云計算技術與架構前瞻

云計算技術與架構前瞻,Presented by: Fubo Zhang,云計算的市場前景云計算的現(xiàn)狀云計算的技術、架構發(fā)展趨勢,內容提要,3,1979: CFD Contributions to the 767,4,2005: CFD Contributions to the 787,Gartner View of Private Cloud Market,可升級的,共享,自動化,彈性計算,451 Group View of the Cloud Market,管理,自動化,安全,存儲,Forrester View on Cloud Computing Platforms,云計算實現(xiàn)的3 條路徑,Time,Scope of sharing,2003,2015,2009,Client/ServerSilos,Public & Private Cloud,9,公共云 - What do we believe?,ESX,Virtual Machine,Virtual Machine,Virtual Machine,Virtual Machine,Virtual Machine,Virtual Machine,物理機,PC cloud,手機移動上網(wǎng) 3G,上網(wǎng)機Internet 接入上網(wǎng),WorkstationVIP 接入,10,虛擬機,公共云 - What do I believe?,11,PC cloud上網(wǎng)筆記本等終端設備臺式電腦、筆記本將會被淘汰虛擬PC服務運營商將會替代MS和google Data Bank解決數(shù)據(jù)、信息安全的最終解決方案衍生許多種數(shù)據(jù)服務， 比如醫(yī)療健康，個人企業(yè)信用征詢，個人數(shù)據(jù)銀行Many SPsPC SPStorage SPGame SPMany SAAS SPsMany other SPs,Platform ISF,HPC Apps,Test & Dev,Analytics,LSF, Symphony, ,LSF, Symphony, ,Private cloud management software,私有云,Java,JBoss, Tomcat, ,數(shù)據(jù)中心體系很復雜， 系統(tǒng)的維護和管理難度大IT成本高，資源占用多，配置峰值資源需求 = 浪費資源系統(tǒng)穩(wěn)定性、可靠性低。人工服務為主，高成本、低滿意度IT傳統(tǒng)模式不能適應業(yè)務部署速度的需求 （至少2個月）,地理上分布的用戶,根據(jù)應用靜態(tài)分配IT資源,Supply and Demand Not Aligned,今天IDC 數(shù)據(jù)中心的挑戰(zhàn),云計算（虛擬化）帶來的好處,資源虛擬化，從而實現(xiàn)全局共享 節(jié)省投資SOA 框架使得IT能對業(yè)務需求提供快速部署， 時間從原來的幾個月縮短到半天 提高了業(yè)務支撐能力一體化管理打破了原來IT煙囪式業(yè)務系統(tǒng)管理，使得系統(tǒng)維護人員的效率大大提高 運維效率提高消除單點故障造成的穩(wěn)定、可靠性問題 提高系統(tǒng)穩(wěn)定性和可靠性幫助IT投資風險和決策 Smart IT節(jié)能減排 綠色IT,實現(xiàn)云計算（虛擬化）的困難,沒有成熟的產(chǎn)品和方案讓運營商有直觀的理解。各個廠商都在推、強調自己的云計算概念和產(chǎn)品，像盲人摸象，無法讓用戶有個全局的認識?，F(xiàn)有IT系統(tǒng)的改動成本和風險傳統(tǒng)業(yè)務系統(tǒng)虛擬化改造少有成功案例可以借鑒組織管理和人們的認識需要提高,Platform在實現(xiàn)云計算方面的優(yōu)勢,Platform已經(jīng)有了自己的云計算產(chǎn)品ISF - IAASPlatform的云計算產(chǎn)品是建立在許多用戶的需求上的， 如 JP Morgen, CityBank, Amazon, Alatum我們幫助用戶實現(xiàn)小步快跑的策略， 降低IT系統(tǒng)的改動成本和風險我們本地有近200名研發(fā)團隊來協(xié)助用戶的業(yè)務系統(tǒng)虛擬化改造Platform已有幾個云計算成功案例Platform可提供業(yè)務培訓和成功用戶案例考察,Java Cloud: Fortune 500 FS Companies,App Server,App 1,煙囪式的業(yè)務系統(tǒng)導致了系統(tǒng)部署時間長，資源利用率低下，IT管理維護成本高,Virtual Servers,Self-Service Portal,100s Java 應用軟件運行在1000s UNIX SMP 服務器,App Server,App 2,App Server,App N,App 1,App 2,App N,挑戰(zhàn),環(huán)境,SaaS Cloud: Fetch Technologies,Platform ISF,CUST A,CUST B,CUST C,CUST X,由于手動部署業(yè)務系統(tǒng)，無法滿足日益增長的業(yè)務需求簡單的用戶系統(tǒng)配置的改變，需要幾個小時乃至幾天,200 virtual servers in production and test/dev using VMware,環(huán)境,挑戰(zhàn),IaaS Cloud: SingTel,在人工操作的方式下， 管理和維護出租的計算機、計算用戶使用情況和計費非常的不方便,Xen, VMware, Hyper-V, and physical blades on 2,400 cores providing public cloud services,挑戰(zhàn),環(huán)境,HPC Cloud: CERN,系統(tǒng)無法為用戶提供靈活的資源配置，需要等待很長的時間,RHEL, KVM, Xen, Hyper-V for widely diverse set of scientific applications on x86 commodity boxes,挑戰(zhàn),利用私有云的技術Platform ISF自動管理和部署系統(tǒng),在固有的資源上， 開發(fā)了更多的服務項目; 對一些應用系統(tǒng)的運行效率提高了一倍多 降低成本， 提高資源利用率 通過自動化管理， 降低了維護費用,解決方案,Benefits,環(huán)境,3 個層次的云計算服務,2017/10/26,21,Each level provides a higher level of service abstraction & may use lower level(s),云計算體系架構,服務調度和管理引擎,云計算平臺VM中間構建,IAAS 調度分配引擎,資源池,IAAS,服務管理,SAAS,PAAS,服務系統(tǒng),IAAS 服務平臺,服務應用系統(tǒng),云計算門戶,IDC 服務,Placement Engine Policy,10/26/2017,23,PlacementEngine,1) Application Topology,DB,AppServer,AppServer,AppServer,WebServer,WebServer,2) ResourceHierarchy,3) Network Access Cost Matrix (Available Bandwidth / Dynamic Utilization),DC,RackGroup,RackGroup,RackGroup,Rack,Rack,Tier 1,Tier 2,Tier 3,Rack,Rack,Rack,Rack,Output: Initial Placement- Run-Time Management- Defragmentations& Migration,DC,/,4) Previous reservations committed & Account Quotas,5) C-language plug-in placement policy:vem_policy_init()vem_policy_fin()vem_policy_distribute()vem_policy_place()vem_policy_housekeep()vem_policy_status()vem_policy_info(),基于資源可用狀態(tài)的分配策略,PTM Framework,Application-Aware 調度,ISF Master,Application Orchestrator,KVM HypervisorResource Group,Deploy &Monitor VMs,Monitor,Get historical usage,Grow/Shrink,Vmware HypervisorResource Group,PhysicalLinux Resource Group,Application PerformanceManager,ICM,PVMO,PTM,Call ISF or sendAlert,1,3,2,需求: 動態(tài)的調整業(yè)務應用對資源的需求改變，而無須修改應用的中間件輸入： 系統(tǒng)負載參數(shù)(cpu, mem) 集成的應用軟件效率管理系統(tǒng) 需求的調度策略,Add custom attributes about hosts from 3rd party tools via APIAttributes can be used in scheduling & placement policiesUseful to organize hosts by location, energy efficiency, model types, lease datesSupports both static and dynamic attribute updates,資產(chǎn)管理與監(jiān)控,AssetDB,ResourceUpdater,Monitoring DB,ISF,CMDB,API,資源使用數(shù)據(jù)用來做計費和能力規(guī)劃,RateDefinition,Usage Data Reserved Amount & Time On-Demand Amount & Time CPU Usage Mem Usage By Account By Application Def By User By Resource Group,Platform ISF,HPC Application,Reporting App,JBOSS App,External Billing System,Other Data,存儲的集成,10G,10TB,Storage Containers,5G,100G,5TB,20TB,2G,10G,StorageAdaptor,ISF Master,25G,50G,適配器來發(fā)現(xiàn)存儲的內容和生成存儲的資源池用戶使用存儲資源就如同使用一個服務的定義系統(tǒng)可以預約存儲資源 系統(tǒng)啟動適配器來生成存儲的分配空間 工作流和后置配置步驟用來連接存儲服務器,Storage Manager APIs (EMC, NetApp),High Performance,Mid-range storage,Low-end storage,網(wǎng)絡集成,企業(yè)私有云,West CoastData Center,East CoastData Center,EuropeData Center,ISF Market Place Broker,Enterprise Messaging Infrastructure / JMS,Enterprise DB Infrastructure,Remote Proxy,Leverage EGO capabilities to distribute ISF services across multiple nodes and provide centralized monitoring, troubleshooting, and HA,Application,End-User,CloudAdmin,ISF Master Node,P-VMO,Portal,ICM,CES,App Manager,BladeLogic,VMware,Amazon,Tivoli,xCAT,ISF Master Node,P-VMO,Portal,ICM,CES,BladeLogic,VMware,Citrix,RedHat,Opsware,xCAT,VMware,xCAT,Nagios,Remote Proxy,VMware,xCAT,Nagios,Platform Computing, Inc.,The leader in cluster, grid and cloud management software:17 years of profitable growth2,000 of the worlds most demanding client organizations5,000,000 CPUs under management500 professionals working across 13 global centersMany Strategic relationships including Dell, Intel, Microsoft, Red Hat and SAS,PlatformClusters, Grids, Clouds, WhateverComputing,Platform Product Portfolio,Thank youwww.platform.comzfbplatform.com,