Data consistency: Analyse, understand and decide
Conference
Architecture, Performance and Security
|
|
Room 2 |
Thursday at 13:20 - 14:10 |
Most applications handle some form of data. And quickly you end up duplicating this data … whether going from a JPA entity to a DTO, offering a caching layer on top of your services or even because the database is replicated and clustered. As soon as there is duplication, there is a risk of inconsistency. How long is my DTO meaningful? How do I invalidate my cache? What happens if asynchronous data replication fails? Answering these questions quickly becomes extremely complicated and most often requires a final decision be made at the application level. By illustrating these concepts using the resilience strategy developed for Ehcache 3, we will see that even for libraries targeted at handling data consistency in the face of failure, only the application may know the answer. Handling these issues from day one, during design and architecture, is a major concern. And after all, isn’t the easy path quite boring? |
|
Louis Jacomet
|
|---|
Louis Jacomet has been perfecting his knowledge of Java for more than 10 years. Always a developer at heart, his role evolved over the last years to include technical team leading, coaching and some management. In addition to the bits and bytes, Louis is interested in people management skills mandatory to create a productive project team. To complete the buzzword bingo, Louis is interested in agile practices for the visibility, communication and result orientation it promotes. Louis works remotely from Belgium for Terracotta, with a focus on Ehcache. |