Hello everyone!
I currently have a problem: we writing long test scenarios (it might take about 13 minutes for each one) and if i execute one single scenario, it always pass. But if i'm executing all feature file, test run will be aborted after 2-4th scenario.
I run test scenario on local Mac, and all scenarios passed, so we got this problem on windows build agents only.
I got same ruby and gems on windows and mac. Here is gem list:
awesome_print (1.7.0)
bigdecimal (1.2.6)
builder (3.2.2)
calabash-android (0.8.0)
cucumber (2.4.0)
cucumber-core (1.5.0)
cucumber-wire (0.0.1)
diff-lcs (1.2.5)
escape (0.0.4)
gherkin (4.0.0)
httpclient (2.8.2.2)
io-console (0.4.3)
json (1.8.1)
luffa (2.0.0)
minitest (5.4.3)
multi_json (1.12.1)
multi_test (0.1.2)
power_assert (0.2.2)
psych (2.0.8)
rake (10.4.2)
rdoc (4.2.0)
retriable (2.0.2)
rubyzip (1.2.0)
slowhandcuke (0.0.3)
test-unit (3.0.8)
thor (0.19.1)
We are not using Bundler and Gemfiles. We tried to run scenario by "calabash-android run" command and by "rake", and got different logs.
Executing by "calabash-android run" log: calabash_android_run_logs.txt
Executing by rake log: rake_logs.txt
We tried to run this on Ruby 2.2 and Ruby 2.3, and got same result
We also tried to run scenarios not from TeamCity build server, but manually, and got same result, so the reason of problem is not in TeamCity.