2Platforms Differentiated by Superior Availability Stratus ConfidentialPlatforms Differentiated by Superior AvailabilityIndustry-standard ftServer® platform>99.999% availability across fleetChoice of hardened operating systemAutomatic availability; no application modificationRed Hat® Enterprise Linux®Red Hat Certified Hardware CatalogVMware® Infrastructure 3Global Alliance PartnerMicrosoft® Windows Server™ 2003Microsoft Hardware Compatibility ListThe Stratus ftServer family is based on industry-standard Intel processors and a choice of the Microsoft Windows 2003 operating system or Red Hat Enterprise Linux 4, along with many other standard hardware and software components. ftServer systems are listed on the Microsoft Hardware Compatibility List and Red Hat Certified Hardware Catalog, ensuring that all standard Windows and certified Red Hat Linux software will run flawlessly on all ftServer models.
3Technology Partners Joint program to insure processor determinism Stratus ConfidentialTechnology PartnersJoint program to insure processor determinismEarly access to new processorsftServer® Series systems in Intel labsCollaborate on OS availability featuresHardened driversMemory resynchronizationSupport for mission-critical applicationsCollaborative support modelftServer systems in partner test labsReseller of CLARiiON® storage systemsftServer Series systems in EMC Qualification labCollaborative support model
4Continuous Availability – 99.999% Stratus ConfidentialContinuous Availability – %Track record of successTrusted availability partner for over 27 years to enterprises, government agencies and telecommunications service providers worldwidePioneer in continuous availability solutionsClients in 70 countries on 6 continentsGlobal world-class expertiseSpecialized availability expertsUnrivaled troubleshooting expertise
5Stratus® ftServer® Family Stratus ConfidentialStratus® ftServer® Family2500 Family4400 Family6200 FamilyDual-coreQuad-corePositioningHighly affordable, entry-level, fault-tolerant serverVersatile, expandable Price/performance leader for applications that must not failEnterprise-class server delivering maximum performance & availability for mission-critical computingDeployment ModelReplicated remotelocationsDepartmentalbusiness processingCritical enterprisebusiness/operations processingWorkloadStandaloneFixed configurationsStableSophisticated SMB environmentsGeneral purposeGrowing or unpredictableTransaction-intensiveDatabase serverServer consolidationThe current ftServer family consists of the entry-level 2500, the price/performance leading 4400 and the enterprise-class quad-core 6200.
6ftServer® Architecture: Lockstep Technology Stratus ConfidentialftServer® Architecture: Lockstep TechnologyDuplex hardware componentsFault Detection&IsolationEmbedded I/OPCIDisksCPUChipsetMemoryFault Detection&IsolationLockstepped CPUsMulti-path I/OFault Detection&IsolationEmbedded I/OPCIDisksCPUChipsetMemoryFault Detection&IsolationLockstep technology uses replicated, fault-tolerant hardware components that process the same instructions at the same time. In the event of a component malfunction, the partner component provides an active spare that continues normal operations and prevents system downtime. The system also eliminates transient hardware errors that, in-turn, could cause software failure.ftServer systems use a standard DMR (Dual Modular Redundancy) configuration which includes two CPU/memory units.The fault-tolerant I/O system is logically separated from the CPU-memory system. Hardware logic, in the form of custom ASICs, acts as a PCI bridge between the CPU and I/O, and also provides the core error detection, fault isolation and synchronization logic for the lockstep architecture. Fault-tolerant I/O is implemented using replicated PCI buses, replicated I/O adapters and replicated devices. Multiple paths are available to any logical I/O operation, including both internal and external storage operations. Any I/O operation failure will result in a retry using an alternate path ensuring successful completion of the I/O operation.Eliminates single points of failureDetects and isolates errorsPresents single software imageDelivers zero switchover time440025006200
7Stratus Lockstep Technology: Automatic Availability Stratus ConfidentialStratus Lockstep Technology: Automatic AvailabilityRedundant components execute same instructions in lockstepSelf-checking logic detects and isolates errors to the component level and notifies Stratus of problemsStratus systems are designed to handle hardware problems automatically, with no failover delay or data loss. Lockstep technology uses replicated fault-tolerant hardware components that process the same instructions at the same time. Hardware logic, in the form of Stratus chipsets, provides the core error detection, fault isolation, and synchronization logic for the lockstep architecture.Errors are detected and isolated to the component level and the system itself then notifies Stratus of any problems so that corrective action can be initiated immediately.Stratus systems are also built to withstand unpredictable transient errors that can lead to system downtime or cause silent data corruption on conventional systems and clusters.In the event of a component malfunction, the redundant component acts as an active spare that continues normal operation, thus averting system downtime.While other servers may offer duplicated power supplies, fans, and disk drives, only Stratus provides protection for core system components including motherboards, processors, memory, I/O buses, and I/O adapters.In-memory data is protected from hardware failures by the ftServer lockstep architecture which maintains at least two copies of all memory data.System distinguishes between transient and hard errors; withstands transient errorsIf hard error occurs, faulty component is taken out of service; duplicate component continues normal processingIn-memory data and disk data are protected
8Failsafe ftServer® System Software Stratus ConfidentialFailsafe ftServer® System SoftwarePrevents, diagnoses, and resolvessoftware issuesHardened drivers prevent software failuresftServer hardware and software combine to shield operating system, middleware and application from the impact of transient errorsSoftware issues are reliably captured, analyzed and correctedExtensive integration / error insertion testingIn-memory data and disk are protectedFailsafe software runs concurrently within the system to prevent many software errors and to capture, analyze, and correct other software issues that might disrupt processing. Even in-memory data is constantly protected and maintained.Unlike standard servers or clusters, ftServer hardware and software handles many errors transparently, shielding errors from the operating system, middleware or application software, thus eliminating many potential software problems.Stratus-supported device drivers include software extensions for full hot-plug operation, transparent failover, and added diagnostic and service capabilities, along with additional integration testing and reliability improvements. Stratus hardened drivers add considerable reliability to the Windows environment on ftServer systems.Stratus uses a rigorous test process that targets fully-integrated systems including all hardware and software options in a variety of configurations including maximum configurations. Systems are tested under extreme processing and I/O loads and errors are continuously simulated during the test process. Finding and resolving these integration and error-insertion test issues results in a higher level of software reliability for ftServer systems.ftServer management and diagnostic features ensure that software issues are reliably captured. With Stratus Assured Availability Plus Service, a user can count on Stratus to take responsibility for diagnosis and resolution of any system crash, including issues within the Window operating system.In-memory data is protected from hardware failures by the ftServer lockstep architecture which maintains at least two copies of all memory data.
9Stratus Failsafe Software: Automatic Availability Stratus ConfidentialStratus Failsafe Software: Automatic AvailabilityReliably distinguishes software issues from hardware issues to speed problem resolutionDriver hardening ensures uninterrupted processing if an I/O interface failsRapid disk resynchronization transparently mirrors disks and protects data integrityAccurate error data capture enables root-cause analysis to prevent problem recurrenceManagement and diagnostic features also capture, analyze and notify Stratus of software issues.Unlike standard servers or clusters, ftServer Series hardware and software combine to handle many errors transparently, shielding the operating system, middleware and application software, thus eliminating many potential software problems.Stratus-supported device drivers include software extensions for full hot-plug operation, transparent failover, and added diagnostic and service capabilities, along with additional integration testing and reliability improvements. Stratus hardened drivers add considerable reliability to the operating system on ftServer Series systems.Stratus uses a rigorous test process that targets fully-integrated systems including all hardware and software options in a variety of configurations including maximum configurations. Systems are tested under extreme processing and I/O loads and errors are continuously simulated during the test process. Finding and resolving these integration and error-insertion test issues results in a higher level of software reliability for ftServer systems.RDR continually monitors the integrity of the RDR-enabled disks, scouting for bad blocks. RDR then fixes bad disk blocks, and copies the correct data from the partner. This online bad block correction maintains data integrity and system availability, key components to the mission-critical functions of Stratus’ ftServer systems.ftServer Series management and diagnostic features ensure that software issues are reliably captured. With Stratus Assured Availability Plus Service, a user can count on Stratus to take responsibility for diagnosis and resolution of any system crash, including issues within the operating systems.
10ActiveService™ Architecture: Automatic Availability Stratus ConfidentialActiveService™ Architecture: Automatic AvailabilityComponent fails. System ISOLATES fault and notifies Stratus that a CPU has failedSystem automatically ordersCORRECT replacement partNext DayDelivery ServiceHot-pluggable components are EASY to replaceThe ActiveService Architecture has its foundation in the design of the hardware and its technology-enabled Access features.The system's self-monitoring and diagnostics features are capable of correctly identifying the specific hardware or software component that causes an abnormal condition. The elimination of guesswork and trial-and error-part swapping clearly sets ftServer systems apart from alternative high-availability solutions.In the event of component failure, the system automatically opens a call with the Stratus Customer Assistance Center that supplies information on exactly what action is needed.Most replacement components arrive within 24 hours. .The hot-pluggable components are easy to replace by users without special tools, commands or a visit from a field engineer. Once in place, the new component is automatically integrated into the running system.Throughout this entire process, you system and application will continue to run normally.System synchronizes with replacement componentNormal processing continues throughout process!
11ftServer® Family User Benefits Stratus ConfidentialftServer® Family User BenefitsContinuousAvailabilityOperational SimplicityFinancial Advantage> % uptime for a suite of enterprise Operating SystemsOperating System availability enhancementsSpace-efficient fault-tolerant designQuick to deploySimple to operateEasy to maintainTechnology-enabled servicesLow TCOCuts cost of downtimeMinimizes business disruptionBetter resource utilizationWith the ftServer family, Status provides a practical and affordable way to achieve the highest levels of availability for Microsoft Windows and Red Hat Enterprise Linux environments.Our ftServer line delivers Continuous Processing through Stratus’ lockstep technology, failsafe software, and the ActiveService Architecture. These key features automatically work together to prevent downtime in the easiest way possible.Our newest generation of systems expands on these inherent strengths with superior price-performance, greater space efficiency, and simpler serviceability.Stratus' Smarter Approach to Uptime also rewards users with a tangible financial advantage over competing approaches. This superior technology reduces the costs associated with complicated deployment, unplanned downtime, and ongoing support expenses.
12Operational Simplicity Stratus ConfidentialOperational SimplicityDesigned for the most demanding local, remote, and lights-out computing environmentsStratus industry-standardplatforms are as easy to install, use, and maintain as conventional serversDay in and day out, organizations deal with the reality that neither their in-house IT staff nor vendor service professionals will be close at hand to every site or able to respond quickly to every situation.This includes centralized computing operations where IT resources are becoming more and more constrained, remote sites where there may not be immediate access to engineers or parts, as well as lights out environments where IT resources may be non-existent.We respond to these needs with systems that have proactive remote service capabilities and are easy to install, use and maintain.
13Stratus ConfidentialEasy to InstallRapid deployment: applications typically up and running in less than a daySingle system; single software licensesCustomer Response“I like the Stratus ftServer system because of its simplicity. With clustering, the typical approach to high availability, comes other issues including complex configuration and management, as well as software compatibility.” Vaughn Book, Senior Vice President and Chief Technology Officer Arrowhead Credit UnionBecause fault-tolerant ftServer systems supply uptime protection automatically, without added effort, they eliminate the time consuming and people-intensive work needed to implement high-availability clusters. Applications running on an ftServer system are unaware of any unique hardware or software elements of the ftServer architecture.Typically, an ftServer Series system can be installed - with the application loaded and ready for operation in less than a day.By contrast, applications should be modified to use cluster APIs in order to operate optimally in a cluster environment.
14 Easy to Use Streamlines day-to-day operations Customer Response Stratus ConfidentialEasy to UseStreamlines day-to-day operationsMinimal operator intervention requiredNo need to think about failure scenariosRemote monitoring, physical indicators; and hot-pluggable components simplify serviceabilityReduces workload of IT support staff; enables redeployment of scarce resourcesCustomer Response“One of the reasons for a change to the ftServer system was to increase operational simplicity. Considering the limited size of our IT team, the easy management of the new server will also allow us to save a lot of costly human effort." Diego Bertazzi, Operations Manager Information Technology Group, Casino de la ValleeftServer systems effectively streamline your day-to-day operations.We've significantly eliminated operator intervention for routine management or when a hardware failure occurs because the system is designed to handle and report problems without disrupting your normal computing operations. And, there is no need to even think about failure scenarios or rewrite scripts which can change any time you make changes in your computing environment.A problem on other systems also means your staff must spend precious time finding the problem first and then figure out the best way to fix it. Our remote monitoring and 24 x 7 diagnostics means we often know about potential problems before you do! Customers often tell us that the first time they are aware of any issues is when they receive a call from us telling them exactly what problem has occurred along with the recommended remedial action.When necessary, our service personnel can also dial into your Stratus system, regardless of where it is located, and take corrective action (even if it is offline). In fact, more than 95% of the time issues are resolved remotely.Should a replacement part be necessary, the system's colored status LEDs indicate exactly what component is affected. Our components are designed to be customer-replaceable.
15 Easy to Maintain Integrates with standard management tools, Stratus ConfidentialEasy to MaintainIntegrates with standard management tools,existing staff skills, operational proceduresManaged as a single, industry-standard serverCustomer Response“Because the ftServer system runs in a Windows environment, we are able to leverage technical expertise that's plentiful throughout our staff. This is a big change from the special knowledge needed to work with the ATM/EFT controller that we replaced." Gomer Jones, Senior Vice President Texas State BankftServer Series systems are administered as single systems. Your organization can use standard management tools and procedures.And, you save even more on operational costs because you can leverage existing skills and staff training.In contract, managing clusters involves special skills as well as additional configuration and management procedures beyond those needed to manage standard servers.
16Immediate Application Availability Stratus ConfidentialImmediate Application AvailabilityApplication transparencyApplications load and goInstant fault-toleranceCustomer Response“Standard Windows server-based applications run without a problem, operating transparently on the fault-tolerant platform."The Stratus approach to applications availability is both automatic and transparent.The ftServer Series family meets qualification for the Microsoft Hardware Compatibility List (HCL), VMware HCL Systems Compatibility Guide, and Red Hat Certified Hardware Catalog. This is your assurance that standard compatible software will run flawlessly on ftServer Series models without ANY modification.You simply power up the server and load your application to benefit from our Continuous Processing technology.This immediate access to application availability is another clear advantage that Stratus servers offer compared with clusters, which recommend (and in some cases require) that applications be modified for cluster operation. In addition, the loss of memory and application state during a cluster failover often dictates design changes within the application to prevent unacceptable data loss.
17Remote Problem Resolution that Goes Beyond the Hardware Stratus ConfidentialRemote Problem Resolution that Goes Beyond the HardwareWhen You call ...... an availability expert answers.The types of problems we resolve dailyOperating systemNetworkDatabaseMulti-vendor issuesHardwareApplicationPatchesSecurityHot fixesStratus ActiveService Network is a secure, fault-tolerant worldwide network that enables instantaneous, around-the-clock monitoring and troubleshooting of systems regardless of physical locations.The Stratus Customer Assistance Center (CAC) provides the critical support necessary to ensure continuous operation. At Stratus 24x7 support is the norm, not the exception. And whenever you call, an availability expert is there to help you immediately.The server's remote support capabilities – made possible by Stratus out-of-band management technology and our global ActiveService Network – also enable service engineers to diagnose, troubleshoot, and resolve problems online as if they were on-site.Running on independent power and network connections, the out-of-band management device allows remote communication to take place between Stratus’ ActiveService Network and the customer’s module — regardless of the system’s state. This allows a service engineer to remotely power on/off or reset/ reboot the system and manages the security of incoming and outgoing communications through the ActiveService Network.But, we don't limit ourselves to solving Stratus server problems. Our customers consistently credit us with understanding availability at every level -- from the server to the operating system to the network to the ongoing managed services needed to maintain optimal performance.Customer Response“You need immediate help when you have a 911 systems issue, and Stratus understands that. Your call doesn't go into a phone tree. You speak to a problem-solver right away." Douglas Bolton, Information Systems AnalystSan Diego Fire & Life Safety Services
18Select the Support Offering that Matches your requirements Stratus ConfidentialSelect the Support Offering that Matches your requirementsServiceTierOfferingYour Business EnvironmentStratus Value AddAssuredAvailabilityPlusExtremely risk adverseHigh cost of downtimeMission-critical applicationsRequire highest level of 24/7 supportImmediate response from an expertSingle point of contactPriority issue escalationRoot-cause analysis and permanent fixAvailability partnershipPremium operating system servicesAvailability program guaranteeSystemTraditional service levels are adequateCompetitive with traditional vendors break/fix servicesPremiumStratus offers two tiers of service offerings Premium” and “Basic”Premium support service customer business concerns include:• Extremely risk adverse• High cost of downtime• Mission-critical applications• Require the highest level of 24/7 supportPremium Value-added service components include:An immediate response from an expertSingle point of contact when they have a serious/critical problemPriority management issue escalationRoot cause and permanent fixManage OS kernel issues directly with Microsoft or Red HatAvailability partnership (take ownership even when vendor ownership is not immediately clearAssured Availability Program GuaranteeBasic maintenance customer business concerns include:• Initial acquisition cost• Traditional service levels are adequateBasic maintenance customer service components include:Be competitive with traditional vendor break/fix servicesBasic
19Availability Alternatives Stratus ConfidentialAvailability AlternativesStandalone Systemssome duplicate componentsClustered SystemsftServer SystemsReduce failureRecover from failurePrevent failureCost of downtime negates advantage of low purchase priceUnpredictable availability - varies by vendor / modelWhen it fails, it fails hardSuccess depends on correct planning and implementationSatisfactory operation highly dependent on skilled personnelLockstep technology, failsafe software, and ActiveService™ architecture provide continuous availabilityNo special skills required to implement and manageSome of today's servers offer subsets of availability features in the form of duplicated components but these vary greatly by vendor and model. The notion is to reduce the probability of failure - but when a standalone server fails, it fails. Such servers aren't designed to prevent downtime or data loss, nor are they able to recover quickly from problems.Clusters are built by assembling hardware and software from multiple vendors. Physical connections and software programs link two or more servers so that when a failure occurs on one, its workload fails over to a server that is still operating. These configurations are described as high availability because they speed recovery from problems. They are not, however, designed to eliminate downtime altogether. They are also complex to build and maintain and require highly skilled staff throughout the lifecycle of the system. Whenever changes are made to hardware or software, its failover capabilities (software scripts) must be updated and retested to make sure that failover will work as planned.Stratus fault-tolerant ftServer systems provide continuous uptime because they are designed to prevent downtime from ever occurring. Normal processing continues without transaction or data loss should component failures or software faults that generally cause other servers to fail occurs.These systems give you the best of all worlds. Operationally they are comparable to a standalone system. They protect applications as well or better than clusters and standalone systems and from a technology availability standpoint they are superior to either alternative.We’ve graphically displayed how each effective each alternative is at addressing Gartner Group’s mix of the causes of unplanned downtime. We believe that the ftServer scores highest in all three categories for the following reasons::-Technology Failures. The combination of lockstepped CPUs, redundant components, ability to detect and ride thru transient errors, Fail Safe software, and the insertion detection process used in the development of the system delivers the.-2- Application Failure. Many failures blamed on the OS or application are actually transient hardware problems that the ftServer will ride thru. Also, OS and applications are deployed with no modification further decreasing the likelihood of problems.-3- Operations Errors. The complexity of clusters demands expert-level administration and management. Conversely, the ftServer can be managed without specialized training.OperationsApplicationTechnologyEffectiveness at addressing downtime
20Value Comparison Features ftServer® System >99.999% Stratus ConfidentialValue ComparisonFeaturesftServer® System >99.999%>Clusters 99.9%SystemArchitectureLockstep TechnologyZero failover timeFailure preventionMinutes of failoverFailure recoveryResourceProtectionFailsafe SoftwareProtects memory & disk dataShields SW from transient HW errorsProtects ONLY disk dataHW errors cause OS and application failuresService technologyActiveService™ ArchitectureFault detection and isolation Hot-swappable componentsBuilt-in serviceabilityLimited self diagnosticsLimited hot-swappabilityAdd-on serviceabilityDeploymentImmediateScript development/testingApplication ModificationNoneRecommendedOperationSingle system imageLights out operationMultiple system imagesRequires IS SupportContinuousAvailabilityOperationalSimplicityFinancial AdvantageThe recovery time comparison assumes hardware failure. For a software failure, both ftServer and a cluster will recover in minutes (ftServer reboots while a cluster fails over).A cluster failover always loses all memory contents, while ftServer preserves all memory contents in the event of a hardware failover, and preserves data in the event of a system software failure. Applications can take advantage of memory persistence to both improve performance and reduce complexity.It is recommended that applications be modified to work best in a cluster environment, although this is not an absolute requirement.
21Ease of Implementation: Availability Alternatives Comparison Stratus ConfidentialEase of Implementation: Availability Alternatives ComparisonImplementation ProcessStandalonesome duplicate componentsClusteredsystemsftServersystemPlanning and designNoneExtensiveAddresses single points of failureExtra work requiredYes; it's built-inSizing / load balancingNodes must be properly sized for adequate performanceApplication modificationHighly recommended to ensure failover, SW compatibility, data integrityDevelop and test failover scriptsNoYes; must repeat with each changeTest for effectivenessYesDocument proceduresExtensive documentationThe Stratus ftServer and competitive standalone servers compare almost identically when it comes to ease of implementation. Essentially, it’s “load and go”. The main difference of course, is that the level of availability (99.999% and greater) delivered by ftServer systems.Ease of implementation has been, and continues to be one of the major drawbacks to high availability clusters. Customers who choose to implement a cluster should be aware of the amount of effort and time it takes to properly plan and test the solution before implementation can occur. Even after the cluster is deployed, the work isn’t over. In fact, it never ends because every time a change is made to the configuration, the cluster needs to be tested to insure proper failover will occur.
22Problem Handling Effectiveness: Availability Alternatives Comparison Stratus ConfidentialProblem Handling Effectiveness: Availability Alternatives ComparisonProblem Handling FeaturesStandalonesome duplicate componentsClusteredsystemsftServersystemRecovery timefrom failureMinutes to daysMinutes to hoursZeroIn-flight dataAll lost on failureAll lost on failoverNo data lossFault detection and isolationLimitedYes, to the component levelTransient errorsMay cause failure; lead to silent data corruptionWithstands transient errors; no data corruptionHard errorsVaries by HW configuration; possible failureSystem failoverSystem continues normal processingIntegrity of database transactionsVaries by HW configurationYesDriver hardeningNoWhile the availability characteristics of standalone servers are – as a rule –increasing, they are still extremely vulnerable to catastrophic hardware crashes. Depending upon configuration, failures to components including processors, memory, disks and their controllers, voltage regulators, power supplies, and fans can all result in an outage. And, depending on the type of failure, a standalone server can be offline until the service provider arrives hours or even days later (hopefully with the right replacement part).High availability clusters – if implemented and managed correctly – will deliver a higher level of availability than a standalone server. (Think of this increase in availability as a reward for all of the effort that goes into the set up and care-taking of the cluster).However, even with expert setup and management, the best a cluster can hope for is “speedy” recovery from a failure. When a failure does occur, any in-memory data that was being worked on is lost and the amount of time that the application is unavailable is highly dependent on such things as the time it takes to recognize that the failure has occurred, the time it takes to fail over to an available server and the time associated with database and application recovery.Both standalone servers and clusters are also susceptible to problems caused by transient errors and misbehaved drivers.
23Solutions Management: Availability Alternatives Comparison Stratus ConfidentialSolutions Management: Availability Alternatives ComparisonOn-going ManagementStandalonesome duplicate componentsClusteredsystemsftServersystemServer platformSingle system view;no special skillsComplex; requires specialized personnelSingle system view; no special skillsFailover testing when changes are madeNoRequiredSystem monitoring, reporting, diagnosticsVaries by vendor and platformIntegrated 24 x 7 system monitoring, status reporting tied to ActiveService NetworkRemote problem identification / repairLimited. Varies by vendor and platformYes: even when server is offlineMemory data capture in event of OS outageYes, default feature; does not delay application restartChange managementChanges easily trackedComplex documentationA particular strength of the ftServer is the ease with which the system can be managed on a daily basis and repaired when a problem does occur. The ftServer appears to the administrator as a single system and requires no special testing or change management procedures thereby greatly reducing the likelihood of administrative errors.Because the ftServer’s ActiveService Architecture is engineered into the systems’ design – not added on as an afterthought –problems are instantly identified and reported and problem resolution can commence immediately. While competitors offer “phone home” technology to identify certain classes of problems, none can match Stratus’ ability to diagnose, root-cause, and remedy most problems remotely.Stratus value-add features include the ability to capture the contents of memory after a failure occurs while simultaneously bringing the system back online. Competitive servers are faced with a choice of either bringing the system back online or delaying the system restart in order to harvest the contents of memory in order to analyze the cause of the failure.
24Managing Planned Downtime (Windows® only) Stratus ConfidentialManaging Planned Downtime (Windows® only)ChallengesAvailabilityriskEase of useMinimize downtimeClustersProne to synchronization and recovery issues_Cluster expertise requiredMinimal (1 to 3 short duration) outagesStratus Active Upgrade™Automatic synchronization between production and updateSystem can be restored to its original state at any timeAutomatic pre-upgrade readiness checkSingle OS and Single ApplicationIntuitive user-interface guides you through every stepNo Application modificationSingle Application outageUpgrade completed in secondsPrior to Active Upgrade, the only option for minimizing planned downtime was cluster rolling upgrades, where one node of the cluster is updated while all users are connected to the other node. While this does accomplish the task of minimizing downtime, it is challenging to insure both systems are the correct configurations, and perhaps more daunting is the recovery required if an issue is encountered during the update. This often requires multiple restores and application check pointing before starting all over again .In contrast, Active Upgrade automatically insures that production and update versions are the same. In addition, if issues are encountered during the update, the system may be returned to its original state by the press of a button with no downtime at any step of the upgrade.From a simplicity standpoint, Active Upgrade requires only a single upgrade to the Operating or application, since there is only one copy of the ftServer, Operating System, and application. Unlike clusters, which may require application modifications to make them cluster aware so that they operate properly in a rolling upgrade, applications may operate on a system with the Operating System or application being upgraded by Active Upgrade with no modification whatsoever.Last but not least, Active Upgrade requires only a single application outage while clusters may require as many as 3 depending on the cluster type (Active-Active or Active-Passive).
25Cluster Customer Experience Stratus ConfidentialCluster Customer ExperienceImplementation ProcessFailure RecoveryProcessRequiresspecialized skills:PlanningDesigningBuildingScriptingConfiguringMeasuringTestingProcess must repeated with each change!Notify service providerWait for arrivalProblem analysisReplace failed componentReboot OSRecover file systemRecover databaseRestart network softwareRestart applicationTime-consuming, labor-intensive
26Stratus Customer Experience Stratus ConfidentialStratus Customer ExperienceAutomatic AvailabilityImplementation ProcessFailure PreventionStratus Continuous Processing® technology prevents failures from occurring24 x 7 monitoring and reporting integrated into every systemRoot-cause analysis to prevent problem recurranceLoad-and-go deploymentMaximizes existing IT skills
27ftServer® Product Specifications Stratus ConfidentialftServer® Product SpecificationsftServer 2500 systemftServer systemftServer 6200systemProcessors1-socket1- and 2-socket2-socketCPU/FSBFrequencyDual-core 2 GHz CPU1333 MHz FSBQuad-core 2.66 GHz CPUCPU Cache4MB L24MB L22 x 4MB L2Memory2GB–6GB DDR667 FBDIMM2GB–12GB DDR667 FBDIMM4GB–24GB DDR667 FBDIMMPCI Slots6 x PCI-X, or 4 x PCI-X and 2 x PCIeRAID 1 Disk3 Enterprise SAS or SATA DisksEmbeddedSAS, 2x ENET (10/100/1000) ,3x USB2.0, 2x COMM, VGA, BMC, R/W DVDOperating SystemsWindows® Server 2003 Standard Edition or Red Hat® Enterprise Linux®Windows® Server 2003 Enterprise Edition VMware ® Virtual Infrastructure 3Red Hat® Enterprise Linux®Form Factor4U Rack and Pedestal4U RackNote: All I/O options specified in logical count, physical count is 2 times numbers specified
28Building Blocks of the ftServer® Family Stratus ConfidentialBuilding Blocks of the ftServer® Family250044006200CPU / IO CRUBlindmate BackplaneThe ftServer family is based upon modular building blocks which are combined to build a fault-tolerant platform. The CPU/IO are rack-optimized CRU’s that are paired for fault-tolerance and inserted in a blindmate backplane.
29ftServer® Modular Family Components Stratus ConfidentialftServer® Modular Family ComponentsCPU/IO2U Blindmate Enclosure, front removable1 or 2-socket CPU6 DIMM Slots1333 MHz Front Side Bus4 GHz Memory BusUp to 3 Hot plug, SAS or SATA Drives, front removablePower On button, front accessible & protected6 Customer Configurable PCI-X/PCI-Express slots, rear accessible2 10/100/1000 Com port, rear accessibleThe CPU-I/O is the main a module in all ftServer systems. Each CPU-I/O contains the equivalent of a single server including disk, PCI slots, memory, and processor(s), as well as integrated USB, SCSI, Serial, and Ethernet in a 1U form-factor.The CPU-I/O contains a variety of LED’s that provide the status of the CPU-I/O status, disk activity and status, PCI slots status, system management status, and Ethernet status, activity, and speed
30ftServer® Modular Family Components Stratus ConfidentialftServer® Modular Family ComponentsPassiveBackplanePassive backplane for seamless component interconnectBlind mates power and signal to CRU’sStandard 19” EIA compatibleAdjustable rail – fits cabinet depth range 24” to 30”Shared connectors located in rearThe blindmate backplane is a single unit into which the CPU-I/O units plug. While configuring a typical cluster may require configuring a myriad of data, heartbeat, and power cables; the ftServer interconnect is achieved by simply sliding in the CRU into the VHDM (Very High Density Metric) connection on the backplane. The backplane provides power and signals (data, clock, system management) to each CRU. In addition to the clock card the backplane contains an IO-Card containing the ports for the embedded Ethernet, USB, Serial, and VGA.The fact that the backplane is passive versus active dramatically improves the MTBF.The backplane mounts in a standard 19” rack and its sliding rails support virtually any rack depth.
31ftServer® 2500 Systems Workloads and Attributes Specifications Stratus ConfidentialftServer® 2500 SystemsWorkloads and AttributesHighly affordable fault-toleranceEasy to deploy and simple to manageIdeal for replicated deployment at remote locationsEntry-level1-socket serverSpecifications4U rack-optimized form factorPedestal option1-way CPU2.0 GHz dual-core Intel® Xeon™ processor with 4MB cacheDual modular redundancySAS/SATA storage supporting 6 internal disksOptional external JBOD or RAID (Windows only)2GB – 6GB memory6 PCI-X slots, or 4 PCI-X and 2 PCI-Express slotsIntegrated 10/100/1000 EthernetWindows® 2003 Standard Edition orRed Hat Enterprise Linux
32ftServer® 4400 Systems Workloads and Attributes Specifications Stratus ConfidentialftServer® 4400 SystemsWorkloads and AttributesPrice / performance leaderDedicated business processing application serverIdeal for departmental business processing; ideal for space-constrained data centersMid-range1- or 2-socket systemSpecifications4U rack-optimized form factorPedestal option2.0 GHz dual-core Intel® Xeon™ processor with 4MB cacheDual modular redundancy2GB – 12GB memorySAS/SATA, storage supporting 6 internal disksOptional EMC (Windows/VMware only) or ftScalable Storage6 PCI-X slots, or 4 PCI-X and 2 PCI-Express slotsIntegrated dual 10/100/1000 EthernetWindows® 2003 Enterprise Edition VMware® Infrastructure Red Hat® Enterprise Linux®
33ftServer® 6200 Systems Workloads and Attributes Specifications Stratus ConfidentialftServer® 6200 SystemsWorkloads and AttributesEnterprise-class, quad-core system for transaction-intensive workloadsTop-of-the-line 2-socket serverSpecifications4U rack-optimized form factorQuad-core 2-way SMP2.66 GHz dual-core Intel® Xeon™ processor with 2 x 4MB cacheDual modular redundancy4GB – 24GB memorySAS/SATA, storage supporting 6 internal disksOptional EMC (Windows/VMware only) or ftScalable Storage6 PCI-X slots, or 4 PCI-X and 2 PCI-Express slotsIntegrated dual 10/100/1000 EthernetWindows® 2003 Enterprise Edition VMware® Infrastructure Red Hat® Enterprise Linux®
34Storage Options Fault-tolerant protection of critical data Internal Stratus ConfidentialStorage OptionsFault-tolerant protection of critical dataInternalSoftware-based RAID 1 and 1+0; 6 disk capacitySAS/SATA73 GB, 146 GB 15K RPM SAS250 GB, 500 GB 7200 RPM SATA3.0 TBftScalable StorageHardware RAIDFibre Channel (Direct or SAN attach)27 TBStratus’ storage portfolio is designed to protect customer’s crucial data while complementing our fault-tolerant server architecture. Whether customers need to optimize the availability of their critical information, improve their storage utilization, protect information from disaster or reduce costs through simplicity, Stratus offers world-class storage options to meet any need. Stratus delivers the reliability and capability customers demand for their data storage environments.This slide depicts the range of capability offered for Stratus ftServer systems, from simple internal disks to an external fault-tolerant storage subsystem ..The internal ftServer storage subsystem supports 2, 4, or 6 or SAS or SATA hot-pluggable disk drives.The ftScalable Storage subsystem is designed to address the cost-sensitive, entry-level and mid-range data storage market segments while offering enterprise-class product attributes.Stratus Hardware Assisted Software Mirrors (HASM)HASM provides RAID Level 1 functionality through software, but exploits the Stratus ftServer system design to deliver maximum availability. HASM uses the fully redundant components within the Stratus system – controllers, busses, power supplies, additional error checking on all operations, hot-swap, SSN remote management - as well as the Stratus hardened device driver/SCSI Disk controller combination to ensure that the maximum amount of work is off-loaded from the main system to maximize overall performance.Stratus ftServers running the Windows Operating System can connect with our EMC attachment kit to either the EMC Symmetrix or CLARiiON family of products. This provides fibre channel connected hardware based RAID capability with comprehensive failover and management software capabilities.
35ftScaleable Storage: Feature Set Stratus ConfidentialftScaleable Storage: Feature SetAvailabilityRedundant, hot-plugActive-Active controllersRolling firmware upgradesEnviroStor™ Battery-free cache protectionASN IntegrationSnapshotsHot sparesPerformance15K RPM SASDupliCache: Patented redundant cache optimizationSwitched drive topologyCapacityDynamic expansion: up to 3 shelves27TB maximumDedicated, dual-host DAS, and SAN topologies
36Stratus Tape Backup Family Stratus ConfidentialStratus Tape Backup FamilyVS160 Tape80GB/160GB native/compressed capacity8MBps/16MBps native/compressed data-transfer ratesExternal 1U rack-mount enclosure with single or dual tape drives availableOutput read-compatible with SDLT600Available as standalone tape or with autoloaderAll ftServer platforms can connect to an external VS160 drive, either as a rack-mountable standalone unit or as an autoloader with a capacity of up to 16 cartridges.