Code based on skeleton from Harvard CS165.
This should be filled with a workflow.
Start by giving (you may have to run automake twice):
automake --add-missing
autoreconf
./configure
Then there you can compile:
make
To create a dataset and workload for a test:
cd testtools
and edit gen.cfg file to customize your test
./dataGen.py
./dataGen_dbinfo.py
./workGen.py
then move the b_* folder and the workload file into the data folder for tests
The workload generated is in ascii format, for partition decision function to work, you need to genereate parsed past workload through the db server, function doSomething in file src/utils/utils.c does this job.
To partition the dataset currently in database according to a specific workload
feed the client binary:
partition(<col_var>, "workload_num")
To run the partition decision algorithm only, feed the client binary:
partition_decision(<col_var>, "workload_num")
To execute many workload commands in one command, feed the client binary:
exec_work("relativeworkloadfilePATH")
To print the results for partitioned table, feed client binary:
show_tbl(<col_var>)
More db operations can be found under folder src/database/dsl.c