Open xhernandez opened 1 year ago
I think the best way would be to have a final test that fail if there is a coredump or something. Of course, it only work if the test clean the coredumps themself after.
Actually the test suite already detects if a test has generated a core dump or not (though it doesn't mark the test as failed). The list of cores appears at the end of the regression, under x test(s) generated core
. Couldn't we parse that list ?
I guess somewhere around here: https://github.com/gluster/fstat/blob/master/fstat/parser.py
I am not familliar with that code base, but I guess no one else is anymore.
I guess somewhere around here: https://github.com/gluster/fstat/blob/master/fstat/parser.py
I am not familliar with that code base, but I guess no one else is anymore.
Most probably. I'll try to check the code and see if I can send a PR.
If a test script passes all tests successfully but generates a core dump, it's not captured in the fstat statistics. These problems should also be collected and displayed (if possible indicating if it's a test failure and/or a core dump).