TPC-H Multi-tenant dbgen. Base on the work from here
Comparing with the original TPC-H dbgen verison, the tpch-mt version extended with serveral MT features and is designed for Multi-Tenancy DB benchmark.
dbgen.java
is the entry point of tpch-mt application. The output is controlled by a combination of command line options
and environment variables. Command line options are assumed to be single letter flags preceded by a minus sign. They may be followed by an optional argument.
-h -- display this message
-s <arg> -- set Scale Factor (SF) to <n> (default: 1)
-t <arg> -- set Number of Tenants to <n> (default: 1)
-m <arg> -- set distribution mode to <mode> (default: uniform, others:zipf)
-T <arg> -- generate tables
Sample usage: java dbgen -s 1 -t 10 -m uniform
.
tpch-mt dbgen will generate 6 separate ascii files under the output
folder. This is not as same as original TPCH dbgen. We eliminate the PART and PARTSUPP table. Each file will contain pipe-delimited load data for one of the tables defined in the TPC-H database schema. The default tables will contain the load data required for a scale factor 1 database. By default the file will be created in the current directory and be named <table>.tbl
. As an example, customer.tbl will contain the load data for the customer table.