SlideShare uses cookies to improve functionality and performance, and to provide you with relevant advertising. If you continue browsing the site, you agree to the use of cookies on this website. See our User Agreement and Privacy Policy.
SlideShare uses cookies to improve functionality and performance, and to provide you with relevant advertising. If you continue browsing the site, you agree to the use of cookies on this website. See our Privacy Policy and User Agreement for details.
Successfully reported this slideshow.
Activate your 14 day free trial to unlock unlimited reading.
What kind of design patterns are useful for applications adopting the cloud? How can apps achieve the scalability and availability promised by the cloud? Presentation from Interop 2011 Enterprise Cloud Summit.
Accelerating Software Innovation and Business Growth
What kind of design patterns are useful for applications adopting the cloud? How can apps achieve the scalability and availability promised by the cloud? Presentation from Interop 2011 Enterprise Cloud Summit.
1.
Ten CloudDesign Patterns<br />Shlomo SwidlerFounderOrchestratus<br />
2.
Shlomo Swidler<br />Founder, Orchestratus<br />Strategic and technicalIT consulting<br />Customers include:<br />Cloud Developer Tips bloghttp://shlomoswidler.com/<br />Among top community-ranked contributors to Amazon Web Services discussion forums<br />2<br />
3.
Ten CloudDesign Patterns<br />Shlomo SwidlerFounderOrchestratus<br />
4.
Ten Cloud ApplicationDesign Patterns<br />Shlomo SwidlerFounderOrchestratus<br />
5.
Ten Cloud ApplicationDesign Patterns<br />Many<br />Shlomo SwidlerFounderOrchestratus<br />
6.
What is a Design Pattern<br />A reusable recipe for building (software) systems that solve a particular problem.<br />6<br />
7.
What is a Design Pattern<br />A reusable recipe for building (software) systems that solve a particular problem.<br />AKA Architectural Pattern<br />7<br />
8.
What is a Design Pattern<br />A reusable recipe for building (software) systems that solve a particular problem.<br />Goal<br />AvailableResources<br />Constraints<br />8<br />
9.
A Design Pattern<br />A reusable recipe for building (software) systems that solve a particular problem.<br />Meets affirmativerequirements<br />Goal<br />Does not violatenegative requirements<br />AvailableResources<br />Constraints<br />Can be implemented<br />9<br />
10.
Challenges Faced by Apps in the Cloud<br />Application Scalability<br />Cloud promises rapid (de)provisioning of resources.<br />How do you tap into that to create scalable systems?<br />Application Availability<br />Underlying resource failures happen … usually more frequently than in traditional data centers.<br />How do you overcome that to create highly available systems?<br />10<br />
11.
The Scalability Challenge<br />Scalability: Handle more (or fewer) requests<br />It’s not Performance (handle requests faster)<br />It’s not Availability (tolerate failures)<br />But improving Scalability often improves Availability<br />11<br />
12.
The Scalability Challenge<br />Two different components to scale:<br />State (inputs, data store, output)<br />Behavior (business logic)<br />Any non-trivial application has both.<br />Scaling one component means scaling the other, too.<br />12<br />
15.
The Availability Challenge<br />Availability: Tolerate failures<br />Traditional IT focuses on increasing MTTF<br />Mean Time to Failure<br />Cloud IT focuses on reducing MTTR<br />Mean Time to Recovery<br />15<br />
16.
The Availability Challenge<br />Availability: Tolerate failures<br />Traditional IT focuses on increasing MTTF<br />Mean Time to Failure<br />Cloud IT focuses on reducing MTTR<br />Mean Time to Recovery<br />What follows is four availability scenarios: [low, high] X [MTTF, MTTR]<br />16<br />
32.
Spot the SPOF:2<br />32<br />Internet<br />Elastic IP Address<br />Cloud<br />App<br />App<br />App Instance<br />App Instance<br />Fail-over<br />
33.
Spot the SPOF:2<br />33<br />Internet<br />Might work…<br />Until you need more App instances<br />Or until another SPOF fails…<br />Elastic IP Address<br />Cloud<br />App<br />App<br />App Instance<br />App Instance<br />Fail-over<br />
59.
Availability Best Practice:Chaos Monkey<br />AKA Error Injection Testing<br />Forcibly create fault conditions in your cloud components.<br />Kill instances, detach disks, screw up DNS, etc.<br />Automate recovery from the errors.<br />The team gets really good at reducing MTTR, increasing availability!<br />Popularized by Netflix, who run it on their live environment.<br />59<br />
60.
For more on Designing forAvailability, Scalability<br />Jonas BonérScalability, Availability, Stability Patterns http://slidesha.re/cK3NJv<br />George ReeseThe AWS Outage: The Cloud’s Shining Momenthttp://oreil.ly/eKCGG9<br />John Ciancutti of Netflix5 Lessons We’ve Learned Using AWShttp://bit.ly/h8rU8b<br />60<br />