eloq_test

Before running the test

./setup

If you have updated anything in test_lib, reinstall test_lib by running

sudo python3 setup.py install

To run the test

1. Run test cases by single file

python3 single_test/smoke_test.py --dbtype mariadb --install_path $your_install_path$

2. Run a particular test method in a test file

python3 single_test/smoke_test.py --dbtype mariadb --install_path $your_install_path$ --method $method_name

3. Run all tests together

python3 run_tests.py --dbtype mariadb --install_path $your_install_path$

There are other optional flags you can set(except the single flag, which the scripts will take care of). Check them out by using the -h flag:

ubuntu@xxx:~/eloq_test$ /bin/python3 /home/ubuntu/eloq_test/run_tests.py -h
usage: run_tests.py [-h] --dbtype --install_path INSTALL_PATH [--bootstrap {true,false}] [--repeat REPEAT] [--method METHOD] [--storage {cassandra,dynamo}] [--force {true,false}] [--group {single,multiple}]

optional arguments:
  -h, --help            show this help message and exit
  --install_path INSTALL_PATH
                        Eloq install path
  --dbtype {mariadb,redis}
                        Specify the database engine
  --bootstrap {true,false}
                        (User should not set this flag!! The scripts will take care of this flag) If this test runs singly or by run_tests. In the latter case, run_tests will set this flag to false, meaning single tests don't need to bootstrap themselves.
  --repeat REPEAT       Number of times to repeat the test, cannot be smaller than 1.
  --method METHOD       Specify the paticular test method to run
  --storage {cassandra,dynamo}
                        Specify the storage engine
  --force {true,false}  Continue to run tests if one test fails.
  --group {single,multiple}
                        For run_tests only. Specify running single-node tests or multi-nodes tests. If this arg is absent, run both(which is the default behavior)

Notes

In most cases, the test will automatically cleanup storage and kill mysqld after it finishes, no matter it's a success, failure or keyboard interrupt. Very occasionally, it will fail to complete auto-cleanup, in which case we need to do this manually. First, kill the mysqld processes:

pkill -9 mysqld

Then, if you are using Cassandra, drop the kepspace by running:

python3 clean_cass.py

or if you are using DynamoDB:

python3 clean_dynamo.py