We use DuckDB to generate TPC-DS data.
After installing DuckDB, you can use these commands to generate the data (more information):
INSTALL tpcds;
LOAD tpcds;
SELECT * FROM dsdgen(sf=0.01) -- sf can be other values, such as 0.1, 1, 10, ...
EXPORT DATABASE '/tmp/tpcds_0_01/' (FORMAT CSV, DELIMITER '|');
Then, move the data to current directory:
mv /tmp/tpcds_0_01/ "$(pwd)/data/"
After that, you can load data to Databend:
./load_data.sh
To run the TPC-DS Benchmark, first build databend-sqllogictests
binary.
Then, execute the following command in your shell:
databend-sqllogictests --handlers mysql --database tpcds --run_dir tpcds --bench