Details
-
Task
-
Resolution: Fixed
-
Major
-
7.1.0
-
1
Description
What's the issue?
Our unit testing in 'cbbs' can be pretty flaky (especially on the CV machines). If I run the unit testing in a loop for long enough (only takes a few tries) I see failures such as the following:
--- FAIL: TestDefaultJanitorRunLocationCheckOnlyOneNodeFails (0.00s)
|
janitor_test.go:306:
|
Error Trace: janitor_test.go:306
|
Error: Received unexpected error:
|
listen tcp :17894: bind: address already in use
|
Test: TestDefaultJanitorRunLocationCheckOnlyOneNodeFails
|
What's the fix?
Our tests should pass consistently, or be removed if they're unstable