Let me explain a little more.
Both are long running tests so when the first one starts it grabs the data sink file and starts writing to it.
Well before this test case ends, the second test case needs access to the file that is being written to.
So for example the data sink may have 40 rows written to it so far, there are another 60 to go, but the second test case needs access now to that file to read out row number 1 for a different request but it can't access it because it's still being written to. This is how I understand data sinks work.
But surely this cant be a problem never encountered before?
