Compression algorithms (dictionary, RLE, sparse, and so on) applied to a table store column.
The compression are only on the main storage
Default - Dictionary compression
Default is the Dictionary compression
- applied to all columns
- mapping of distinct column values to consecutive numbers. More distinct value, more compression.
Others - Advanced compression
Each column can be further compressed using different compression methods:
- prefix encoding (PREFIXED),
- run length encoding (RLE),
- cluster encoding (CLUSTERED),
- sparse encoding (SPARSE),
- and indirect encoding (INDIRECT).
Columns with the PAGE LOADABLE attribute are compressed with the NBit algorithm only.
SELECT SCHEMA_NAME, TABLE_NAME, COLUMN_NAME, COMPRESSION_TYPE, LOADED FROM PUBLIC.M_CS_COLUMNS WHERE SCHEMA_NAME='<your_schema>' and TABLE_NAME='<your_table>'
The columns of the selected table are listed with the type of compression applied. The following values are possible: