2. Grid Computing

8/4/2019 2. Grid Computing

1/9

GRID COMPUTING

INTRODUCTION

1. Enterprise grid computing is an emerging IT architecture that delivers flexible

enterprise information systems that are more resilient and less expensive thantraditional legacy systems. In grid computing, groups of independent hardware andsoftware components are pooled and provisioned on demand to meet changingbusiness needs. The accelerating adoption of grid technology is in direct response tothe challenges facing information technology (IT) organizations. With todays rapidlychanging and unpredictable business climate, IT departments are under increasingpressure to manage costs, increase operational agility, and meet IT service-levelagreements (SLAs).

2. Using enterprise grid computing technology, IT departments can adapt to rapidchanges in the business environment while delivering high service levels. Enterprise

grid computing has revolutionized IT economics by extending the life of existingsystems and exploiting rapid advances in processing power, storage capacity, energyand space efficiency, and network bandwidth.

3. This article provides an overview of grid computing, highlights the benefits ofusing it, and describes key grid computing techniques that enable IT resourceconsolidation, agile IT operations, predictable high performance and scalability, andcontinuous availability.

WHAT IS GRID COMPUTING?

4. One way to think about grid computing is as the virtualization and pooling of ITresources compute power, storage, network capacity, and so on into a single set of

shared services that can be provisioned or distributed, and then redistributed asneeded. Just as an electric utility uses a grid to deal with wide variations in powerdemands without affecting customer service levels, grid computing provides ITresources with levels of control and adaptability that are transparent to end users, butthat let IT professionals respond quickly to changing computing workloads.

5. The term utility computing is often used to describe the metered (or pay per use)IT services enabled by grid computing. Cloud computing (where dynamically scalableand often virtualized resources are provided as a service over the internet) is anotherterm that describes how enterprises are using computing resources on both private andpublic networks over the internet. Because grid computing provides superior flexibility, itis the natural architectural foundation for both utility and cloud computing.

6. As workloads fluctuate during the course of a month, week, or even through asingle day, the grid computing infrastructure analyzes the demand for resources in realtime and adjusts the supply accordingly.


2/9

2

Compare the difference in infrastructures between traditional computing and gridcomputing.

GRID COMPUTING OPERATES ON THESE TECHNOLOGY PRINCIPLES

(a) Standardization. IT departments have enjoyed greater interoperabilityand reduced their systems management overhead by standardizing on operatingsystems, servers, storage hardware, middleware components, and networkcomponents. Standardizing also helps reduce operational complexity in the datacenter by simplifying application deployment, configuration, and integration.

(b) Virtualization. Virtualizing IT resources means that applications are not

tied to specific server, storage, or network components and can use anyvirtualized IT resource. Virtualization occurs through a sophisticated softwarelayer that hides the underlying complexity of IT resources and presents asimplified, coherent interface used by applications and other IT resources.

(c) Automation. Because of the potentially large number of components bothvirtual and physical grid computing demands large-scale automation of IToperations. Each component requires configuration management, on-demandprovisioning, top-down monitoring, and other management tasks. A grid


3/9

3

management solution must ensure that infrastructure cost savings do notevaporate as a result of hiring additional staff to manage the grid. ITadministrators need a top-down view from the end-user or application level sothey can effectively measure service levels and proactively resolve problems.Combining these capabilities into a single, automated, integrated solution for

managing grids gives organizations a maximum return on their grid investment.THE BENEFITS OF GRID COMPUTING

7. Using a grid computing architecture, organizations can quickly and easily createa large-scale computing infrastructure from inexpensive, off-the-shelf components.Other benefits of grid computing includes as follows:-

(a) Quick response to volatile business needs

(b) Real-time responsiveness to dynamic workloads

(c) Predictable IT service levels

(d) Reduced costs as a result of improved efficiency and smarter capacity

planning.

GRID COMPUTING OPERATES ON THREE BASIC TECHNOLOGY PRINCIPLES

(a) Standardize hardware and software components to reduce incompatibilityand simplify configuration and deployment.

(b) Virtualized IT resources by pooling hardware and software into sharedresources.

(c) Automate systems management, including resource provisioning andmonitoring.

RESPONDING QUICKLY TO VOLATILE BUSINESS NEEDS

8. Businesses today operate in an unpredictable, global environment. Staying ontop of business demands, competitive threats, supply chain risks, and regulatoryrequirements is increasingly challenging. Businesses expect their IT groups to turn ona dime and be able to, for example, change a pricing model to beat a competitor, adjustthe order management process to accommodate new regulatory requirements, andintegrate acquired companies. With an underlying grid infrastructure, IT has the agility torespond quickly to these types of changing business needs.

RESPONDING TO DYNAMIC WORKLOADS IN REAL TIME

9. Most of todays applications are tied to specific software and hardware silos,

limiting their ability to adapt to changing workloads. This costly and inefficient use of ITresources means that IT departments must overprovision their hardware so that eachapplication can handle peak or worst-case workload scenarios. Grid computing lets ITprofessionals dynamically allocate and deallocate IT resources as needed, providingmuch better responsiveness to workloads that change on a global scale.

PROVIDING PREDICTABLE SERVICE LEVELS

10. Through the use of service-level agreements (SLAs), organizations can tiebusiness requirements to IT architecture to get demonstrable metrics and proactive


4/9

4

monitoring and maintenance. This encourages a shared-service-bureau approach to IT,with the focus on measuring and meeting higher service levels and better aligning ITand business goals. A grid-based architecture eliminates single sources of failure andprovides powerful, high-availability capabilities throughout the entire software stack,protection for valuable information assets, and business continuity. It lets IT groups

eliminate expensive systems administration overhead, costly integration projects, andrunaway budgets. Reducing Costs with Improved Efficiency and Smarter CapacityPlanning Grid computing practices focus on operational efficiency and predictability.Easier grid workload management and resource provisioning puts more power in thehands of the IT staff, letting them maintain current staffing levels even as computingdemands skyrocket. With a new generation of server virtualization and clusteringcapabilities from Oracle, IT departments no longer have to overprovision to meet worst-case scenarios during peak periods. And because computing resources can be appliedincrementally when needed, organizations enjoy much higher computing and storagecapacity utilization at a reduced cost.

11. With grid computing, organizations can choose a more cost-effective scaling with

a pay-as-yougrow procurement strategy. Companies can avoid buying extra hardwareor additional software licenses before they actually need them, and they can takeadvantage of the price/performance benefits that come with rapid growth in processingpower and greater energy efficiency. Using Grid Techniques to Modernize Data CentersGrid computing IT architecture and methodology include these technologies and bestpractices.

(a) Consolidated IT resources(b) Agile IT operations(c) Predictable high performance and scalability(d) Continuous availability

12. Not every IT department will adopt every grid computing technology or technique,however, many groups are already seeing dramatic benefits by using selected Oraclegrid technologies and best practices.

CONSOLIDATING IT RESOURCES

13. Consolidating IT resources can provide dramatic cost and energy savings.Forrester Research estimates that in data centers today, the average server utilizationis only about 30 percent. Considering how many servers can be in an enterprise, thatinefficiency is staggering. Because application usage varies greatly by certain times ofthe day or year, it provides an opportunity to apply grid techniques for a combination ofbetter management, utilization, and overall efficiency. IT departments are significant

users of electricity, so they must also consider those energy costs in their data centeroperations. When metrics such as efficiency and operating margins are scrutinized, IDCestimates that power, cooling, and other management costs account for 70 percent of aservers lifetime cost. Many organizations are now putting energy efficiency andgreencomputing initiatives into their buying criteria for technology components. With thepower and space optimization that comes from consolidating resources into a gridinfrastructure, organizations can have a greener data center. As organizationscentralize and consolidate servers and storage, their overall server and storage


5/9

5

utilization increases. The IT staffs no longer needs to overprovision hardware and canimprove overall energy efficiency. Two key grid computing technologies, servervirtualization and clustering, enable the sharing of IT resources and the consolidation ofservers, storage, and even entire data centers.

14. Grid computing enables the sharing of IT resources and the consolidation of

servers, storage, applications, and even data centers. Results include reduced costsand lower power, cooling, and space requirements.

Server virtualization and clustering are two key grid computing techniques that canmake one resource look like many and many resources look like one.

CONSOLIDATING WITH VIRTUAL SERVERS

15. Many subscribers are using hypervisor-based virtual machines (VMs) toconsolidate multiple applications onto fewer centrally managed shared servers. A virtualmachine (or virtual server) is software that simulates the operation of computerhardware and allows an application to run on it as if it were a physical computer. Theadvantage of this approach is that many virtual machines can run on a single physicalcomputer, allowing the consolidation of multiple small servers onto one larger server.This approach helps establish a standardized computing environment in which to run

applications, middleware servers, database servers, and storage servers.

16. Oracles server virtualization product, Oracle VM, provides a highly efficient wayto run multiple Oracle and non-Oracle databases, middleware, and applicationenvironments on a single server running Oracle Enterprise Linux. Oracle VM lets the ITstaff quickly add or release more server resources to match spikes or lulls in demand.This increases server utilization and reduces energy costs. Using Oracle VM, ITprofessionals can create preconfigured virtual machine images for quick deployment,


6/9

6

and then quickly perform live migration to other servers to maintain high levels ofavailability.

KEEPING IT OPERATIONS AGILE

17. Managing a grid computing system requires a new breed of agile, efficient

systems management software that can provide rapid resource provisioning, real-timevisibility into end-user service levels, proactive monitoring, and diagnostics. As theadministrative workload associated with grid computing grows and evolves, the newsystems management software must be able to automate administrative tasks so that ITstaff can manage the growing complexity. In addition, the systems managementsoftware must be able to work within the underlying complexity of hardware andsoftware infrastructures, and be able to configure and modify those infrastructures tomeet the dynamic needs of the business.

18. Manual management of grid computing infrastructures is not economicallyfeasible. However, the comprehensive functionality and grid automation provides top-down, end-to-end application management with broad coverage across databases,

middleware, and applications. WebLogic Operations Control automates and optimizesthe dynamic allocation of resources across the application grid. It also ensures thatindividual applications have the memory, storage, and computing resources they needto meet their SLAs. Automated provisioning of database and middleware servers is dueto the server cloning or bare metal provisioning capabilities. By cloning servers, ITdepartments can create reference copies of test, development, and production serversincluding all patches, installed applications, and configuration data and then rapidlydeploy them to new server machines. Grid computing relies on agile, efficient systemsmanagement that includes rapid resource provisioning, proactive monitoring, andautomation of administrative tasks. The utility provisioned RAC cluster nodes to serveproduction, development, and test servers, and achieved the following results.

(a) Significant savings in hardware costs due to the consolidated datawarehouse.(b) Tremendous performance increases in business intelligence applications.(c) Low cost and high availability in remote locations.(d) A stable, scalable environment for both transactional and businessintelligence applications.

DELIVERING PREDICTABLE HIGH PERFORMANCE AND SCALABILITY

19. Recent trends toward more customer-facing Web applications, cloud computing,and service oriented architecture (SOA) are driving the need for predictable highperformance and scalability.

(a) Customer-facing Web applications have significantly more users thaninternal applications and more-challenging requirements for response time,scalability, and availability.

(b) Cloud computing service providers often have service-level commitmentsfor performance and availability. They must be able to scale quickly as newclients are added and workload usage peaks.


7/9

7

(c) SOA enables disparate applications and data sources to be integrated intoloosely coupled, composite applications. As more and more applications areexposed as Web services by SOA, other programs are quickly consuming thoseservices and creating greater, often unpredictable workloads. This is especiallytrue if Web services are exposed outside the organizations firewall. Such

exposure can put a strain on Web services and create spikes in response timeand bottlenecks in throughput.

20. Grid computing leverages clustering and virtualization technologies at all layersmiddleware, database and storage to deliver predictable high performance andscalability for applications. Enterprise grid computing delivers consistent highperformance as workloads increase from initial rollout to full-scale deployment. Inaddition, grid architecture provides the ability to scale all levels of the technology stackby using clustering and virtualization technologies to add storage, network, andcomputing capacity as they are needed. Capacity can be added in smaller, lessexpensive increments at any time, so organizations can take advantage of lower pricingopportunities.

FLEXIBLE, COST-EFFICIENT APPLICATIONS

21. The Exadata Storage Server and Database Machine are revolutionary newproducts that provide a smart storage grid connected to a database grid with fast Bandconnections. These products combine hardware and optimized software, and offerunlimited scalability (by adding more physical units) and faster query performance thatis at least 10 times faster than traditional data storage warehouses.

IN-MEMORY DATA GRIDS

22. For continuous operation, predictable high performance, and agile scalability,enterprise applications need virtualization and clustering capabilities in the middle tier.

They also need in-memory data grids. Grid computing Coherence an in-memory datagrid application lets Java and .NET applications quickly access objects that are stored inmemory spanning multiple physical machines in the middle tier. Grid computingCoherence provides horizontal scalability, predictable high performance, and highavailability. Its architecture scales linearly as additional nodes are added (or removed)on the fly. Grid computing Coherence achieves high performance by storing data inmemory on middle-tier computers to avoid disk access and network delays. It ensureshigh availability by storing copies of the data on different servers in the data grid,thereby avoiding a single point of failure in case a node crashes or is taken offline formaintenance.

23. AbeBooks.com is the worlds largest online marketplace for books, listing more

than 110 million new, used, rare, and out-of-print books from 13,500 booksellers.AbeBooks.com needed to enhance its online customer checkout experience andprovide fast, reliable access to frequently used information such as publication inventorydata, while increasing Web site performance and scalability. To achieve this,AbeBooks.com implemented Grid computing Coherence across all six AbeBooksinternational Web sites, to support the companys growing customer base and extremeprocessing requirements. With Grid computing Coherence, AbeBooks reducedtransactional strain on system resources, seamlessly managed millions of daily updates


8/9

8

from its seller community as well as 30,000 customer orders, and accelerated bythreefold the delivery of information to the companys shopping basket application.

SERVICE-LEVEL MANAGEMENT

24. IT departments are under great pressure to deliver on their service-level

commitments and be accountable to the lines of businesses they serve within anorganization. It is critical that IT departments be able to quantify and validate theservices they provide as information technology becomes more tightly linked withbusiness operations. As part of an organizations utility computing strategy, gridcomputing depends on accurate service-level management to ensure that the physicalcapacity of the data center is providing adequate service levels to end users.

ENABLING CONTINUOUS AVAILABILITY

25. Grid computing continues to be at the forefront of developing high availabilityproducts and practices. From server failover with Grid computing RAC and Gridcomputing Application Server, it provides IT organizations with a comprehensiveportfolio of solutions to keep the data center and the business running smoothly.

SERVER FAILOVER

26. Server failover has been available for many years from both hardware andsoftware manufacturers. The protections offered by a successful failover of a server arecritical, and yet the hardware and software costs of setting up a standby server usedonly if disaster strikes can be prohibitive. With grid computing, standby resources canbe used as active resources, resulting in higher utilization and lower costs.

27. Another consideration of server failover is failover time. Some applications are socritical that the business cannot afford to be without them even for a few minutes, whileother applications can tolerate some interruption of operation. As a pioneer in server-

failover techniques, Grid computing offers automatic failover capabilities for severalserver types. IT departments simply remove the failed servers from service, repair orreplace them, and then add them back to the server grid. Automatic migration andfailover of services (or whole servers) along with load balancing, workloadmanagement, and overload protection ensure that mission-critical applications stay upand running.

DISASTER PROTECTION

28. Even clusters cannot survive a complete data center failure from naturaldisasters such as fires and floods. In these cases, failover to a remote location isrequired. An enterprise grid can be designed to encompass multiple locations,dynamically shifting workloads to other locations for the highest reliability. Oracle ActiveData Guard provides the ability to create up-to-date replicas of the production databasefor standby and disaster recovery. These replicas can also be used for resourceintensive, read-only operations such as queries, reporting, and backup. Fidelity NationalFinancial (FNF) is a leading provider of outsourced products and services, includingtechnology solutions, financial and insurance services, and claims management. EDocmanages more than 50 million documents and affects numerous systems that arecritical to FNFs service offerings.


9/9

9

29. FNF faced heightened uptime requirements for EDoc because its subscriberswere increasingly using online documents for faster real-estate transactions. FNF alsowanted to lower operating costs by reducing manual operations and promoting a moreefficient e-business model. The previous implementation of EDoc was continuallyrunning at full server capacity. FNF decided to migrate to Grid computing RAC so it

could scale the application to meet the companys evolving growth and quality-of-service requirements. FNF determined that an Grid computing RAC implementation ofEDoc would provide increased CPU horsepower, higher utilization of resource levels,and load-balancing capabilities. According to the FNF case study, collectively, theutilization of Grid computing high-availability features and Grid computing maximumavailability architecture best practices has enabled FNF to meet service-levelagreements at the lowest cost. The new implementation has proven to be morereliable by eliminating single points of failure, and more scalable by providing the abilityto add capacity on demand.

CONCLUSION

30. Grid computing first introduced enterprise grid computing in 2003. The state-of-the-art technologies and new database and middleware capabilities helped change theway IT departments operate. At the time, data center projects including serverconsolidation, SOA development, space and power optimization, and large-scaleimplementations of rack-mounted Linux servers seemed unrelated. We can now seethat taken together these techniques can be described as a grid computing approach todata center modernization. The subscriber examples in this article illustrate howbenefits derived from these techniques can be compounded. For example, server andstorage consolidation increases utilization levels, which allows IT departments to saveenergy, reduce systems management costs, and get a better return on their hardwareinvestments. The widespread adoption of open standards, IT resource virtualization, on-demand provisioning, highly automated systems management, and real-time monitoringhas created a new generation of data center best practices. A second generation of gridcomputing technology builds on its strong foundation of scalable, fault-tolerant databaseand middleware clusters, virtualized computing and storage resources, and highlyautomated, end-to-end systems management. Today, with more than 10,000 Oraclecustomers deploying some level of grid computing technology, Oracle continues to leadthe software industry in its commitment to grid computing solutions, products, andpractices.

2. Grid Computing

Documents

Transcript of 2. Grid Computing