Details
-
Improvement
-
Status: Open
-
Major
-
Resolution: Unresolved
-
None
-
None
-
None
Description
This issue was brought up while running Mesos integration tests on docker. Some tests fail with timeout, caused by insufficient CPU resources in the virtual environment.
There are 5 tests that require (relatively) large resources: CloseEvaluatorTest (16 Evaluators) and 4 Vortex Tests (2 Evaluators each of which has 4 CPU cores). We may reduce the amount of the resources to request a bit, as the main goal is to test the functionality, rather than the performance.
However, I'm not sure yet which numbers would be good enough; would it be fair enough to set the numbers to pass the mesos integration tests running on Docker in OSX with the , which only causes trouble for now? Any suggestion is welcome.