Skip to content
Fabio Simeoni edited this page Aug 19, 2014 · 5 revisions

Memory beans implement the Bean SPIs without an underlying persistent store.

For each xxx.Bean SPI there is a corresponding Mxxx memory bean. For example, MIdentified is the memory bean for Identified.Bean and MCodelist is the memory bean for Codelist.Bean.

Since they do not incur the setup and runtime penalties of persistence, we use them as "bean dummies" in:

  • demo installations for increased performance.
  • during development and testing for faster feedback.

Memory beans are not dummies, however. They continue to play a crucial rule even in combination with a persistent bean family. In particular we use them:

All the entities created through the data DSL are based on memory beans. As a result, memory beans are responsible for initialising entities and changesets. This includes validation and cloning, which is the basis for versioning.

Clone this wiki locally