Compilation-based execution engine for a database
- Instructions will be for SF 1 but same holds for other scale factors.
- Run dbgen to generate .tbl files.
- For TPC-H, use https://github.com/electrum/tpch-dbgen.
- For JCC-H, use https://github.com/ldbc/dbgen.JCC-H.
- Move them to
benchmark/tpch1/raw/
. - Run
bazel build //benchmark/tpch1:load && bazel-bin/benchmark/tpch1/load
. - Run
python benchmark/tpch1/monetdb_load.py
. - Run
python benchmark/tpch1/duckdb_load.py
.
To bench, run python benchmark/tpch1/bench.py > tpch1.csv
.
To bench, run python benchmark/tpch1/monetdb_bench.py > monetdb_tpch1.csv
.
To bench, run python benchmark/tpch1/duckdb_bench.py > duckdb_tpch1.csv
.
- Download CSV Files from JOB here.
- Make the following changes to the CSV files:
- Update line 268268 of
person_info.csv
to escape the\
inCanada\US
- Update line 2671662 of
person_info.csv
to escape the\
in(qv); \,
- Update line 2514451 of
title.csv
to escape the\
in\Frag'ile\
- Run
python benchmark/job/monetdb_load.py
. - Run
python benchmark/job/duckdb_load.py
.
To bench, run python benchmark/job/monetdb_bench.py > monetdb_job.csv
.
To bench, run python benchmark/job/duckdb_bench.py > duckdb_job.csv
.