I and @sharif on this thread noticed a CI build failure on ref application. This is a blocker for the qa technical work. However, have taken time to deeply look into this and would like to help in the resolution. Have gone a head to create a ticket that am working out locally and then do the necessity. @sharif do you mind migrating this ticket to make it ready for work.
cc: @dkayiwa @k.joseph
@k.joseph am current enrolled at this blocker and looks like qa branch has some conflicts with the master branch. Have made efforts here and the builds locally passes. On triggering a commit against qa branch the result seems to be promising. The ci result message received on my mail shows the build passes. But looking at the pr, the Travis CI build fails with some functions flagging a message Could not start a new session. Possible causes are invalid address of the remote server or browser start-up failure. Any advise on this
Thanks @kdaud for looking into this , i dont think that we need this ticket for now because referenceapplication ci is breaking because of many uitest that are annotated @ignored , btw did you see that ci passed when you removed @ignored annotation and with this change RA-1895: Fix ci plan failure on qa branch by kdaud · Pull Request #303 · openmrs/openmrs-distro-referenceapplication · GitHub , so ideally thats the whole cause can be done through the tickets that have been created recently
From my discovery, after refactoring the code by removing the @ignore annotation, all the tests passes locally. BTW even before refactoring the code, the tests passes on build plan locally. However, when you trigger a commit on a refApp qa branch the ci plan fails. Diving into the error log of ci plan, they show an exception which is similar to the local build failure when you run refapplication repo before building the uitestframework
Picking from above, i thought this could be a bug with the host branch qa branch on refApp distro, I then created that ticket you are pointing at. Locally after refactoring the code as you see here the local build plan passes and the experience of the remote ci plan is shared in the first comment i made on this topic
Are you sure about this , how did you run these tests and they locally successfull run, All the tests should first run succesfull locally remember we are working on tests that are @ignored so run each test and run it as junit test . every tests that is ignored will successfull run then when its un ignored its should fail and hence ci build will fail too
Applied mvn clean install
Do i need to run this test in ide otherwise I used terminal
since you are working on [RATEST-96] - OpenMRS Issues, run this specific class test in your IDE (eclipse will do it well)as junit test without touching on anything
Do you mean after addressing the requirement in the ide, I then run mvn clean install as a whole repo on terminal
Run each single test as junit test especially those which are ignored in your IDE not on commandline
Please set qa branch travis to look up its respective qa branch, check through my comment at: RA-1895: Fix ci plan failure on qa branch by kdaud · Pull Request #303 · openmrs/openmrs-distro-referenceapplication · GitHub
1 Like