solidDB^® enables compressing the data in the database to save disk capacity, memory capacity and disk I/O. Compression is not automatically turned on since it might have some undesirable side effects or for example behave badly on data that is already compressed. This section explains the underlying technology of solidDB^® compression, how to operate it and how to analyze whether it can be expected to be useful in particular circumstances.

Typical key reasons for compressing data in databases are related to capability of reducing requirements of disk space and memory or processing higher logical data volumes in same capacity. As a side impact, disk I/O is also reduced due to processing compressed blocks instead of uncompressed blocks. All these benefits are applicable in solidDB^®.

In solidDB^® compression, the database files are compressed for disk-based tables only. Both data tables and indexes for them are compressed. The storage for BLOBs is not compressed even though the blobs are referred from disk-based system tables. The BLOB directory in system tables can be compressed but the actual BLOB content can’t. Memory image is not compressed as such, but due to processing compressed tree nodes in memory more data will fit in cache or less memory is required for similar cache hit rate. In-memory data inside solidDB^® is not compressed.

Transaction log files are not compressed. The disk capacity benefits are related to database files only.

The compression works with both solidDB^® replication mechanisms HotStandby and Advanced Replication. Also, all linking models (Client/Server, LLA and SMA) are supported as are all Client Drivers.