-
Bug
-
Resolution: Unresolved
-
Minor
-
None
It's not entirely clear if this should work, but seems like it should be detected as a deadlock if job X #Y is waiting on resource held by job X #Y.
It's not entirely clear if this should work, but seems like it should be detected as a deadlock if job X #Y is waiting on resource held by job X #Y.
Will try to describe bug more clearly:
Assume we have job with 2 defined resources, 1 of them is defined as reference:
In situation when ${RESOURCE_REF} is assigned to RES1, we'll have deadlock with the following log (current run is "test-job #3"):
Note: This will not happen if we define both resources immediately:
In this case job will fail with exception:
(it occurs because we try to add RES1 to IdAllocator context twice)