Too much contention in the server when performing QE reads can cause the internal transactions to exceed the retry limit of 120 and return a NoSuchTransaction error. Because of this, the fle2_crud.js FSM test may fail when many worker threads are performing "findOne" in the readDocs phase. So, to stop the test from terminating when it hits this transient error, the test must be fixed so that a worker thread can catch the NoSuchTransaction error and back off for a few seconds before retrying the read.