A Gentle Introduction to Cloud Computing Source: Wikipedia
Platform Computing, Inc. Platform Clusters, Grids, Clouds, Whatever Computing The leader in managing large scale shared environments o 18 years of profitable growth o 2,000 of the world s most demanding client organizations o 5,000,000 CPUs under management o Headquarters in Toronto, Canada o 500+ professionals working across 13 global centers o Partnerships with Dell, HP, IBM, Intel, Microsoft, Red Hat and VMWare 2
What is Cloud? 3
Evolution of Cloud Computing HPC/ Enterprise Grid Public & Private Cloud Customer Profile Internet Internet Services Client / Server Silos Mainframe PC VM Cluster Big Data,BI/Analytics, 1980 1980 1990 2000 2010 2015 Time 4
Cloud Computing Defined Model for enabling convenient, on-demand network access to a shared pool of configurable computing resources that can be rapidly provisioned and released with minimal management effort or service provider interaction. - NIST Server Storage 5 Key Characteristics On-demand, self-service Shared resource pools Rapid elasticity Measured service Broad network access Network 5
Cloud Taxonomy Service Models SaaS Software-as-a-Service End user applications Single or multi-tenant PaaS Platform-as-a-Service Application development platform E.g., middleware, multi-tier IaaS Infrastructure-as-a-Service Full range of IT components E.g., VMs, database, OS
More Cloud Taxonomy Deployment Models Public Made available to the public Pay as you go Security, compliance, performance questions Private Operated for one organization Behind corporate firewall Maximize control Requires infrastructure and management Hybrid Composition of private and public cloud usage Federated data and process sharing
Public Clouds TORONTO 3/21/2011
Public Clouds Amazon EC2/S3 Microsoft Azure Google AppEngine Salesforce.com 9
Amazon EC2 10
Amazon EC2 Concepts Network Security Model: Explicit access control Security groups Start/stop programmatically Amazon Machine Images Bootable root disk Pre-defined or user-built Catalog of user-built AMIs Instance Running copy of an AMI Launch in less than 2 minutes Start/stop programmatically
Microsoft Azure!
Google AppEngine 14
Salesforce.com 15
Risks in Public Cloud Confidentiality name, home addresses, birth dates, medical information, or credit card or other financial data Security Protection against physical and electronic security vulnerabilities Service Usability Reliability or Supplier Viability Will your web 2.0 service provider goes bust/bought over? Example: twitter ( no money- business model, just burning cash) Data Protection Risk Data backups and ability to restore data in the event data is lost or corrupted Your org is prosecuted under Data Protection legislation. Is a Web 2.0 SP a Data Processor? Can users maintain their privacy? Are personal details send outside your country/area? Can you delete information? Have you clarified what you are doing?
Private/Hybrid Enterprise Clouds TORONTO 3/21/2011
Enterprise Data Center Challenges Google & Amazon Have Reset the Bar for Delivering IT Services End users want systems much faster High IT costs Lots of progress but - Low utilization - Manual, repetitive processes - Excessive space & power consumption Support existing apps and IT tools while improving Vendor lock-in 19
Expectations for Cloud Computing Increase agility and innovation - Enable self-service delivery (minutes) - Deliver on SLAs - Simplify process for what-if experimentation - Gain control over public cloud usage Decrease costs - Increase utilization - Increase operational efficiency (100s server per admin) - Achieve a greener data center - Maintain vendor choice 20
Common Use Cases Development Test / QA Production Help Desk HPC 21
Platform ISF for Cloud Management Applications and Middleware Complete, Modular Cloud Management Product Operational Management Platform ISF Application Support Reporting & Self-Service Chargeback Allocation Engine Resource Integrations Workflow & Approvals Supports Entire Cloud Lifecycle from Request to Reclamation VM Management H/W Provisioning External Clouds Server Storage Network Data Center Systems 22
Top Insurance Company Problem Slow: weeks to get resources from IT High cost: still low utilization with VMs Manual labour: managing silos of IT Idle resources: not reclaimed Test/Dev App Users IT Users Platform Solution Start with self-service Dev/Test cloud (VM and WebSphere) Integrated with internal app configuration system & vlans for isolation Future: enable customer SaaS & cloud bursting Self-service Portal Infrastructure Middleware Platform ISF 23 Results Reduce provisioning time from weeks to minutes Increase innovation by reducing IT friction Reduce cost of IT infrastructure Data Centers
Major Bank Problem Low utilization of >1000 UNIX SMP server silos High operating costs and long delivery times with manual provisioning App dev expensive & not standardized App 1 App Server App 2 App Server App N App Server 24 Platform Solution Java cloud spanning dev, test, UAT & production Full commodity stack: Linux & Xen/KVM, JBoss Multi-data center Active-Active for HA Results Guaranteed 2 sec Web response time IT delivery time in minutes not weeks Increase productivity of 1000s app developers Reduce capital cost by 50% Enable vendor choice and flexibility App 1 App 2 App N Self Service Portal App Server Platform ISF Virtual Servers
Leading Government Contractor Problem High costs: infrastructure and admin duplication across 100s of silo LOB programs Long IT delivery times for end users No global visibility to IT usage Platform Solution Corporate cloud data center with secure access to each group and agency Hierarchical self service: end users, Line-of- Business managers, cloud managers Resource quotas to balance requests with funding Results <1 minute response to self-service requests Reduced admin cost Improved capacity, performance, support and maintenance Self-service Portal LOBs & Agencies Infrastructure Platform ISF Middleware Resource Pools 25
Building a Cloud: Technology & Process Challenges TORONTO 3/21/2011
Where Are You With Your Cloud Strategy? 1. Trying To Understand Cloud 2. Creating The Business Case 3. Building a Proof of Concept 4. Planning to Implement in 2012 5. Building Now 6. Implemented 27
Vendor Landscape IBM BMC CA Platform ISF Product Footprint VMware VCE/ Cisco HP Surgient DynamicOps Cloud.com Abiquo Eucalyptus Openstack Openness
Journey To Private Cloud 1 Discovery 2 3 Use Cases Proof of Concept 4 Production 29
Cloud Is More Than VM 1 2 3 4 Agility / Self- Service Server Consolidation Virtual Data Center Cloud Data Center Automation, Orchestration, Scalability! " 30
People & Politics Business Sponsorship - Executives - End users 1 2 3 4 IT from Cost Center to Services - Culture - Process and operations Cloud Governance - Security - SLA - Usage guidelines (e.g., bursting) - Chargeback 31
From Infrastructure to Applications App Users Applications (SaaS Enable) 1 2 3 4 App Developers Middleware & App Platforms (PaaS Enable) IT Users Infrastructure Services (IaaS) Cloud Management Resource Pools 32
Build Cloud Business Case Depends upon use case 1 2 3 4 Strategic justification - Cloud is transformative - Return = speed, agility, customers Use of the classic ROI - Cost = OpEx (# servers / admin), CapEx (commodity stack, # servers), investment (license, services) - Risk = ready to deploy vs. stitching together Assumptions are key need real data points 33
Run A Proof of Concept (POC) Focus on key customer requirements and include customers in the process (build support) Manageable size with focus Dedicate resources (people and technology) Plan & sign off by stakeholders Defined schedule (4-8 weeks) Compile results for your business case 1 2 3 4 34
Expand Production Over Time Pilot - Start small and build on success - Implement functions in stages: self-service, reservations, bursting, elasticity, chargeback, etc. Adapt your IT processes, policies & procedures - Service provider vs. cost center - Build & train your cloud operations team - New metrics for SLA Develop robust service catalog - Reusable services / design for elasticity Expand capabilities and use cases 1 2 3 4 35
Private Cloud is Your Journey 1 2 3 4 Applications Company A Platforms Company B Company C Infrastructure Dev/ Test Production Internal Production External 36
Continuously Seek To Consolidate 1 2 3 4 Consolidate Apps Based on Resources and Op Requirements 37
Summary Evolution of Cloud Cloud Definitions & Concepts Public Cloud Examples Private/Hybrid Clouds Building A Cloud 38