Very few trends in IT have generated as much buzz as cloud computing. This talk will cut through the hype and quickly clarify the ontology for cloud computing. The bulk of the conversation will focus on the open source software that can be used to build compute clouds (infrastructure-as-a-service) and the complimentary open source management tools that can be combined to automate the management of cloud computing environments. The discussion will appeal to anyone who has a good grasp of traditional data center infrastructure but is struggling with the benefits and migration path to a cloud computing environment. Systems administrators and IT generalists will leave the discussion with a general overview of the options at their disposal to effectively build and manage their own cloud computing environments using free and open source software. [Presented as part of the Open Source Build a Cloud program on 2/28/2012 - http://cloudstack.org/about-cloudstack/cloudstack-events.html?categoryid=6]
- 1.C R A S H CO U RS E O NO P E N S O U RC E C LO U D CO M P U T I N G Mark R. HinkleDirector, Cloud Computing Community CloudStack.org
2. AG E N DA1. Introduction2. Quick Cloud Computing Overview3. Open Source Building Blocks for Cloud Computing4. Open Source Tools for Cloud Management5. Questions 3. % W H OA M I 4. Q U I C K C LO U D CO M P U T I N G OV E RV I E W :O R T H E O B L I G ATO RY W H AT I S T H EC LO U D ? S L I D ES 5. F I V E C H A R AC T E R I ST I C S O F C LO U D S 6. C LO U D CO M P U T I N G S E RV I C E M O D E L SUSER CLOUD a.k.a. SOFTWARE AS A SERVICESingle application, multi-tenancy, network-based, one-to-many delivery ofapplications, all users have same access to features.Examples: Salesforce.com, Google Docs, Red Hat Network/RHELDEVELOPMENT CLOUD a.k.a. PLATFORM-AS-A-SERVICEApplication developer model, Application deployed to an elastic service thatautoscales, low administrative overhead. No concept of virtual machines oroperating system. Code it and deploy it.Examples: Google AppEngine, Windows Azure, Rackspace Site, Red HatMakaraSYSTEMS CLOUD a.k.a INFRASTRUCTURE-AS-A-SERVICEServers and storage are made available in a scalable way over a network.Examples: EC2,Rackspace CloudFiles, OpenStack, CloudStack,Eucalyptus, Ubuntu Enterprise Cloud, OpenNebula 7. D E P LOY M E N T M O D E L SP U B L I C , P R I VAT E & H Y B R I D C LO U D S 8. C LO U D ST I L L R EQ U I R ESA RC H I T EC T U R A L D ES I G N Cloud Computing isnt a magicalsolution apps need to be able toscale out Design your architecture with theend in mind Make your infrastructure easilyreplicable 9. B U I L D I N G CO M P U T E C LO U D SW I T H O P E N S O U RC E S O F T WA R E 10. W H Y O P E N S O U RC E ? User-Driven Solutions to Real Problems Lower barrier to participation Larger user base, users helping users Aggressive release cycles stay current with the state-of-the-art Open data, Open standards, Open APIs 11. O P E N V I RT UA L M AC H I N E FO R M AT SOpen Virtualization Format (OVF) is an openstandard for packaging and distributing virtualappliances or more generally software to be run invirtual machines. Formats for hypervisors/cloud technologies: Amazon - AMI KVM QCOW2 VMware VMDK Xen IMG VHD Virtual Hard Disk - Hyper-V 12. S O U RC I N G O P E N S O U RC E S O F T WA R EV M S A N D C LO U D A P P L I A N C ES 13. O P E N S O U RC E H Y P E RV I S O RSOpen Source Xen, Xen Cloud Platform (XCP) KVM Kernel-based Virtualization VirtualBox* - Oracle supported Virtualization Solutions OpenVZ* - Container-based, Similar to Solaris Containers or BSD Zones LXC User Space chrooted installsProprietary VMware Citrix Xenserver Microsoft Hyper-V OracleVM (Based on OS Xen) 14. O P E N S O U RC E CO M P U T E C LO U D SYear Started License Virtualization Technologies2008 GPL Xenserver, Xen Cloud CloudStackPlatform, KVM, VMware2006 GPL Xen, KVM, VMware Eucalyptus(commercial version)2010 (Developed by ApacheVMware ESX and ESXi, , OpenStackNASA by Anso LabsXen, Xen Cloud Platformpreviously)KVM, LXC, QEMU and Virtual Box 15. SCALE-UP OR SCALE-OUTVertical Scaling (Scale-Up) Allocate additional resources toVMs, requires a reboot, no needfor distributed app logic, single-point of OS failureHorizontal Scaling (Scale-Out) Application needs logic to workin distributed fashion (e.g. HA-Proxy and Apache, Hadoop) 16. O P E N S O U RC EP L AT FO R M - A S - A - S E RV I C E Year Started SponsorsLanguages/FrameworksCloudFoundry 2011 VMwareSpring for Java, Ruby forRails and Sinatra,node.js, Grails, Scala onLift and more viapartners (e.g. Python,PHP)OpenShift ** 2011 Red Hat Java, Ruby, PHP, Perl andPythonPHPFog*2011 AppfogPHP, NodeJS, Ruby,Python, Java, .NET,MySQL, PostgreSQLStackato* ActiveState Java, Python, PHP, Ruby,Perl, Node.js, othersWSO2 Stratus 2010 WSO2Jboss, Java EE6 17. O P E N S O U RC EC LO U D CO M P U T I N G STO R AG E GlusterFS Scale Out NAS system aggregating storage overEthernet or Infiniband Ceph Distributed file storage system developed byDreamHost OpenStack Object Storage (SWIFT) Long-term storageobject storage system Sheepdog Distributed storage for KVM hypervisors NFS Old standby, tried and true, not designed for cloud scaleor performance 18. C LO U D A P I S A R E N T C R EAT E D EQ UA LO P E N S O U RC E A B ST R AC T I O N S jclouds libcloud deltacloud fog 19. P R I VAT E C LO U D A RC H I T EC T U R E 20. M A N AG I N G C LO U D S W I T HO P E N S O U RC E TO O L S 21. AU TO M AT I O N U N LO C KST H E P OT E N T I A L O F T H E C LO U D 22. T H E M Y T H O F T H E N I N ESAvailability % Downtime per Downtime perDowntime per Year Month Week99.9% (three nines)8.76 hours 43.2 minutes10.1 minutes99.95% 4.38 hours 21.56 minutes 5.04 minutes99.99% (four nines)52.6 minutes 4.32 minutes1.01 minutes99.999% (five nines) 5.26 minutes 25.9 seconds6.05 seconds99.9999% (six nines) 31.5 seconds 2.59 seconds.0605 seconds 23. 4 T Y P ES O F M A N AG E M E N T TO O L SProvisioningInstallation of operating systems and othersoftwareConfiguration ManagementSets the parameters for servers, canspecify installation parametersOrchestration/AutomationAutomate tasks across systemsMonitoringRecords errors and health of ITinfrastructure 24. M A N AG E M E N T TO O LC H A I N S MonitoringPatchingandProvisioningConfiguration 25. O P E N S O U RC EP ROV I S I O N I N G TO O L S Year Started License InstallationTargetsKickstart?GPL Most .dep and RPMbased Linux distrosCobbler (Plus koan 2007 GPL Red Hat, OpenSUSEfor PXE boot of Fedora, Debian,VMs)UbuntuSpacewalk2008 GPL Fedora, CentosCrowbar2011 Apache(Bare metalprovisioning) 26. OPEN SOURCEC O N F I G U R AT I O N M A N A G E M E N T TO O L SYear Started Language License Client/Server Cfengine 1993 CApache Yes Chef 2009 Ruby Apache Chef Solo No Chef Server - Yes Puppet 2004 Ruby GPLYes & standalone Salt 2011 Python Apache yes 27. O P E N S O U RC EM O N I TO R I N G TO O L SLicenseType of Monitoring CollectionMethodsCacti / RRDTool GPLPerformanceSNMP, syslogGraphiteApache 2.0 PerformanceAgentNagiosGPLAvailability SNMP,TCP, ICMP,IPMI, syslogZabbixGPLAvailability/SNMP, TCP/ICMP, Performance andIPMI, Synthetic more TransactionsZenossGPLAvailability,SNMP, ICMP, SSH, Performance, Event syslog, WMI Management 28. O P E N S O U RC EAU TO M AT I O N / O RC H EST R AT I O N TO O L SYear Started Language License Client/Server SupportOrganizationCapistrano2006RubyMIT Yes NoneRunDeck 2010JavaApacheYes DTO SolutionsFunc2007PythonGPL Yes Fedora ProjectMCollective 2009RubyApacheYes PuppetLabsSalt2011PythonApacheYes SaltStack Inc.? 29. CO N C E P T UA L AU TO M AT E D TO O LC H A I NGenerate Images BootStrapped Image ProvisionConfigurationSUSE Studio CloudStackCobblerPuppetBoxGrinderOpenStackKickstartChefMonitoringNagios Start/Stop ServicesZenossRunDeck Cacti Capistrano MCollective 30. Questions?SLIDES CAN BE VIEWED AND DOWNLOADEDAT: H T T P : / / W W W. S L I D E S H A R E . N E T/ S O C I A L I Z E D S O F T WA R E / 31. CO N TAC T M E 32. A D D I T I O N A L R ES O U RC ES Devops Toolchains Group DevOps Wikipedia Page Open Cloud Initiative NIST Cloud Computing Platform Open Virtualization Format Specs Clouderati Twitter Account Planet DevOps 33. C R A S H CO U RS E I N O P E N S O U RC E C LO U D CO M P U T I N GB Y M A R K R . H I N K L E I S L I C E N S E D U N D E R A C R E AT I V E C O M M O N S AT T R I B U T I O N - S H A R E A L I K E 3 . 0 U N I T E D S TAT E S L I C E N S E .