FWD » Core Development » Database

I put in a tentative fix into rev. 10928 for a race condition in DynamicTablesHelper.createDynamicDMO, which passed regression testing and some unit testing. However, to truly confirm it is fixed, I would like to force the situation in the debugger, using breakpoints to simulate the race condition. This will involve creating the same dynamic temp-table in two contexts and halting both after the first cache check misses, then allowing one to build the table and put it to the cache. The original error (NPE) occurred when the second thread would get a cache hit on the second try, so the test would be to confirm this now works without the NPE.