Run a GB-sized integration test

Create a GB-sized dataset

Once a Qserv mono-node instance is configured, you can automatically download a GB-sized dataset from a remote server using next command:


This part require rsync and a ssh key to access


Example below (case #04) provides around 16GB data extracted from Winter13 and stored at NCSA computing center. It requires a 40GB storage for partitioning and loading the data, and will create a 33GB MySQL database (for both Qserv and plain MySQL), in QSERV_RUN_DIR. --download --work-dir=/path/to/large/storage --case-id=04 --custom-case-id=<new-id>
# See advanced options with --help option

Here data will be duplicated from built-in test case #04, located in $QSERV_TESTDATA_DIR/datasets, to /path/to/large/storage/case<new-id>, and some original data files will be overridden with big remote data files.


Built-in test datasets are good models if you want to create your own custom test dataset by hand.

Run integration test on a custom dataset

First start Qserv and then: --load --testdata-dir=/path/to/large/storage --case-id=<new-id>
# See advanced options with --help option

This test will create a Qserv database named qservTest_case<new-id>_qserv, and a plain MySQL database named qservTest_case<new-id>_mysql. Results are stored in QSERV_RUN_DIR/tmp/qservTest_case<new-id>/outputs/, and are erased before each run.