Dan Stone "Scalabale Application Frameworks"
1 like376 views
The document compares two frameworks for building scalable applications: Gigaspaces and Terracotta. It highlights their functional similarities and differences, such as language support, learning curve, and configuration requirements. The document emphasizes the benefits of using these frameworks in cloud environments, including cost reduction and improved application performance.
![Done! (with seconds to spare) [email_address]](https://image.slidesharecdn.com/3-danscalabaleapplicationframeworks3-090813125410-phpapp01/85/Dan-Stone-Scalabale-Application-Frameworks-9-320.jpg)
Dan Stone "Scalabale Application Frameworks"
- 1. Frameworks for Scalable Applications
- 2. Top Five Reasons for Moving to Cloud Finance Director Convert fixed costs to variable Shrink balance sheet Capture economies of scale Reduce overall cost Looks good on CV IT Professional More flexibility Quicker delivery Improve testing Looks great on CV Build scalable applications
- 3. GigaSpaces and Terracotta Frameworks for building scalable applications Have functional similarities support spreading data across many nodes can bring data and code together (locality) have built-in high availability features work with Spring and Hibernate Memory-centric models But they are as different as chalk and cheese!
- 4. GigaSpaces Based on the JavaSpaces specification Does much more Partitioned spaces Transparent fail-over Cloud deployment and management framework Sync/async replication Automatic routing Virtual services Distributed tasks .net support
- 5. GigaSpaces Big Ideas Data written to in-memory spaces Spaces partitioned across many hosts Content-based routing delivers data access calls to correct partition Can address individual partitions and/or entire space Space-based Architecture (SBA) – Deploy entire application on each partition and share nothing
- 6. Terracotta Java only Shares data and events between JVMs No API's or programming changes Central server(s)-based Looks like a one trick pony. Don't be deceived! X
- 7. Terracotta Big Ideas Data, event, and lock sharing between JVMs and hosts Virtualized heap – handle big data Real-time performance analysis dashboard High availability – automatic server fail-over High reliability – can persist to disk Integration modules – share Spring, Hibernate, etc. objects at run-time with minimal fuss
- 8. Quick Comparison GigaSpaces Terracotta Language support Java, .net Java What you have to learn API, configuration configuration What you have to know Spring, high level concurrency Low level concurrency (for primitives) Support for locality ubiquitous application logic Learning curve Steep Steep (for primitives) Distribution Part open source, part proprietary Open source
- 9. Done! (with seconds to spare) [email_address]