BenchMark

Prepare Table Schema

Look at json files in 'benchmark/testdata' and put your table schema into json files. The benchmark script also provides functionality to generate data. The following data types can be generated:

int:

• random. default -100~100
• random_range. "range": [min, max]
• increment. same as the auto-increment in MySQL
• random_pool. choose rand value in "pool"

float:

• random. default -100.0~100.0
• random_range. "range": [min, max]
• random_pool. choose rand value in "pool"

string:

• random_pool. choose rand value in "pool"
• random. same as random_pool but choose rand value in a default pool
• increment. same as the auto-increment in MySQL

Run Benchmark

put your query into 'benchmark/testdata/query.json' and run the script:

cd benchmark
make

then you can see logs and op cost in 'benchmark/log', or you may make plot to create png file.