This repository is the source/build code for weka/tools/wekachecker
The scripts.d/ directory contains a number of subdirectories that contain bash scripts. Use -w or --workload to specify which set of scripts to execute. Of course, if not specified, the default/ directory will be used.
Currently, there is scripts.d/default, scripts.d/client and scripts.d/ta.
Example would be ./wekachecker -w ta to run the scripts in the scripts.d/ta directory.
The scripts are written in bash. The return code indicates pass/fail of the tests.
Specifically, the scripts can return:
| Return Code | Meaning |
|---|---|
| 0 | PASS |
| 255 | HARDFAIL |
| 254 | WARN |
| anything else | FAIL |
Simply use return 0 for a passing grade, or return 254 for a warning, etc.
There is a preamble script in each directory. The preable is injected before the test script, and handles some standard argument parsing and such (take a look for details).
wekachecker can run the scripts in 4 ways - single host, sequentially on all hosts, parallel on all hosts, and parallel comparing (to compare items from all hosts).
The run method is specified by defining SCRIPT_TYPE in the script. For example, SCRIPT_TYPE="parallel-compare-backends" or SCRIPT_TYPE=parallel
By setting the DESCRIPTION variable in the script, wekachecker will display the value of DESCRIPTION first, the the result (PASS/FAIL, etc) as it executes the scripts.
That's about all there is to it... your script is free to do most anything on the target system(s); but please avoid modifying things that might upset customers. ;)