Autocompression Interactions With User‑Specified Compression Methods
Teradata Database applies user-specified compression for columns within a multicolumn column partition value, but Teradata Database applies autocompression to a column partition value as a whole.
Teradata Database might decide not to use one or more of the user‑specified compression techniques with autocompression if it determines that other autocompression techniques, including none, provide better compression for a container.
When the system constructs a container, it first applies any user-specified compression. When the container reaches a certain size limit, which is defined internally, Teradata Database examines the container to determine what autocompression techniques and user‑specified compression can be used to reduce its size. After compressing the container, the system can append additional column partition values to it using the method of autocompression it has determined to be optimal and any user‑specified techniques until a container size limit is reached, at which time Teradata Database constructs a new container.
As an alternative to the previous method, the system may decide that it is more efficient to reuse the compression from the previous container and apply that compression when constructing a container.
If you specify block-level compression, which is not an autocompression technique, for a column‑partitioned table or join index, it is applied for data blocks independently of whether Teradata Database applies autocompression.