Query & Analyze
RainStor provides fast response to multiple query types, often on par or faster than a relational database. The fast query response time is due to the sophisticated algorithms that learnĀ the patterns in the data and store them as single instances. When RainStor receives a query, it parses and processes the query to identify the subset of patterns that contain the query result. The query is then run over this reduced set of data, realizing huge benefits in performance.
SQL-92, ODBC/JDBC Access
Data in RainStor can be accessed at any time for retrieval using all versions of SQL Query including SQL language extensions from popular databases such as Oracle and SQL Server. The data is also available for analysis with a number of business intelligence and analytics tools including SAP Business Objects, IBM COGNOS and MicroStrategy among others.
MapReduce and Pig Analysis
Once data is loaded into RainStor running on Apache Hadoop, it can be accessed using MapReduce jobs. Additionally, RainStor provides full support for Pig. RainStor’s compression becomes a multiplier and MapReduce jobs run much faster because they are reading fewer partitions or files. Additionally, Hadoop users can flexibly use either SQL or MapReduce Pig depending on the type of query and the desired result. Having access to both query and analytical tools improvers overall productivity for the IT team and ultimately the enterprise.
Cross Schema Query
RainStor’s versioned approach to schema change ensures that queries written against the latest or current schema are able to access data previously stored under multiple historical schemas in a consistent manner. This is especially useful for data sets that need to be retained for long time periods and also for data that is ingested from different source database environments where schemas are inevitably different and will change over time.
Point-in-time History View
RainStor also supports query and export of data as it was at a selected historical point in time. Users can set their logical view of data to comprise an exact view of what would have been visible had they run the query at that prior point in time. Tables or columns created since the specified time are masked from view, while tables or columns dropped since then are exposed and remain visible.
