-
Type:
Bug
-
Status: Closed
-
Priority:
Major - P3
-
Resolution: Fixed
-
Affects Version/s: None
-
Component/s: None
-
Labels:None
-
Epic Link:
-
Story Points:5
-
Sprint:Storage Engines 2020-01-27, Storage Engines 2020-02-10
The new Evergreen task format-stress-sanitizer-test failed consistently ever since being migrated from Jenkins (WT-5000). The new model in Evergreen is to run the test for ~24 hrs to simulate the continuously-triggered testing pattern we used to have in Jenkins.
In one of the failed task runs, the test ran through 24 hrs (as expected) and hit a few instances of "heap-use-after-free" (WT-5219) and cache stuck failures (WT-3694, WT-5370 and friends).
We need to either address those known failures (which could be hard esp. the cache stuck failures), or find another way to avoid consistently failing the task.