You can seamlessly deploy OneData in clustered and other distributed environments using the OneData distributed caching framework. In such environments, different OneData instances will run on different JVMs (and possibly different physical servers). To ensure that metadata changes made on one node are reflected across all other nodes, caches of all nodes must be synchronized. Caching framework performs automatic synchronization of caches of all OneData nodes in a distributed environment. This improves performance by “borrowing” already cached metadata from the cache server.

The Technology Cache server is embedded within OneData. You can start the cache server from the user interface. OneData embeds the cache server, which can be started in the user interface or from the command prompt. The server runs as an independent process and uses Java RMI for communications between nodes. It is possible to configure failover cache servers to protect against primary cache server going down.

The caching framework is built on the Java Caching System (JCS) caching library and integrates tightly with OneData. The cache server is configured using OneData property files. Temporary configuration files and libraries that the cache server needs are created dynamically on OneData start-up in <ONEDATA_HOME>/temp/remotecache directory. The cache server requires JRE version 1.5 or later.