The advent of BigMemory two years ago allowed for 1000x per-node jumps in scale for in-memory management. Instead of 2GB per node, it is now possible to manage 2TB per node. At the same time, BigMemory introduced lower latencies that were more predictable.
Since then, BigMemory has continued to evolved with search capabilities and on-heap bytes based tuning (ARC or automated resource control).
Further, to protect that data in-memory, they added a fault tolerant restartable store (FRS). In addition, to help secure monitoring and management, they created the Terracotta Management Console (TMC).
To engage users, Terracotta released BigMemory Go, which offers 32GB of single-node in-memory management of data – absolutely free. This service provides the ability to track and test the value-add of BigMemory Go.
Check it out, and the incredible developer community over at Terracotta.