《利用内存数据库构建云计算平台.pptx》由会员分享，可在线阅读，更多相关《利用内存数据库构建云计算平台.pptx（71页珍藏版）》请在taowenge.com淘文阁网|工程机械CAD图纸|机械工程制图|CAD装配图下载|SolidWorks_CaTia_CAD_UG_PROE_设计图分享下载上搜索。
1、基于云计算实现IT扁平化管理云云计算平台构建与管理工具计算平台构建与管理工具多数据中心间灾难恢复桌面云的评估、迁移、监控及管理议题应用基础架构管理现状每个新应用、每次变更都需要进行大量的配置与调整耗费时间、价格昂贵且拖延产品面市时间没有一个能对单个组件实现管理控制的应用整体管理平台防火墙负载平衡器Web serversWeb serversWeb serversWeb 及应用服务器Web serversWeb serversWeb serversWeb servers数据库服务器存储3交换机$4我们需要什么？我们需要什么？在云计算的环境中，我们需要一种比传统方式更灵活、更节省、更快速的应用操作
2、模式！云计算管理的特点1)云计算管理关注的是架构虚拟，而不仅仅是一个服务器虚拟工具.2)将架构及应用都作为虚拟对象对待.3)将一系列普通PC服务器转化为可扩展的共享资源用于运行在线应用.56软件硬件操作系统Web 服务器操作系统数据库服务器操作系统7层架构代码防火墙网关服务器SAN用软件+更少的硬件如何实现上述目标？在虚拟数据中心中虚拟对象(模板)目录7创建网络、配置组件.Components are networked togetherBy drawing lines8通过“画线”将组件连接互通为组件提供资源.分配内存,CPU,带宽等资源 9创建卷组(Volume)创建“卷组”(Volume
3、)并将其分配给一个Web服务器10虚拟应用创建成功!Once the application or“Virtualized business service”is started-we can open it up and log into one of the components where we are able to set up alarm monitoring11应用启动后，可以登录到应用各组件中进行告警监控等设置云计算需要是一个可满足建模、部署和管理应用所涉及的一切要求的简单产品全局目录全局目录 组件，架构，应用，版本云环境云环境计费Cloud WISA Stack将整个基础设施
4、实现虚拟化来支持应用运行把这种虚拟服务组合作为一个单一对象来进行部署和管理效果：降低模型成本；创建了一个一对多的环境；同时增加了新的服务种类 Cloud LAMPCloud App WikiCloud JAVA StackCloud ERPCloud CRM虚拟数据中心私有云公共云组合组合运行时管理运行时管理自动化,优化,监测,故障容错1213对客户意味着什么？降低成本 单一集成的云平台显著改善了以下几个方面：计算成本计算成本其它解决方案成本其它解决方案成本资源利用率资源利用率业务持续性成本业务持续性成本提高效率 模型驱动的服务创建与复用显著提高了开发速度开发速度灵活性和控制力灵活性和控制力可
5、移植性可移植性增加收益 通过为企业提供私有云，或为MSPs 提供SaaS 和IaaS 14四小时内从无到云具有磁盘及Gigabit网络的PC服务器每台服务器均提供运算、网络及存储能力 统一资源池为所有应用提供服务 降低成本从今天开始 应用虚拟化管理15降低成本从今天开始 虚拟化基础架构16降低成本从今天开始 综合管理17服务性能瓶颈及关联关系分析服务性能瓶颈及关联关系分析云云基础设施（基础设施（IaaS）监控）监控分布式存储监控分布式存储监控云云动态资源监控动态资源监控资源池监控调度资源池监控调度云计算平台构建与管理工具多数据中心间灾难恢复多数据中心间灾难恢复桌面云的评估、迁移、监控及管理议题
6、多数据中心的灾难恢复l利用分布式内存数据库实现数据中心间的数据同步l利用一对多虚拟化容灾实现简单高效的应用级容灾利用内存数据库实现多数据中心的数据同步平台的可用性技术平台的可用性技术基础架构池化主要优势=动态、灵活的基础架构操作系统应用程序数据库配置操作系统应用程序数据库配置数据一致性池化资源的过程即是把物理设备、操作系统、应用配置、数据库文件化的过程 虚拟机文件封装了系统，数据，配置及应用，保障四者的一致性，避免重复建设在灾备中心提供备用虚拟机，在数据级容灾建设的同时即可提供业务级容灾能力备用虚机不实际占用物理资源，真正实现按需分配的动态灾备中心虚拟化容灾备用虚拟机生产中心建立备用虚拟机容灾
7、中心数据级容灾应用级容灾云计算平台构建与管理工具多数据中心间灾难恢复桌面云的评估、迁移、监控及管理桌面云的评估、迁移、监控及管理议题桌面云面临的挑战Are you selecting the right desktops to be virtualized,migrated?Are you confident if your Virtual Infrastructure Design is viable?Can you measure and monitor user experience?(if you can measure it,it matters)Can you identify i
8、ssues at the user,machine,application,network,and storage level?Can you monitor latency of access to network resources and performance of storage?Are you wasting too much time troubleshooting slowness,bottlenecks,issues?How do you migrate Users“goodwill and personalization+user data”over to the new
9、architecture?How do you deliver a consistent desktop experience?How do you avoid vendor-lock?桌面云的核心阶段桌面云的核心阶段评估测试/POC 迁移 管理PhaseDescription评估估Application Inventory,Application Virt Compatibility,User Behavior,Infrastructure Constraints,Performance Baseline,VDI Classification,Workload Bursts设计Capacit
10、y Planning,Infrastructure Requirements,Deployment/Migration PlanningPOCValidation to Baseline,Implement Profile Strategy,Simultaneous Environment Support,Swing to Production迁移迁移User and Application Settings,Profile Encapsulation&Data Migration,Folder Redirection管理管理Application Provisioning,Performan
11、ce Monitoring&Alerting,License Audit,VM Sprawl Audit,Health Check&Diagnostics桌面云的核心阶段在开始桌面云之前Assessment Features and Functionality 桌面合适度桌面合适度Provides an intuitive view of your infrastructure Delivers decision support data for identifying users/desktops for the next generation desktop定定义用用户习惯Monitors
12、 usage and correlates data to identify users and groups that warrant additional resources or updated hardware定定义应用用负载Identifies what users are using which applications and to what extent获取现有配置情况获取现有配置情况Examine details ofexisting configurations:Filter byLocationUser or GroupApplication See distributi
13、on of:OS versions#ProcessorsCPU speedsMemory sizes#MonitorsResolutions#NICsNIC speedsPrintersAge of machines获取资源使用情况获取资源使用情况Examine distribution of resource utilization:CPU,Memory,Network Filter byApplicationUser or GroupLocationDays and times Identify top consumers定义配置群集定义配置群集Find key clustersand c
14、ommonality:Filter byLocationUser or GroupApplication Find clusters across:OS typeCPU typeMemory sizeTotal diskNICs#,speed#Monitors#Printers决定评估结果决定评估结果See how well specificmachines,users and applications fit in VDI:Analyze by group Workload statisticsSystem CPUUser CPUMemory UsePagingDisk I/ONetwork
15、 I/O Advanced metricsUser logonsApp loadingNetwork latencyGraphics intensity Categorize workloads Customizable ratingsBest FitWorst Fit桌面云和适度对比-GoodVs.OthersScopeFindingsVDIFITRecommendationsScopeFindingsVDIFITRecommendations关键数据采集结果分析HourBeginningActiveDesktopsActiveUsersAverageLogonDurationAverage
16、AppLoadAverageSystemCPUAverageUserCPUAverageCPUMhzAverageMemoryUsageAverageDiskI/OAverageDiskI/ORateAverageDiskR/WRatioAverageNetworkI/ORateAverageNetworkLatencyAverageGraphicsIntensity8AM5548n/a5.62%6.00%48.29%21.29IOPS420.99KB/s22.86KB/s9AM6660n/a5.30%5.84%45.79%30.90IOPS622.23KB/s26.16KB/s10AM696
17、2n/a4.25%5.69%46.40%22.62IOPS455.38KB/s20.48KB/s11AM7264n/a4.20%6.95%47.07%19.51IOPS397.94KB/s17.26KB/s12PM7265n/a4.03%7.84%48.35%17.05IOPS332.25KB/s12.39KB/s1PM7266n/a4.26%7.74%49.21%18.34IOPS359.23KB/s13.82KB/s2PM7266n/a4.19%7.70%49.47%16.04IOPS312.55KB/s12.60KB/s3PM7266n/a3.84%7.69%50.07%13.54IOP
18、S262.97KB/s9.83KB/s4PM7167n/a3.72%7.48%49.59%13.89IOPS274.67KB/s12.16KB/s5PM7166n/a4.35%8.22%49.19%21.03IOPS439.53KB/s11.88KB/s6PM6560n/a4.98%8.86%49.77%22.78IOPS491.75KB/s8.78KB/s7PM6055n/a5.39%8.32%49.60%15.23IOPS368.71KB/s9.55KB/s8PM5954n/a5.18%8.30%50.11%18.55IOPS373.14KB/s14.04KB/sScopeFindings
19、VDIFITRecommendationsService Level AssuranceFeatures and Functionality 桌面云建设及维护桌面云建设及维护UX Rating(User Experience)Provides intuitive reporting of user experience rating over timeDiagnosticsIdentifies issues by correlating data and aggregating resultsAlerts proper teams to manage issue(network,hypervi
20、sor,SAN/storage,or desktopDrill down to granular details and reportingProvides user performance reporting by user or groupProvides hardware/software performance reporting by application,machine,or storage device基于策略的用户体验监控基于策略的用户体验监控Rating User Experience(UX)Disk IOPS,Disk Queues,Disk LatencyLogin T
21、ime,Application Load Time,Applications Not Responding(ANR)Network Response Time Latency(NRT),Application Response Time(ART),Incomplete&Failed ConnectionsDetailed View of Endpoint Device(from Desktop through Cloud to the Datacenter)Detailed view of Windows Endpoints(Per App.CPU,Memory,Disk,Network Wo
22、rkload,Page Faults,etc)Network Latency Measured Across Critical Hops Through The CloudNetwork Utilization by Protocol(SLA/Policy can be set per protocol,user,group,machine,IP subnet)性能监控性能监控Monitor and Alert on Application Performance/ResponsivenessApplication Performance(CPU,Mem,Disk,Net,Latency,Pa
23、ging,Launch,Responsiveness)License usage(Who uses App,How Often,What are the patterns,Local or Remote)Detailed connection data for:Monitoring application accessTroubleshooting performance issuesIdentifying dropped connectionsMonitoring service levels网络监控网络监控Visibility to the virtual storage for:Iden
24、tification of storage utilizationMonitoring I/O performanceTroubleshooting I/O problems and trendOverall utilizationPeaks and bottlenecksUser and application storage trendsI/O and performance analysisWhich Vendor Solution is best given observationsStreaming/Single Image/De-duplicated/ZFS 存储监控存储监控自动发
25、现与清单自动发现与清单Dissolvable agent reports:MachinesUsersApplicationsProcessesVirtual Apps 机器连接映射机器连接映射系统诊断系统诊断ScopeFindingsVDIUXRecommendationsUserExperience-OverviewPopulationuser loginApp launch cpu queuedisk queuegraphic intensity Application Non-RespondingPhysical 21.02 2.240.520.00391.160.7Virtual22.
26、232.161.160.0170.5426.14ScopeFindingsVDIUXRecommendationsResourcesusageOverviewPopulationCPU Memory(MB)Disk(IOPS)Network(KB/s)Physical-SITE X3.05%699.4814.482.64Virtual SITE X All 2.72%621.07 6.17 3.17 ScopeFindingsVDIUXRecommendationsUserExperience-OverviewPopulationuser loginApp launch cpu queuedi
27、sk queuegraphic intensity Application Non-RespondingPhysical SITE X21.02 2.240.520.00391.160.7Virtual SITE X All22.232.161.160.0170.5426.14ScopeFindingsVDIUXRecommendationsResourcesusageMachinesBandwidth Avg.Network(KB/s)Virtual SITE X East&West Remote Connection14.04Machine/application3.5 Physical-
28、SITE XRemote ConnectionN/AMachine/application2.64350-HVD-15(10.47.16.132)38.2311.4652.711.28350-HVD-237(10.47.16.126)4.2216.37100718.11350-HVD-227(10.47.17.94)1.8523.06527.416.35350-HVD-287(10.47.17.114)9.2127.689.450.09350-HVD-53(10.47.16.42)1.7627.7859.510.6350-HVD-12(10.47.16.86)9.446.580.480Targ
29、et VMAvg.NRTAvg.Session durationAverage Data TransferredAvg.Transfer Rate350-HVD-15(10.47.16.132)38.2311.4652.711.28350-HVD-237(10.47.16.126)4.2216.37100718.11350-HVD-227(10.47.17.94)1.8523.06527.416.35350-HVD-287(10.47.17.114)9.2127.689.450.09350-HVD-53(10.47.16.42)1.7627.7859.510.6350-HVD-12(10.47
30、.16.86)9.446.580.480RDP/TSconnectionsexceeding8hoursScopeFindingsVDIUXRecommendationsRemoteconnectionbandwidthRDP/TSApplicationAvg.NRTAvg.Session durationAverage Data TransferredAvg.Transfer RateAverage Active transfer rateRemote Desktop Protocol(RDP)(TCP/3389)3.98 ms2.35 hr*135.18 MB14.04 KB/s604 K
31、B/s*Excluding sessions over 8 hours(see next slide)Monitor the complete desktop deliveryWWWserversVM End point VI Network Machine Configuration Info(physical or virtual desktops)OS versions and patch levelsCPUs make,model,speedAge of machines(BIOS date)Monitors and resolutionsDisks,local and network
32、NICS,make,model,speedPrinters,local and networkOther peripherals(USB,etc.)Hosts(VMs only)Full data access via Custom Report DesignerMachine Utilization Info(physical or virtual desktops)Total uptimeTotal time users logged inSystem CPU consumptionUser CPU consumptionCPU queue lengthMemory,page file c
33、onsumptionDisk%used,IOPS,KB/sDisk queue lengthNIC%used,KB/sApplication Usage Info(physical or virtual desktops)Application names and versionsNumber of executionsUsersLoad times(launch,initialize)Total running timesNon-responding instancesGraphic intensity(GDI objects)System CPU consumptionUser CPU c
34、onsumptionMemory consumptionPage faultsIO consumption(disk,network)Network Usage Info(taken from virtual switch)Track users(Connector ID)Track machines(Connector ID)Applications(includes RDP,ICA)Number of connectionsNumber of packetsFailed connectionsUnauthorized connectionsBandwidth by user,applica
35、tionApp Response Times(ART)Network Roundtrip Times(NRT)User Login InfoMachines used by each userTime of loginsDuration of logins(login delay)Time of logoffsSAN Usage Info(FC,iSCSi,NFS)IOPS,KB/s per LUN%Storage used per LUNMachines active on each LUNUsers active on each LUNMetadataUser groups(import
36、from AD/LDAP)Machine groups(import from vCenter,XenCenter)Service level policies,alert thresholds是什么是什么Windows 个性化个性化?By default,Windows embeds personality into the OS:Local Windows User AccountMy Documents,My Photos,etcSpellchecker/DictionariesFavoritesWallpaperEtc.System Registry SettingsApplicati
37、on-level settingsCustom settings/buttonsCertain windows settingsUnique Configuration per user/department/groupDrive mappings/PrintersIP settingsGranular custom configurationsProfileUnity separates personalization(profiles and data)and desktop configuration(settings,etc)from the Windows OS.Profiles a
38、nd Configurations are stored locally and on the network.They are accessed at logon and saved at logoff without a client-side agent.Delivery MethodWindows OSApplicationsLocal/NativeMicrosoft App-VVMware ThinAppCitrix StreamingImplement a Layered Desktop EnvironmentUser Prefs&App SettingsCustom Settin
39、gsSpellcheckers/DictionariesMail Settings Sigs&MoreUser ProfileMap Drives Folder RedirectsInternet&Browser SettingsPrinter Mapping&AssignmentEnvironment&MoreAdmin SettingsBreak free from Roaming ProfilesSpeed Logon/LogoffImport Legacy ProfilesEliminate Roaming Profile Corruption and Mandatory Profil
40、e ComplexitiesRetain Centralize Control without ScriptingLayer User Settings with Target OS Awareness(Supports User/Group/IP Subnet awareness)Allow Users to Switch Between Desktops&Operating Systems(No Agents/Databases to support!)Compatible with VMware RTO and Citrix UPMProfile and Configuration 管理
41、解决方案管理解决方案It separates personalization(profiles and data)and desktop configuration(settings,etc)from the Windows OS.Profiles and Configurations are stored locally and on the network.They are quickly accessed at logon and saved at logoff without a client-side agent.PHYSICAL OR VIRTUAL DESKTOPSScalabl
42、eScalableNo-Touch AgentNo-Touch AgentNo SQL DatabaseNo SQL DatabaseNo Data Lock-inNo Data Lock-inNo Kidding!No Kidding!Profile and Configuration Management SolutionKey Features OverviewData stays in Windows native formatWorks independent of Roaming ProfilesMigrates&Redirects User Authored DataMigrates existing XP/2000 Windows profiles to new platformCompatible with RTO Virtual Profiles and Citrix UPMVMware ThinApp DeploymentNetwork side momentary agent no desktop software to install Zero RAM/CPU usage during sessionSupports Vista,Windows 7No SQL&IIS Servers neededCan configure user settings