You can save yourself from huge pain of maintaining data consistency between two different storages. Take a look at Apache Ignite data grid. Using it you can get several benefits:
transparent caching
access data as key-value store
access same data as relational DB with distributed SQL queries
full transaction support
close to linear horizontal scalability for both storage and performance
tunable redundancy and consistency characteristics for every "table" ("cache" in Ignite terms)
It also provides additional services like queues, pub-sub, distributed computing, built-in file system and many more useful things.
You can save yourself from huge pain of maintaining data consistency between two different storages. Take a look at Apache Ignite data grid. Using it you can get several benefits:
It also provides additional services like queues, pub-sub, distributed computing, built-in file system and many more useful things.
Very cool, I've never looked at that before! I'll certainly consider it as an option.