# Description of problem: Today, if a regression test fails regression due to some errors, by default, there is no way for the author to figure out if there are any more tests failing. Most of the times, they have to fix it and send another patch to figure out there is another one which fails. This process continues till all tests pass, or author has to send a patch with changing the 'run-tests.sh' to not fail on failure, and then again send another one disabling that. The best way is to have an option to run the regression tests with '--exit-on-failures=no' option given a review command called 'run all-tests-in-regression'. Not critical about changing the command, but need a announcement on this. # Additional info: Search bugs with 'Process-Automation' in Whiteboard for the series of bugs.
https://review.gluster.org/#/c/20042 added, and triggered the sample run in https://review.gluster.org/20035 (https://build.gluster.org/job/regression-on-demand-full-run/1/)