To solve random tests failures, you can use the use the --bisect
argument when running rspec
.
A second argument --seed N
is required to correctly bisect the tests.
Take note of the seed number used in the latest failing test and run:
bundle exec rspec spec --seed SEED_NUMBER --bisect
Example result: