Cacheonix distributes cached data across a cluster of computers connected by a high-speed network and converts the per-node memory into a huge cache of the shared address space. A cache coherence protocol ensures that all nodes have a consistent view of the shared data.
To see Cacheonix in action, download Cacheonix or check out user guide and code examples.
At the maximum, a single network operation is required to retrieve cached data, without regard to its location on the network. This allows Cacheonix to provide extremely high throughput and very low latency on a standard local area network.
Each Cacheonix node maintains an exceptionally fast coherent front cache that provides memory-like speed when accessing remote data by capturing information access patterns.
Cacheonix provides uninterrupted access to the cache data by replicating the data in a cluster of servers and by seamlessly restoring the data if any of the servers fails. Cacheonix architecture allows to scale enterprise applications horizontally by evenly partitioning cache data so that servers carry a fair share of load:
Cacheonix uses multicast and TCP sockets for cluster membership maintenance and for inter-node communications. An advanced clustering protocol provides wire-level speeds when exchanging information between the cluster nodes.