Downloaded 174 times
Uploaded byRightScale
Tips For a Successful Cloud Proof-of-Concept - RightScale Compute 2013
The document provides a guide for successful cloud proof-of-concepts (POCs), emphasizing the need to adapt evaluation methodologies to fit the evolving technology landscape. It outlines key dos and don'ts for selecting clouds and applications, encouraging lightweight, real-world testing while measuring success through agility, efficiency, and improved solutions. Ultimately, it advocates for starting with simple use cases and treating POCs as learning experiences to leverage cloud advantages without attempting to replicate traditional data center setups.
More Related Content
Similar to Tips For a Successful Cloud Proof-of-Concept - RightScale Compute 2013
![Governance, Deployment & Methodologies for Agentic Automation [2/3]](https://cdn.slidesharecdn.com/ss_thumbnails/agenticcommunityseries-day2-251112135038-a386476d-thumbnail.jpg?width=640&height=640&fit=bounds)
Tips For a Successful Cloud Proof-of-Concept - RightScale Compute 2013
- 1. april25-26 sanfrancisco cloud success startshere Tips for a Successful Cloud Proof-of-Concept Vijay Tolani, Cloud Solutions Engineer
- 2. #2#2 #RightscaleCompute Agenda • State ofthe Cloud • Cloud POCs: Dos and Don’ts • Choosing the Right Cloud(s) • Choosing the Right Application • Navigating the Technical Hurdles • Measure the Success of Your Cloud POC
- 3. #3#3 #RightscaleCompute Conventional Vendor EvaluationTechniques • Steering Committees to Determine Requirements • Analyst Reports for Vendor Selection • RFPs to Filter Qualified Vendors • POCs for Qualified Vendors
- 4. #4#4 #RightscaleCompute Today’s Challenge: BusinessExpectations Have Changed Instant On Self Service Unprecedented Scale Public & Private Cloud Business Expectations
- 5. #5#5 #RightscaleCompute And The TechnologyIs Evolving Quickly Open Sourced Goes OpenStackLaunched Adds Australia Grizzly Released Apr Jun Aug Oct Dec Jan Mar 2012 2013 GALaunched Cloud Services
- 6. #6#6 #RightscaleCompute And If You’reNot Careful • You’ll find yourself applying conventional methodologies to evaluate revolutionary technology: • By the time Steering Committees determine requirements, needs have changed • Information coming from analysts provide point-in-time snapshots, but may lack real-time feedback • RFPs ask questions which are too specific to apply to revolutionary technology • POCs test for the worst case, while the technology is being built to start with the simplest use case
- 7. #7 #rightscalecompute Cloud POCs: Dosand Don’ts
- 8. #RightscaleCompute #8#8 Pick the right approach Alearning experience, not once and done Cloud POC Mantra
- 9. #9#9 #RightscaleCompute Pick the RightApproach • Keep it lightweight • Start simple • Keep it short – 30-60 days • Accept cloud as it is, don’t try to make it into your data center • Don’t start with the hardest use case • Decide if the architecture meets the needs of the application, not if it matches your data center • Treat it as a learning experience • It’s not once and done • Your cloud deployments will evolve • Your expertise and knowledge will evolve • Consider how you will manage the application over time
- 10. #10 #rightscalecompute Cloud POCs: Dosand Don’ts Choosing a Cloud
- 11. #11#11 #RightscaleCompute Choosing a Cloudfor POCs • Public & Private Clouds built on same technologies • OpenStack • CloudStack • Private Clouds are good choices for applications which require: • High Flexibility & Performance • Increased Security & Compliance • CapEx Procurement Model • Public Clouds offer more advanced automation than Private Clouds
- 12. #12#12 #RightscaleCompute Public Clouds EnableAdditional Agility • “It’s a marathon, not a sprint” • Don’t sprint directly to the finish line, without training. You’ll hurt yourself • Start with something small and show it works • Start with the simplest application, then test more complex
- 13. #13 #rightscalecompute Cloud POCs: Dosand Don’ts Choosing an Application
- 14. #RightscaleCompute #14#14 Pick the right application Cloudis a fit for everyone, not for every application Cloud POC Mantra
- 15. #15#15 #RightscaleCompute Pick the RightApplication • Keep it real • Pick a real application for a real project with real expectations • Don’t start with the most complex use case • Low to moderate security/compliance profile • Focus on where cloud adds value • Speed to deploy, scalable, variable load • Good options for your first deployment in the cloud • Web Application • Mobile/Social/Collaboration Apps • Big Data • Batch Processing • Dev/Test solutions
- 16. #16#16 #RightscaleCompute Clouds Are Different •… so don’t treat them like a data center in the sky • Don’t expect to be able to move a VM image to a cloud and have it work out of the box • Successful cloud deployments include HA Application Architectures designed to run on ephemeral, commodity infrastructure • NoSQL Databases vs. High Performance Hardware
- 17. #17 #rightscalecompute Cloud POCs: Dosand Don’ts Navigating Technical Hurdles
- 18. #18#18 #RightscaleCompute Next Step: TechnicalPOC • After proving your application can run on a cloud architecture, consider testing with a Private Cloud • Test for Performance • Ensure Security & Compliance • Integration with Existing ITSM Processes • Consider Including More Complex Workloads
- 19. #19#19 #RightscaleCompute Technical Pitfalls toWatch Out For • Storage Performance • Consider Striping, Sharding, & NoSQL Architectures • Shared Storage • Consider OpenNFS • Software Licensing • Consider Licensing through the MultiCloud Marketplace • Network Flexibility • Consider Dynamic DNS • Security & Compliance • Design for Public Clouds
- 20. #20 #rightscalecompute Cloud POCs: Dosand Don’ts Metrics for Success
- 21. #RightscaleCompute #21#21 Pick the right metrics Beginyour cloud journey with the end in mind Cloud POC Mantra
- 22. #22 #rightscalecompute Cloud POC SuccessMetric #1 Business Agility How quickly can cloud help us respond to business needs and opportunities? Business Opportunity Market Window Time to Market $£¥€ ✖ Shadow IT
- 23. #23#23 #RightscaleCompute Evaluation Criteria • Speed/Easeto: • Launch new applications or services • Scale applications or services up or down to meet demand • Offer services in new geographies • Self-service IT • Give control to business users • Give control to developers • Automation • Automate routine work for Operations
- 24. #24 #rightscalecompute Cloud POC SuccessMetric #2 Improve Efficiency Can cloud make us more efficient with time and money?
- 25. #25#25 #RightscaleCompute Evaluation Criteria • Costs •Data Center vs Cloud costs • CapEx vs OpEx • Opportunity costs? • Time • Time to plan/prepare applications or services • Time to deploy applications or services • Time to manage applications or services
- 26. #26 #rightscalecompute Cloud POC SuccessMetric #3 Better Solutions Can cloud helps us deliver better solutions that meet customer needs?
- 27. #27#27 #RightscaleCompute Evaluation Criteria • CanI support new solutions that were difficult before? • Web • Mobile • Social • Gaming • Big Data • Batch Processing • Can I support better solutions? • Geographically Available • Highly Scalable • Highly Available
- 28. #28#28 #RightscaleCompute Summary • Start witha real application that’s a good fit for cloud • Treat it as a learning experience • Use public cloud to evaluate the feasibility of cloud • Then build a private cloud • Leverage the strength of cloud – don’t try to recreate your data center • Measure success based on broad business goals • Agility • Efficiency • Better solutions
- 29. #29#29 #RightscaleCompute Thinking about aCloud POC? • Ask about a Cloud Strategy Briefing • Personalized briefing from the cloud experts for your IT executives and managers: • Real-World Cloud Trends • Strategic Opportunities in the Cloud • 5 Technical Considerations for Cloud • Best Practices for Cloud POCs Contact RightScale (866) 720-0208 [email protected] www.rightscale.com
- 30. april25-26 sanfrancisco cloud success startshere Questions?
Editor's Notes
- #6 Some core IaaS components are converging – example, virtual machinesSome core IaaS remain distinct – networking, storage, higher level services