Date Added: Jun 2012
Compression in traditional database systems is known to improve performance significantly, it reduces the size of the data and improves I/O performance by reducing seek times (the data are stored nearer to each other), reducing transfer times (there is less data to transfer), and increasing buffer hit rate (a larger fraction of the DBMS fits in buffer pool). For queries that are I/O limited, the CPU overhead of decompression is often compensated for by the I/O improvements. The authors revisit this literature on compression in the context of column-oriented database systems. Storing data in columns presents a number of opportunities for improved performance from compression algorithms when compared to row-oriented architectures.