Comment 8 for bug 1394576

Andrew, yes, we are. Also I was mentioned we check if there some time sqew on ther nodes, and if so sync it manually. As an additional check for rabbit here( that creates queues publish it and than consumes) and it pass. But when we manually created instances, or secgroups or other objects it unpredictable failed with oslo traces. It can be 2-4 times from 20 attempts