The testGetCurrent testGetCurrent method, if run at any point between 23 UTC and 24 UTC, will fail. Is this test necessary? If so, we should try to beef it up. Having an hour of each day where DSpace reliably fails testing is a bit odd.
UPDATE: after fixing testGetCurrent, two more tests were discovered to have similar problems: InstallItemTest and RestoreItemTest. DSPR#1449 fixes all of these tests.