Kelley Blue Book and Cloud Computing
•Download as PPTX, PDF•
6 likes•2,547 views
Deck delivered at Web 2.0 Expo 2010 San Francisco, for the "Cars on the Ground, Customers in the Clouds" session.
Recommended
Recommended
More Related Content
What's hot
What's hot (20)
Viewers also liked
Similar to Kelley Blue Book and Cloud Computing
Similar to Kelley Blue Book and Cloud Computing (20)
More from David Chou
More from David Chou (20)
Recently uploaded
Recently uploaded (20)
Kelley Blue Book and Cloud Computing
- 1. Cars on the Ground, Customers in the Clouds Scaling a Website While Enhancing Innovation
- 2. Cloud Computing as a Platform Andy Lapin Director, Enterprise Architecture, Kelley Blue Book alapin@kbb.com linkedin.com/in/andylapin twitter:@andylapin
- 3. About Kelley Blue Book
- 4. Cloud as a Utility Computing Platform Why? Cost Flexibility
- 5. Cloud as a Utility Computing Platform How we operate servers on premise
- 6. Cloud as a Utility Computing Platform How we use our servers on premise
- 7. Cloud as a Utility Computing Platform How we really operate our servers on premise
- 8. Cloud as a Utility Computing Platform How do we pay for servers on premise?
- 9. Cloud as a Utility Computing Platform Viral Marketing
- 10. Cloud as a Utility Computing Platform How much flexibility is there?
- 11. Cloud as a Utility Computing Platform Problems with this approach Focus on planning for scale rather than building applications Hard to handle the unexpected Load and stress testing becomes critical up front Large initial cost
- 12. Cloud as a Utility Computing Platform Why not just virtualize everything? (“The Private Cloud”)
- 13. Cloud as a Utility Computing Platform Why not just virtualize everything? (“The Private Cloud”) Is this realistic? How many problems have we really addressed? Focus on planning for scale rather than building applications Hard to handle the unexpected Load and stress testing becomes critical up front Large initial cost
- 14. Cloud as a Storage Platform Same benefits as Compute Built-in redundancy Built-in scalability Lower initial cost, and most likely lower overall cost
- 15. Demo
- 17. 14M UU/month
- 18. Multiple physical data centers
- 20. Client – Silverlight, DeepZoom, WPF
- 21. Web – IIS 7, ASP.NET MVC
- 22. Server – SQL Server 2008, Windows Server 2008
- 23. Tools – Visual Studio 2008, .NET Framework 3.5
- 24. 63,000 lines of .NET code
- 25. 2.5 GB SQL Server DB w/ 125 tables, 15 indexed views, 117 stored procs, etc
- 26. Cloud Solution
- 28. Windows Azure Web Role
- 30. Benefits
- 31. ~$100,000 savings / year in hosting costs alone
- 32. Retire failover data center (27 Web servers & 9 SQL Servers)
- 33. <1% code changes needed for application compatibility
- 35. Less Data, More Apps Kevin Boyle microsoft corporation kevin.boyle@microsoft.com twitter:@kevinmsft
- 36. @kevinmsft
- 38. Marketplace Standard Data Access (OData) Services forContent Providers Billing Account Management
- 39. Scale as a Competitive Advantage David Chou microsoft corporation david.chou@microsoft.com twitter:@davidcchou
- 40. Cloud as a Platform Utility computing On-demand infrastructure Self-provisioning and servicing Rapid elasticity Economy of scale Operational expenditures Infrastructure-as-a-Service Service delivery model
- 41. Cloud as a Platform Native cloud applications Horizontal scaling (scale-out) Parallelization Shared-nothing Multi-tenancy Failure resilient (or fail-in-place) Service-oriented Staged production Federated composition Platform-as-a-Service Application development model
- 42. Scale as a competitive advantage 2009 stats: +200B page views/month >3.9T feed actions/day +300M active users >1B chat messages/day 100M search queries/day >6B minutes spent/day +20B photos, +2B/month growth 600K photos served /sec 25TB log data /day processed 120M queries /sec on memcache Infinite storage. Clouds of processors. Our ability to capture, warehouse, and understand massive amounts of data is changing science, medicine, business, and technology. As our collection of facts and figures grows, so will the opportunity to find answers to fundamental questions. Because in the era of big data, more isn't just more. More is different. Wired Magazine Issue 16.07 – 06.23.2008
- 43. Cloud as a scale enabler Applications at Internet scale Social networking Social and geo analysis Web as a platform Big data (“The Petabyte Age”) Real-time web Semantic web Derived intelligence Next-gen applications Bio-engineering Financial modeling Federated business processes
- 44. Thank you! © 2010 Microsoft Corporation. All rights reserved. Microsoft, Windows, and other product names are or may be registered trademarks and/or trademarks in the U.S. and/or other countries. The information herein is for informational purposes only and represents the current view of Microsoft Corporation as of the date of this presentation. Because Microsoft must respond to changing market conditions, it should not be interpreted to be a commitment on the part of Microsoft, and Microsoft cannot guarantee the accuracy of any information provided after the date of this presentation. MICROSOFT MAKES NO WARRANTIES, EXPRESS, IMPLIED OR STATUTORY, AS TO THE INFORMATION IN THIS PRESENTATION.
Editor's Notes
- CostPay for what you useCapEx to OpExRemove the huge spikes when buying hardwareMinimize the need for load/stress testingImprovements see a direct savings. It’s hard to realize immediate savings when you have already invested in hardware.FlexibilityAlways have additional capacityReduce the planning cycles for marketingFocus on our business, not on ITPlanning time can be spent in product planning, not scalability planning
- Consistent resources based on long planning and purchasing cycles
- For the most part, predictable trafficDaily, weekly and seasonal fluctuations in trafficUp to a 95% drop-off in peak vs. non-peak
- We actually try to run our servers under 50%, even at peak traffic.This isn’t uncommon, but why?Waiting for something to go wrong?Able to handle extra load?Because you don’t want to have to ask for money too often?Because this is the way it’s done?
- We buy our hardware so we see big spikes in cost and there is a lower operational cost to running those servers.Everything is doubled because we have 2 datacenters, one is our primary and the other one is for DR and to handle burst capacity.Important note: your savings increases greatly the more variable your load is. If you plan to operate the cloud like your on-premise data center you’ll see a reduction in un-front costs, but probably end up paying more long term. Renting is typically more expensive than buying.
- What happens when something goes “viral”?
- It can take 6 weeks to get new servers deployed to the datacenter. In an emergency we can get everything ready in 2.
- Virtualization advantages:Lower operational costsResource managementThis can also apply in a global environment, but you probably want to get your data closer to the edge for a good experience.
- MarketplaceTrusted source to find high-quality, accurate, and authoritative contentQuickly understand data formats and query capabilities with the Service ExplorerChoose from flexible purchasing options and begin using data immediatelyData accessStandard protocol for accessing data through DallasFor developers, interoperable and simple to consume (RESTful protocol with an XML representation for the data)For end users, OData can be consumed by a growing list of products(we’ll show the Excel integration later)BillingProvider users choice on how they pay (subscriptions, per transaction)For V1, payments with credit cards, but provide choice over time