際際滷

際際滷Share a Scribd company logo
Lightblue: Re-thinking the old way
Lessons of How to turn internal Fog into a simple Cloud
DISCLAMER
 The views expressed in the following slides are solely those of
the author, which aims just to share knowledge and help the
community
 Putting this in another way, the contents of this presentation do
not necessarily represent the views or policies of anyone else
than the author
Introduction
 Luan Cestari
 Software Engineer
 @BR_LuanCestari and http://www.slideshare.net/LuanCestari
 You can interrupt me to ask me, I'm here to help =)
Agenda
 Quick story: from small first application in a company to a big
interconnected services
 Some problems that may occur
 The Lightblue project that aims to fix those problems
 Some details of the techonologies behind Lightblue
 Bonus: Some interressting collections libraries
The Beginning
The Application Delivery
The Company and the systems grow
A Simple Change Involves a Lot of Systems
The Result of Change
What are the patterns that you may find?
 Very hard to maintain services
 Services which too much specific and compled
 Slowness
 Hard coded
 Static content
 Hard to garantee the security end to end
 Hard to deploy in different enviroment
What some people thought
 To re-write the services creating an open source solution
 Using cutting edge tecnologies, good practices and techniques
 Make it simple to use and maintain
 Make it fast but it also needs to have monitoring
 Need to be flexible for changes
 Make use a good techonologies which can provide some of the Cloud
characteristics (Measured service,Rapid elasticity,Resource
pooling,Broad network access,On-demand self-service)
 Need to be very secure
This is Ligthblue
Lightblue
 It aims to be capable to provide high availability, resiliency,
consistency, and performance integrating with others SOA
Services
 To provide a high level of flexibility, it is metadata driven project
 You can easily extend and change its behavior
 You can use it for your own business
Lightblue: Overview
Lightblue: Simple Deployment
Lightblue: Deployed on the Cloud
Features: Hystrix
https://github.com/Netflix/Hystrix
Features: Hystrix
 Hystrix: Latency and Fault Tolerance for Distributed Systems
 Can be used with Servo for monitoring (Graphite and CloudWatch)
 https://github.com/Netflix/servo
 Is there overhead?
 https://github.com/Netflix/Hystrix/wiki/FAQ#what-is-the-processing-overhea
Features: Hystrix
https://github.com/Netflix/Hystrix/wiki/Dashboard
Features: MongoDB
 https://en.wikipedia.org/wiki/MongoDB#Main_features
 http://docs.mongodb.org/manual/core/replication-introduction/
Lightblue: more information
 https://github.com/lightblue-platform/lightblue
 https://github.com/lightblue-platform/lightblue/wiki
 https://github.com/lightblue-platform/lightblue-quickstarts
Bonus: Beyond JDK Collections
 https://github.com/goldmansachs/gs-collections
 http://trove4j.sourceforge.net/index.html
 http://high-scale-lib.sourceforge.net/
 http://javolution.org/
 https://code.google.com/p/guava-libraries/
 http://www.goldmansachs.com/gs-collections/presentations/GSC_Memory_Te
 https://commons.apache.org/
 colt
Bonus: Beyond Others
 https://github.com/OpenHFT/OpenHFT
 https://github.com/peter-lawrey/Java-Chronicle
 https://lmax-exchange.github.io/disruptor/
 https://sites.google.com/site/piotrwendykier/software/parallelcolt
 https://code.google.com/p/thrift-protobuf-compare/wiki/Benchmarking
 https://github.com/Netflix/RxJava/wiki
 https://bitbucket.org/robeden/trove/
Conclusion
 There are a lot of other tecnologies other than Java EE which
are very good to use
 Migrating to Lightblue can help you with some hard to maintain
services
Thank you!
 Thank you for watching and sharing your ideas
 @BR_LuanCestari and http://www.slideshare.net/LuanCestari
Any questions?

More Related Content

Lightblue project

  • 1. Lightblue: Re-thinking the old way Lessons of How to turn internal Fog into a simple Cloud
  • 2. DISCLAMER The views expressed in the following slides are solely those of the author, which aims just to share knowledge and help the community Putting this in another way, the contents of this presentation do not necessarily represent the views or policies of anyone else than the author
  • 3. Introduction Luan Cestari Software Engineer @BR_LuanCestari and http://www.slideshare.net/LuanCestari You can interrupt me to ask me, I'm here to help =)
  • 4. Agenda Quick story: from small first application in a company to a big interconnected services Some problems that may occur The Lightblue project that aims to fix those problems Some details of the techonologies behind Lightblue Bonus: Some interressting collections libraries
  • 7. The Company and the systems grow
  • 8. A Simple Change Involves a Lot of Systems
  • 9. The Result of Change
  • 10. What are the patterns that you may find? Very hard to maintain services Services which too much specific and compled Slowness Hard coded Static content Hard to garantee the security end to end Hard to deploy in different enviroment
  • 11. What some people thought To re-write the services creating an open source solution Using cutting edge tecnologies, good practices and techniques Make it simple to use and maintain Make it fast but it also needs to have monitoring Need to be flexible for changes Make use a good techonologies which can provide some of the Cloud characteristics (Measured service,Rapid elasticity,Resource pooling,Broad network access,On-demand self-service) Need to be very secure
  • 13. Lightblue It aims to be capable to provide high availability, resiliency, consistency, and performance integrating with others SOA Services To provide a high level of flexibility, it is metadata driven project You can easily extend and change its behavior You can use it for your own business
  • 18. Features: Hystrix Hystrix: Latency and Fault Tolerance for Distributed Systems Can be used with Servo for monitoring (Graphite and CloudWatch) https://github.com/Netflix/servo Is there overhead? https://github.com/Netflix/Hystrix/wiki/FAQ#what-is-the-processing-overhea
  • 20. Features: MongoDB https://en.wikipedia.org/wiki/MongoDB#Main_features http://docs.mongodb.org/manual/core/replication-introduction/
  • 21. Lightblue: more information https://github.com/lightblue-platform/lightblue https://github.com/lightblue-platform/lightblue/wiki https://github.com/lightblue-platform/lightblue-quickstarts
  • 22. Bonus: Beyond JDK Collections https://github.com/goldmansachs/gs-collections http://trove4j.sourceforge.net/index.html http://high-scale-lib.sourceforge.net/ http://javolution.org/ https://code.google.com/p/guava-libraries/ http://www.goldmansachs.com/gs-collections/presentations/GSC_Memory_Te https://commons.apache.org/ colt
  • 23. Bonus: Beyond Others https://github.com/OpenHFT/OpenHFT https://github.com/peter-lawrey/Java-Chronicle https://lmax-exchange.github.io/disruptor/ https://sites.google.com/site/piotrwendykier/software/parallelcolt https://code.google.com/p/thrift-protobuf-compare/wiki/Benchmarking https://github.com/Netflix/RxJava/wiki https://bitbucket.org/robeden/trove/
  • 24. Conclusion There are a lot of other tecnologies other than Java EE which are very good to use Migrating to Lightblue can help you with some hard to maintain services
  • 25. Thank you! Thank you for watching and sharing your ideas @BR_LuanCestari and http://www.slideshare.net/LuanCestari Any questions?