tpch-kit

TPC-H benchmark kit with some modifications/additions

Official TPC-H benchmark - http://www.tpc.org/tpch

Modifications

The following modifications have been added on top of the official TPC-H kit:

• modify dbgen to not print trailing delimiter
• add option for dbgen to output to stdout
• add compile support for macOS
• add define for PostgreSQL to support LIMIT N for qgen
• adjust Makefile defaults

Setup

Linux

Make sure the required development tools are installed:

Ubuntu:

sudo apt-get install git make gcc

CentOS/RHEL:

sudo yum install git make gcc

Then run the following commands to clone the repo and build the tools:

git clone https://github.com/gregrahn/tpch-kit.git
cd tpch-kit/dbgen
make MACHINE=LINUX DATABASE=POSTGRESQL

macOS

Make sure the required development tools are installed:

xcode-select --install

Then run the following commands to clone the repo and build the tools:

git clone https://github.com/gregrahn/tpch-kit.git
cd tpch-kit/dbgen
make MACHINE=MACOS DATABASE=POSTGRESQL

Using the TPC-H tools

Environment

Set these env variables correctly:

export DSS_CONFIG=/.../tpch-kit/dbgen
export DSS_QUERY=$DSS_CONFIG/queries
export DSS_PATH=/path-to-dir-for-output-files

SQL dialect

See Makefile for the valid DATABASE values. Details for each dialect can be found in tpcd.h. Adjust the query templates in tpch-kit/dbgen/queries as need be.

Data generation

Data generation is done via dbgen. See dbgen -h for all options. The environment variable DSS_PATH can be used to set the desired output location.

Query generation

Query generation is done via qgen. See qgen -h for all options.

The following command can be used to generate all 22 queries in numerical order for the 1GB scale factor (-s 1) using the default substitution variables (-d).

qgen -v -c -d -s 1 > tpch-stream.sql

To generate one query per file for SF 3000 (3TB) use:

for ((i=1;i<=22;i++)); do
  ./qgen -v -c -s 3000 ${i} > /tmp/sf3000/tpch-q${i}.sql
done