IT Operations at MicrosoftPublishedPublished: : May 2004May 2004AgendaIT MissionFacts and FiguresCentralized IT ModelOperations Service OfferingsStandardizationManagement ToolsLessons LearnedIT Mission PrioritiesBe the first and best Microsoft customerProvide thought leadershipSet a coordinated IT strategyRun a world class utilityE-mail 5-7 M e-mail messages per day internally 8.5M inbound mails/day 7.5M inbound mails filtered / dayMicrosoft IT EnvironmentDevices 200K PCs 8K data center servers 275K network hosts 10K network devicesIncident Mgmt 90K help desk calls/month 7K infrastructure Service Requests/month 6K changes/monthMonthly Remote Access 45K RAS 49K OWA 18K RPC over httpApps Single Instance SAP 1,371LOB appsPhones 11M campus calls/month 6.3 M PSS calls / monthLocations 317 sites 25% Internet only connectUsers 50K Employees 5K Contractors 17K Vendors317 WAN sites in 65 317 WAN sites in 65 countriescountriesMicrosoft Network OverviewCharlotteCharlotteHub and spoke Hub and spoke topologytopologyMore than 8,000 serversMore than 8,000 serversSydneySydneyChofu & OtemachiChofu & OtemachiLes UlisLes UlisTVPTVPDublinDublinBeneluxBeneluxMadridMadridDubaiDubaiSingaporeSingaporeJohannesburgJohannesburgSao PauloSao PauloCanyon ParkCanyon Park RedmondRedmondLos ColinasLos ColinasChicagoChicago MilanMilanStockholmStockholmMunichMunichSilicon ValleySilicon ValleyCISCO for production CISCO for production networknetworkIT OrganizationCentralized IT operationsReduced operation costs38% lower than distributed Gartner1Required for effective standardizationRegional representationAccount managers are the “face” of IT in regionsProvides understanding of local business and cultural issuesAllows for local accountability and involvement with the business Links IT service levels directly to localized business driversWorld wide service offering definitionsServer Offering 1: Power, cooling, and physical securityServer Offering 2: Add best effort (reactive) supportServer Offering 3: Add full service supportKey guidelinesHigher compliance with standards as you go upSpecific software patches/upgrades not optionalClarify roles, responsibilities, and expectations between IT and CustomerOperations Service OfferingsStandardizationAchieving standardizationService level determines IT support based on degree of standardizationAllow most anything, but provide top support for those who run standard hardware & software platformModel makes it more expensive to be non-standard as business units can not match our efficiencies/economies of scaleSecurity patches are not optionalKey to efficiency and effectivenessReduced complexity, support, and training requirementsReduced mean-time to resolve (MTTR) through greater experience/expertise with standard systems/platformsLeverage CentralizationMonitoring spaceGNOC, DCOPS, GTOCTools consoleGlobal view and correlationProcess and escalation templatesShared resources and best practicesTechNetIssue ResolutionDrive to solve issues remotelyTier 3 escalation group/infrastructure serviceService management group/infrastructure serviceClear line between application/infrastructure supportToolsMicrosoft Operations Manager as top tier consoleAll tools instrumented for Windows NT eventsAuto ticket generation based on alertsBusiness Prioritization of Issues Measure, track, and improve key metricsVolume (overall activity)Latency (96% of alerts at console in less than two minutes)Alert to ticket ratio (from 35:1 to 3:1 in 2 years)Emphasis on remote resolution of software issuesLessons LearnedChargeback modelPartial cost recovery direct cost only to influence behaviorIncludes datacenter, network, backup, service, and support costsUnit prices do not vary in a fiscal yearDoes not undermine budgeting effortsCustomers live with their volume variancesEmpower customers to calculate cost impact of decisionsIncrease visibility and accountability for IT costsLessons LearnedFacilities competency as a requirement for high availabilityIf the design availability of the facility is not known, neither is that of the serviceLessons LearnedEffective Worldwide IT includesCentralized infrastructureRegional IT representationStandards for hardware, software, tools, and networkCentral strategy for distribution of key infrastructure assetsLeverage active directoryDatacenters and e-mailLessons LearnedDisaster recovery preparedness Prepare nowPlan for a substantial and lengthy effortCreate a global command team for major incident responseLessons LearnedExecutive supportSecurity decisionsConsolidation strategiesPrioritization decisions into budget reality Metrics, metrics, and metricsFocus on what to measure and whyShared goalsFor More InformationAdditional content on Microsoft IT deployments and best practices can be found on http:/www.microsoft.comMicrosoft TechNet http