advanced TCA suppliers claim that low operating-expenses and total-cost-of-ownership can make Advanced TCA-based solutions a cost-effective alternative for many building blocks of the next generation telecom network. The first commercialized blade server architecture was invented by Christopher Hipp and David Kirkeby and their US 6411506was assigned to Houston-based RLX Technologies. RLX, which consisted of mostly former Compaq Computer Corp employees, including Hipp and Kirkeby, shipped the first commercial blade server in 2001 and were acquired by Hewlett Packard (HP) in 2005. In February 2006, Blade.org was established to increase the number of blade platform solutions available for customers and to accelerate the process of bringing them to market. It is a collaborative organization and developer community focused on accelerating the development and adoption of IBM blade server platforms. The name blade server appeared when a card included the processor, memory, I/O and non-volatile program storage (flash memory or small hard(s)). This allowed manufacturers to package a complete server, with its operating system and applications, on a single card / board / blade. These blades could then operate independently within a common chassis, doing the work of multiple separate server boxes more efficiently. In addition to the most obvious benefit of this packaging (less space-consumption), additional efficiency benefits have become clear in power, cooling, management, and networking due to the pooling or sharing of common infrastructure to supports the entire chassis, rather than providing each of these on a per server box basis. Dept. of Computer Engg. 8 MTI Thrissur
- 9. Seminar 2013-14 BLADE SERVER 3. FEATURES Essentially, blade servers are a card on the server: a single motherboard that contains a complete computer system, including processors, memory, network connections and related electronic devices. If you insert a number of blade server rack or cabinet in the plane, then the rack or cabinet infrastructure can be shared, at the same time with redundant features. Recognized the advantages of blade servers have two, one to overcome the shortcomings of the cluster server chip, and the other is optimized to achieve the cabinet. Then choose the main blade server for several reasons to consider the following: 1.Lower Hardware Costs Server with the traditional difference is that each blade server rack and does not need a separate infrastructure, relatively low-cost blade servers. Multiple systems through the sharing of power and cooling equipment, management of hardware and cabling systems can greatly reduce the cost of each server. 2. Simplified Deployment and Maintenance The deployment of multiple servers is a time-consuming and resource-intensive Dept. of Computer Engg. 9 MTI Thrissur
- 10. Seminar 2013-14 BLADE SERVER process. Administrators need to be installed on each server rack for power and network cabling, and to install the software. In high-density environment, in particular the deployment of wiring is a headache in one of the problems. The use of blade servers, administrators only need to be installed on the rack and wiring, no need to separate the blade server routing. The addition of new computer resources or just insert a new blade server, just like now to add a hard drive as easy. Because blade servers can share a number of redundant power supply, so to minimize the wiring of the rack. In addition, the built-in converter to Ethernet data and management networks, and even integrated with KVM connection. The use of automated software tools, administrators can operate through a network; we can quickly and easily install the software to one or more blade servers. Upon completion of the blade server software installed, the administrator will be able to use remote management tools to conduct a comprehensive management. 3. To Maximize the Use of Data Center Space According to the design and the different suppliers, blade servers can make the server than the current density of rack-optimized 1U system increased 100% to 8000%. 4. To Reduce Power Consumption Blade servers in order to reduce power consumption, the majority of suppliers will be in Dept. of Computer Engg. 10 MTI Thrissur
- 11. Seminar 2013-14 BLADE SERVER some blade servers use low-power processor. Even if the blade server does not use lowpower processors, the power consumption is less than "comprehensive" server, blade server power consumption because fewer components. In addition, suppliers can afford to use high-efficiency power supply costs, because power will be shared across multiple servers. In addition, blade servers as a result of lower power consumption, they produce less heat, thus reducing the power consumption of the cooling system. At present, such high-density blade server market prospects brought about by foreign countries has been IBM, HP, SUN and DELL and other manufacturers of the value, they have announced the launch of its own blade servers. Domestic dawn, Lenovo, and other companies have also launched a wave of their respective products, as a result of the overall performance of blade servers more performance, so blade servers will be more business users of all ages. Dept. of Computer Engg. 11 MTI Thrissur
- 12. Seminar 2013-14 BLADE SERVER 4.TYPES Since blade enclosures provide a standard method for delivering basic services to computer devices, other types of devices can also utilize blade enclosures. Blades providing switching, routing, storage, SAN and fiber-channel access can slot into the enclosure to provide these services to all members of the enclosure. Blade vs. Rack mount Blade servers are outgrowing at a faster rate than traditional rack mount servers. A recent Gartner study found that blade servers are the fastest growing segment of the server market. One reason is simply because blades are easy to configure and manage. Using a blade can be as easy as using an expansion card — only this "expansion card" comes with one to four processors, memory and disk storage. Blades are considered to be hot-swappable, which means you can add new blades or remove existing ones while the system is powered on. Traditionally, blade servers have been deployed in data centers and large enterprise environments, but the small business is looking at blades for the same reason enterprise has previously: They take up less floor space than traditional rack mount servers, they require less power and fewer IT management resources are required than with a rack mount. Blade servers are scalable to any physical infrastructure. It is important to remember that blades are not suited to all applications and cannot replace a large-scale server in all instances. There is an also proprietary interest at stake. An HP blade, for example, cannot be plugged into an IBM blade chassis. As a result, third-party vendor blades have to be designed for specific branded chassis. Dept. of Computer Engg. 12 MTI Thrissur
- 13. Seminar 2013-14 BLADE SERVER Blade vs. Rack mount — A Quick Comparison Blade Servers Rack mount Servers Each has its own power supply, fan and cables. Shared infrastructure for fans, power supplies, Ethernet switching, storage. Networking and storage is built into the chassis, which eliminates cables. Large physical floor space required to Small form factor can use up to half the space of a rack mount house rack mount. server. More difficult deployment. SMB may Installation requires no special tools or require on-site technicians to make expertise, semi-technical or non-technical additions to the rack mount. staff can deploy the blades. Able to hotswap. Proprietary nature limits the ability to More choice in system suppliers for mix and match components from multiple acquisitions. Multiple components from suppliers different suppliers can be used in one in one chassis. chassis. Many blades still have cooling issues due A variety of rack mount coolers are to available. Separate fans help cooling shared cooling on the chassis issues. Dept. of Computer Engg. 13 MTI Thrissur
- 14. Seminar 2013-14 BLADE SERVER Virtualization & Common Blade Server Computing Environments Virtualization is another area of computing that has been a driving force behind blades. Virtualization involves emulating multiple servers on one hardware platform. Running multiple operating systems on a single computer or storage virtualization where you have the amalgamation of multiple network storage devices into what appears to be a single storage unit are examples of virtualization. With a blade server you have the option to combine blades with virtualization software to consolidate workloads, each running on its own instance of the OS (using the same or a different OS). With blades, separate operating systems and applications can co-exist on one server and users of the system are able to access more memory and processing power as their workload demands it. Blades are frequently deployed in data centers and high-performance computing environments (a branch of computer science that concentrates on developing supercomputers and software to run on supercomputers.), and can serve as application servers, databases, e-mail or Web servers, and more. Large data centers and telecommunications service providers benefit from the use of blades as they provide the means for a large business to respond quickly to changes in business conditions. High-traffic Web sites are another example of where blades can help — if you plan to host an online event, broadcast events live or something of that nature blades are a perfect solution as they allow you to quickly add memory and processing power to compensate for unusually high traffic to the Web site. Overall, where a business or group would use several different servers for different applications, it makes sense to combine the multiple servers into one blade server to make for better manageability. Blades are often viewed as a solution for large Dept. of Computer Engg. 14 MTI Thrissur
- 15. Seminar 2013-14 BLADE SERVER enterprise, but really the IT cost and manageability of a blade solution makes it wellsuited for smaller businesses and organizations. To this end, many of the main blade vendors market specific blade solutions and packages to the SMB. BLADE SERVER IMAGES 8U Rack mounts Blade Sever Chassis Dept. of Computer Engg. 15 MTI Thrissur
- 16. Seminar 2013-14 BLADE SERVER HP ProLiant DL385 IBM eServer Blade Center T Dept. of Computer Engg. 16 MTI Thrissur
- 17. Seminar 2013-14 Dept. of Computer Engg. BLADE SERVER 17 MTI Thrissur
- 18. Seminar 2013-14 BLADE SERVER 5. ADVANTAGES AND DISADVANTAGE ADVANTAGES Condensed, high-density... Blade servers allow more processing power in less space which simplifies cabling, storage, and maintenance. Blades are often used for grid computing. The advantage of blade servers is not only that one "cabinet" houses multiple servers that share power source and other components, but also from the consolidation of related resources, such as storage and networking equipment, into a smaller architecture than would be the case with a farm of regular servers. Load balancing and failover... Like most clustering applications, another advantage of blade servers is that blade servers can be managed to include load balancing and failover capabilities. This can be done with a farm of regular servers as well, but since blade servers share much simpler and slimmer infrastructure and often come designed for this task from the manufacturer, load balancing and failover management tends to be easier with blades. In case any hardware failure is found on a server blade or the chassis itself, selfdiagnostics function automatically starts and the faulty area can be identified by the indicator display per server blade. Dept. of Computer Engg. 18 MTI Thrissur
- 19. Seminar 2013-14 BLADE SERVER Power consumption & power management... A blade server most importantly reduces power consumption and improves power management. Consolidating power supplies into the blade chassis reduces the number of separate power supplies needed and also reduces the power requirements per server. Because individual server blades are stripped to bare minimum and do not have other features found in conventional servers such as keyboard, graphic cards, and others, that employ less devices that need to be powered. This reduces overall power consumption. One blade server with 16 server blades uses much less power than 16 individual fullsize servers. Lower management cost... Another big advantage over traditional servers. Blade servers often come with a single interface that is used to manage all individual servers within chassis. Server consolidation and resource centralization also simplifies server deployment, management, and administration. Dept. of Computer Engg. 19 MTI Thrissur
- 20. Seminar 2013-14 BLADE SERVER Hardware configuration management, operation status monitoring, and fault monitoring are centrally controlled, thus reducing the system administrator’s burden. Administrator can be notified of an error by such means as e-mail in a timely manner Network and other cabling... Imagine that you have 16 individual stand-alone servers. Each of them needs to be connected to the network which calls for yards or even miles of cabling. Blade servers simplify cabling requirements and reduce wiring by a great percentage. Power cabling, operator wiring (keyboard, mouse, etc.) and communications cabling (Ethernet, SAN connections, and cluster connection) are greatly reduced as well. Flexibility, modularity, and ease of upgrading... Modern blade servers are designed in a way that it is possible to take out and add in server blades while the system is up and running with just a minor configuration in the admin interface. New processor, communications, storage and interconnect technology can be implemented in blades that install into existing equipment with no or minimal disruption to the functionality of the whole system. You can expand or reconfigure the system without stopping your jobs other than those related to the server blade subject to upgrade or replacement. Modules can mix in the same blade server chassis. Server blades within a chassis do not necessarily need to be identical. Depending on the particular type of your chassis, one server blade can have for example Intel Xeon and the other Intel Itanium. You can have one server blade with Windows and another one with Linux. Dept. of Computer Engg. 20 MTI Thrissur
- 21. Seminar 2013-14 BLADE SERVER Deployment and scalability... The same applies for adding new server blades into the system. Once a blade server chassis has been installed, additional servers can be added by simply sliding them into additional bays while the system as a whole is up and running. Blades can be scaled up as needed by just simply adding a new server blade. It is certainly easier to deploy additional server blade than a new stand-alone server. Disaster management... Even though the purpose of packing server blades into a single chassis is to share servicing resources, chassis can be configured with redundant power modules for failover. DISADVANTAGES As with anything, benefits do not come for free. Blades also include some disadvantages. Expensive configuration... Although plugging in a new server blade into the blade server is easy once the system is running, initial configuration can be labor-intensive and expensive in complex application environments. This disadvantage comes with the fact that blade servers are specialized computing equipment and their configuration and administration often requires training provided by the vendor which may not be cheap unless you have a special free-training deal with the vendor. Dept. of Computer Engg. 21 MTI Thrissur
- 22. Seminar 2013-14 BLADE SERVER Expensive tool, i.e. economies of scale... If you do not fill the blade chassis with server blades, you are not fully utilizing it. It does not make much sense to purchase a blade chassis for $5,000 and then run the whole system with just 2 server blades in it. Blade chassis are often made to hold 14 or 16 server blades. The general rule of thumb is that blade servers are not suitable and economical for applications requiring less than 5-10 servers. Applications requiring less than 5-10 server blades (10-20 CPUs) are best devoted to standalone server systems. Vendor-lock... Blade systems vary between manufacturers. Once you spend $50,000 on a blade server from a particular vendor, it is not always easy to switch to another vendor because of servicing agreements and also because a competitor is unlikely going to have the same expertise in your equipment as your vendor. You could theoretically use your server blades in competitor's blade chassis, but practically blade chassis are not standardized. It is unlikely that IBM would get together with Dell and HP to share chassis. Chassis is what makes their products unique. Server blades are often designed to only run in the company's own chassis. Business case... Blade servers are not the best solution for everything. If you have a very large transaction processing application requiring high read/write ratios, then you may run Dept. of Computer Engg. 22 MTI Thrissur
- 23. Seminar 2013-14 BLADE SERVER into bottle neck with your bus speeds, memory limitations, disk access, and network I/O. Email and Web serving are situations where blade computing suits well. Heating and cooling... One often forgotten disadvantage is HVAC. While individual stand-alone servers can be distributed throughout the building and may not necessarily need special accommodations for cooling, blade servers being very powerful these days produce massive amounts of heat per square foot. If untreated, this could melt them down. When purchasing blade servers, it is important to keep in mind that additional resources will be needed also for HVAC. 6. APPLICATIONS As blade servers pack increasingly more punch into smaller form-factors, their use is spreading beyond data center applications. The integration of newer technologies makes blades attractive for high-performance computing purposes such as distributed computing, rendering/imaging and data analysis. Blade servers have gained ground over the past few years as an efficient, condensed computing solution in large data center implementations. They are widely used as replacements for pizza box servers and large rack mounts (Figure 1). Dept. of Computer Engg. 23 MTI Thrissur
- 24. Seminar 2013-14 BLADE SERVER The advantages inherent in blade server technology include smaller form-factors, denser computing, expandability, hot-swap capabilities, flexible and fail-safe architecture, reduced downtime, increased redundancy, simplified server management, easier hardware and software integration, and lower heat dissipation and power requirements. These improvements over traditional data center servers provide for a massive increase in deployable resource density and an overall reduction in long-term costs. Each blade in a chassis is typically a self-contained server. Data center consolidation, advanced communications and remote management of servers required to run 24/7 are just a few of the reasons for large-scale deployments. These deployments are becoming more popular for telecom, telephone and cellular carriers, insurance companies, tax preparers, state and local government agencies and educational institutions, among others. Although blade servers have a long way to go before they are the standard deployed technology throughout the data center, they are beginning to appear in less traditional implementations such as high-performance computing (HPC). Dept. of Computer Engg. 24 MTI Thrissur
- 25. Seminar 2013-14 BLADE SERVER POWER SELECTION AND DISTRIBUTION FOR HIGH DENSITY CABINETS There are many different circuit combinations that can be used to provide power to today’s densely populated cabinets. Today’s total power requirements for rack mounted devices, per cabinet, can be as low as a 2-3 kW’s but may exceed 40kW depending on the application, whether redundant power is required and of course, the devices that are installed within the cabinet. Many servers today are designed to take a large range of input voltage from 120V to 208V. Typically smaller servers that may or may not be rack mountable require 120 V. Most of today’s larger high density servers, such as blade servers require 208V power. While the floor space required achieving a constant level of computing and storage capacity has shrunk, the energy efficiency of the equipment has not increased at the same rate. As a result, the power consumed and the heat dissipated has increased significantly within the cabinet foot print. With increased power consumption (a trend that is likely to continue) comes important decisions that must be made with regards to cabinet level power requirements and selecting the proper CDU (Cabinet Distribution Unit). Tough choices like what power is brought down to the cabinet level such as 120V 15 or 20A, 208V 20 or 30A, 208V 30A 3-Phase or even 208V 60A 3-Phase are important decisions that affect costs, performance and the ability to provide for future expansion. The number of power drops that must be run to each cabinet greatly affects the cost and whether or not there is sufficient power available for current needs and expansion. This is especially true in critical applications where redundant power is necessary. Each power drop run to a cabinet costs $500 to $1,500 or more; as well as having other consequences such as the number of CDU’s required or whether or not cable management is a problem. Dept. of Computer Engg. 25 MTI Thrissur
- 26. Seminar 2013-14 BLADE SERVER 8. FUTURE SCOPE Only time will tell if future capacity demands will continue to increase power consumption and cooling levels to higher and higher levels. Current ASHRAE projections certainly predict that they will. Other technologies such as virtualization, dual core processors, and new more efficient processor designs along with current and future software solutions will ultimately determine the outcome. Blades help enable virtualization. Virtualization allows data centers to tune resources dynamically to business needs—as they change. Virtualization in the modular computing paradigm decouples resources from any particular need: resources are seen as pools that can be reconfigured and redeployed according to demand. Blade modularity plus virtualization tools allow IT departments to step away from traditional data Center approaches and make resources more pliable to changes in business activities and needs, while reducing management burden. Resources can be dynamically optimized and adapted for any application. Virtualization tools will enable automated redeployment of servers, operating systems and application software and allow reconfiguration of switches on the fly, based on data center policies, business planning and strategy and system monitors, threshold triggers and control mechanisms. And it will happen without an administrator visiting the equipment. For example, retail stores or airlines running special Web-only sales events can automatically redeploy underutilized resources to the front-end Web services to maintain service levels to customers. Also, if sales or service demands unexpectedly increase; policy-driven, blades-based modular data center can shift resources in minutes instead of hours to accommodate the loads while administrator is working on some other critical initiative. Dept. of Computer Engg. 26 MTI Thrissur
- 27. Seminar 2013-14 BLADE SERVER 9. CONCLUSION It is clear that by close review of the application and it’s power requirements decisions can be made that will allow for future expansion, denser cabinets, greatly reduced costs (in many cases), power redundancy (if required) and will optimize cable and CDU requirements. It is only through making the power requirements a priority at the beginning of the project rather than an afterthought that these goals are achieved. This has become especially important with today’s higher power demands and the ever increasing push for lower costs, greater efficiencies and improved uptime availability. Dept. of Computer Engg. 27 MTI Thrissur
- 28. Seminar 2013-14 BLADE SERVER 10. REFERENCE http://phys.org/news/2013-06-google-internet-beaming- BLADE SERVER _1.html www.RavenInd.com www.google.// BLADE SERVER Dept. of Computer Engg. 28 MTI Thrissur
- 29. Seminar 2013-14 BLADE SERVER Project presented by SARATH THALEKKARA Mob : 9567781224 Dept. of Computer Engg. 29 MTI Thrissur