A kind of memory database engine based on column-storage techniques

摘要

Big data query requests higher performance in cloud computing environment, but traditional relational DBMS can not achieve high query speed when processing large volume data, so it is a trend to apply column-storage techniques to process big data query in the field of data management. But deploying row-storage and column-storage databases simultaneously for OL TP and OLAP applications makes heavy burdens to enterprises. To enhance the query performance on big data and do not deploy column-storage databases, this paper proposed a database engine which is based on column-storage and memory techniques. It can rapidly load big data from row-storage databases and return query result in a short time. Experiments show that this engine is more efficient than the typical column-storage database MonetDB when querying big data.