As usual, we covered a lot of ground in this chapter. Although we dealt mostly with concepts and patterns that apply to any IMDG, we saw plenty of WebSphere eXtreme Scale example code.
We started the chapter with XTP. XTP systems require high-performance, high-scalability, and strong manageability. In the past years, we've had middleware that supply all three to varying degrees of success. IMDGs are new tools used to build XTP-class software due to their leaps forward in scalability and manageability. Co-locating data and application logic in the same JVM offers huge performance increase over previous generations of tools.
A common requirement when using IMDGs is a reduction of network hops. Accessing data stored exclusively in different partitions is a scalability limitation when using a grid. We duplicate data where it is used, instead of fetching it every time. We saw in depth how to duplicate objects in each partition when needed.
Staying on the data duplication theme, we discussed...