|Recommended Data Types
|Used by default. Extremely fast; good compression; balanced speed and efficiency
|Good compression; pretty fast; best for high compression needs. Don’t use levels highter than 3.
|LZ4 High Compression algorithm with configurable level; slower but better compression than LZ4, but decmpression is still fast.
|Integer Types, Time Series Data, Timestamps
|Preprocessor (should be followed by some compression codec). Stores difference between neighboring values; good for monotonically increasing data.
|Integer Types, Time Series Data
|Stores difference between neighboring delta values; suitable for time series data
|Floating Point Types
|Calculates XOR between current and previous value; suitable for slowly changing numbers
|Integer, Time Series Data, Timestamps
|Preprocessor (should be followed by some compression codec). Crops unused high bits; puts them into a 64x64 bit matrix; optimized for 64-bit data types
|Preprocessor (should be followed by some compression codec). Greatest common divisor compression; divides values by a common divisor; effective for divisible integer sequences
|Floating Point Numbers
|Designed for Float64; Algorithm detailed in FPC paper, ClickHouse PR #37553
|Requires hardware support for QuickAssist Technology (QAT) hardware; provides accelerated compression tasks
|Requires hardware support for Intel’s QuickAssist Technology for DEFLATE compression; enhanced performance for specific hardware
|It’s not a codec, but a datatype modifier. Reduces representation size; effective for columns with low cardinality
|Non-compressable data with very high entropy, like some random string, or some AggregateFunction states
|No compression at all. Can be used on the columns that can not be compressed anyway.
Last modified 2024.01.27: Update _index.md (1b0a47d)