FLUID-5387: Random failure of tests run in all-tests

Metadata

Source
FLUID-5387
Type
Bug
Priority
Major
Status
Closed
Resolution
Cannot Reproduce
Assignee
Antranig Basman
Reporter
Justin Obara
Created
2014-06-02T10:09:31.446-0400
Updated
2015-06-26T10:10:30.867-0400
Versions
N/A
Fixed Versions
  1. 1.9
Component
  1. Prefs Framework
  2. UIEnhancer
  3. UI Options

Description

There appears to be some tests failing randomly in the all-tests.html. However, when they are run individually they pass. Different tests are failing on different runs, so it is hard to track down what is causing it exactly. It seems that it could be the case where elements that are needed for the test are destroyed before the test completes.

Environments

IE 8 (Windows 7)

Comments

  • Justin Obara commented 2014-06-02T13:23:38.377-0400

    This issue seems to have arisen with this merge https://github.com/fluid-project/infusion/commit/f77fac2d2d8ecd7fe5ffd2f497cf404e8f13814b

  • Anastasia Cheetham commented 2014-06-09T13:40:59.230-0400

    GIven that this is only in IE8, perhaps we shouldn't worry too much about it?

  • Antranig Basman commented 2014-06-11T10:16:01.875-0400

    This issue results from a timing issue (race condition) in the IoC testing framework. I believe it will require an upgrade to qunit in order to resolve it but I haven't had time to investigate it in detail. I have seen these failures on several browsers and not just IE8, especially if the machine is running slowly. The workaround is to run the failing tests individually and verify that they run. We should push this out for Infusion 2.0 since it will take a lot of diagnosis to resolve.

  • Justin Obara commented 2014-06-11T10:17:13.913-0400

    Spoke with @@Antranig Basman in the channel today about this. It seems there is a timing issue, that may require a fix in qunit. This seems to be out of the scope for the 1.5 release, and should be addressed afterwards.

    https://botbot.me/freenode/fluid-work/2014-06-11/

  • Justin Obara commented 2015-06-15T14:56:31.722-0400

    Could not reproduce with IE 11 (which is currently the latest version of IE supported)

  • Justin Obara commented 2015-06-26T09:51:17.190-0400

    Reopening to change the fix version to 1.9