White PapersApplication virtualization provides a solution to the disparate nature of most datacentres, creating a real-time infrastructure (RTI) that can adapt dynamically to changes in application end-user demand while still operating within the boundaries of predefined policies.
RTI yields an infrastructure in which every server is essentially a backup for every other server — with the ability to lend and borrow resources across platforms during periods of unanticipated need or peak demand.
The ChallengeIT must be run as a business. Innovation for its own sake and robust IT are beingreplaced by targeted innovation with the goal of competitive advantage andslimmer IT budgets that require operational efficiency and a clear cost-cutting orreturn on investment. End-users have expectations, usually expressed in theform of business policies and Service Level Agreements (SLAs). SLAs govern IT's relationship with the business. Application availabilityrequirements and business policies represent the standard to which IT ismeasured, making it possible to evaluate the IT department's performanceaccording to distinct metrics. Typically, IT focuses singularly on SLA compliance,regardless of the cost or workload it creates. This is especially the case in aheterogeneous environment, where disparate systems power mission-criticalapplications while offering no opportunity to share equipment or streamline andautomate data center operations and management.In a mixed architecture environment, a central technology solution is necessaryfor consolidated IT management and comprehensive compliance with SLAs andbusiness policies. Application virtualization provides a solution to the disparatenature of most data centers, creating a real-time infrastructure (RTI) that canadapt dynamically to changes in application end-user demand while still operatingwithin the boundaries of predefined policies. RTI yields an infrastructure in whichevery server is essentially a backup for every other server with the ability tolend and borrow resources across platforms during periods of unanticipated needor peak demand. SLA Compliance on a BudgetSLAs govern the IT environment in order to ensure that end-users havepredictable and orderly access to the applications on which they rely. Clearmetrics facilitate IT management and translate the performance of IT into clearbusiness impact. As SLAs drive IT management, operational inefficiency is oftenaccepted as necessary to comply with SLAs. As a result of the cost, this is not aviable long-term IT strategy. Instead, a provisioning and maintenance approachthat is economical and efficient while addressing the strategic goals of thebusiness should be pursued; especially as IT budgets are tightening almostuniversally. Current IT practices segment platforms, leading to unnecessarily complex SLAmanagement practices. Mixed architecture data centers hold siloed applicationenvironments with dedicated equipment for specific applications, which are over-provisioned to satisfy projected application demand in any situation. Best ofbreed software implementations, a steady flow of mergers and acquisitions (M&A)activity and delays in upgrading or migrating legacy technology result in aheterogeneous technology environment and prevents centralized performancereporting and SLA compliance. Meet Service Levels with a Real-Time InfrastructureCentralized Command and Control of Enterprise ApplicationsFabricServer focuses on simplifyingmanageability and deployment, increasingoperational efficiency and flexibility, whilereducing cost and complexity.Untitled DocumentMeet Service Levels with a Real-Time InfrastructurePage 2FabricServer provisions capacity on-demand, automating the process of SLAcompliance without requiring the need topurchase, deploy and maintain excessservers. Applications only use the capacitythat they need, making under-utilizedapplication servers available to otherapplications and making SLAmanagement easier and less expensiveacross the entire company. Planned over-provisioning for SLA compliance seeks to mitigate the risk ofperformance degradation or outage. While this may satisfy the objective ofoperating within SLAs, the strategy lacks long-term viability, as planned over-provisioning requires planned over-spending. SLA compliance requires excessexpenditure on hardware, software licenses (and maintenance) and ongoing systemadministrator time, despite the fact that the equipment remains dark through mostof its useful live.DataSynapse FabricServerTMDataSynapse FabricServer centralizes the command and control of applicationdeployment and execution by virtualizing applications and guaranteeing that capacityis available on demand. Policy-driven IT management requires a technology environment that is: (a) Capable of meeting the demands stated in policies and SLAs(b) Standardized, where performance goals are stated and measurable (c) Implemented to measure performance against stated SLAsFabricServer provides a platform for creating a real-time infrastructure (RTI), makingsystems on different platforms capable of providing processors and memory acrossthe infrastructure to help other applications perform through periods of peakutilization and unanticipated demand. Additionally, FabricServer generatesperformance data that allow system administrators to monitor applicationperformance relative to SLAs and identify opportunities for increased efficiency. FabricServer stores application servers, runtime environments, and configurations ina component called the FabricBroker. FabricBroker is the heart of FabricServer,dynamically disseminating applications across the server farm to run in a flexibleenvironment based on specific roles and priorities defined by the business. The ITteam creates rules and policies in FabricBroker, and FabricServer executes them inreal-time and at runtime. Enterprise, custom, and composite applications respond tounexpected spikes in demand or degraded platform performance by reassigningresources swiftly and without manual intervention. The dynamic IT environment enabled by FabricServer collects, and stores anextensive set of metrics in the FabricServer reporting database to help you manageyour systems proactively. Administrators can monitor system performance relativeto IT policies and existing SLAs by integrating FabricServer with enterprise consoletools such as IBM Tivoli, CA Unicenter and HP OpenView to provide a comprehensiveand consolidated infrastructure reporting and analysis environment. The flexiblereporting engine in FabricServer can generate reports based on specific policies tomonitor IT performance and identify opportunities for improvement. Untitled DocumentMeet Service Levels with a Real-Time InfrastructurePage 3Comply Using FabricServerWhen in doubt, add more capacity? This has been the mantra of IT departmentsover the past decade, especially with the advent of "high availability" (HA)configurations. IT ensures that service levels are met simply by adding servers to anapplication environment in anticipation of utilization rates that the end-users areunlikely to need. Provisioned for "just in case" scenarios, these resources oftenremain under-utilized for most of their useful lives, generating significant expenses inexchange for little value. The initial cost of provisioning extra capacity is exacerbatedby ongoing maintenance fees and internal IT management costs that result from thedeployment of application capacity regardless of its utilization. FabricServer provisions capacity on-demand, automating the process of SLAcompliance without requiring the need to purchase, deploy and maintain excessservers. FabricServer balances workloads across the data center, rather thanconfiguring each application environment to balance workloads on its own (and withmost of the application capacity remaining dark). Applications only use the capacitythat they need, making under-utilized application servers available to otherapplications and making SLA management easier and less expensive across theentire company. FabricServer replaces the tangible costs of over-provisioning whileproviding a platform for meeting and monitoring compliance with SLAs andaddressing end-user application demand in real-time and without systemadministrator intervention. Even in a heterogeneous IT environment FabricServerkeeps the business running. About DataSynapseDataSynapse, Inc. is a global provider of application virtualization software. Thecompany's flagship products, GridServer and FabricServer", virtualize business-critical applications and adaptively provision them across a real-time infrastructure.DataSynapse drives business agility through shared services, helping clients reducethe cost and complexity of their IT infrastructure. The company is headquartered inNew York City.
