This problem surfaced when the WriteBehind configuration testcases introduced in
OPENJPA-1237 caused the existing WriteBehind testcases in trunk to fail. Apparently, the WriteBehindCallback sleepTime interval of one EMF datacache is being used and adversely impacting other EMF datacaches. This is not the desired behavior as the WriteBehind datacache for each EMF should exist and function independently of any other WriteBehind datacache(s). The easiest way to demonstrate this failure is to increase the WriteBehindCallback sleepTime internval from 1000 to 15000 milliseconds in each of the testcases in TestWriteBehindConfigurations, as it will cause other WriteBehind testcases in trunk to fail.