Query requests can get stuck causing rebalance to fail to run to completion

In a recent case on 7.2.4 we saw that apparently stuck requests caused Query to not run its part of the rebalance to completion with the rebalance failing with an inactivity_timeout after 10 minutes. This happened in 12 rebalance instances over the course of a couple of days. These could possibly be long-running requests, but 10 minutes is a very long time and since it happened a bunch of times it indicates that there may actually be a request processing bug.  

Also: we should consider simply terminating requests that take this length of time.