Scale

RainStor provides unique scale in terms of both data volume ingestion rates but more importantly scale as volumes grow over time to accommodate future growth rates which is especially important as compliance regulations become more stringent.

RainStor’s efficient data ingestion of thousands of records per second when it splits the file into chunks of say one million records which are then processed in parallel, streamed in acros a number of CPU’s or servers at the same time which ultimately provides ingestion rates typically 10x faster than a traditional RDBMS.

As data volumes grow to petabyte scale, RainStor relies on it’s massive compression and de-duplication to do far more with less and essentially use this as a hardware multiplier utilizing smaller amount of storage, less CPU and less network I/O operations and whereby the underlying hardware and storage platforms can be easily extended to meet additional demand.

By leveraging a Massively Parallel Processing (MPP) architecture and shared-nothing approach at the server layer, RainStor scales across servers and cores during data import and query. Because RainStor stores data in partitions, which are separately stored as physical containers and files, every node within a multi-nodal RainStor deployment is able to see and access data across all partitions, and data can reside on low-cost commodity local disk. RainStor has also been designed to run on the cloud, supporting a variety of chip architectures, to deliver low-cost scalability without the need for specialized hardware or skills.

RainStor supports SAN, NAS, DAS, CAS, or cloud storage platforms, and additional storage can be provisioned on demand using existing management software. RainStor can literally port thousands of tables, thousands of archives and hundreds of customers all at the same time on one system through a multi-tenancy support which is especially attractive for cloud deployments.