The GridGain® in-memory computing platform is inserted between existing application and data layers. Deployed as a cluster, the GridGain distributed computing solution delivers database acceleration for any RDBMS, NoSQL or Hadoop database including popular solutions such as MySQL, PostgreSQL and Hadoop.
When using the in-memory data grid capabilities of GridGain, your entire dataset is uploaded from your disk-based database into RAM. Your application will then communicate directly with GridGain and transactions will be processed in-memory and then written to your underlying database with ACID transaction guarantees. ANSI-99 SQL support ensures that queries can also easily be run against the in-memory version of your data set.
Because your data is held in memory, data does not need to be retrieved from disk before processing, a very time consuming step. Performance can be increased by 1,000x or more. Further database acceleration is accomplished via parallel processing of tasks across the nodes in the GridGain cluster.
GridGain also allows you to scale out your database. If your database begins to slow down because it has become large relative to your database server, GridGain can allow you to seamlessly scale out your database by adding nodes to your in-memory computing cluster. GridGain automatically shards your database across all of the nodes in the cluster, providing high availability as well as database acceleration.