The requests that we send from the Balancer thread from the primary node of the CSRS doen't have a reasonable time out: the opCtx of that thread doesn't define the maxTimeMs and, since we do not specify the maxTimeMsOverride it is set to Milliseconds::max().

The main problem is that if any of these requests cannot be fulfilled we will hang there forever.

The goal of this ticket is to define a reasonable maxTimeMsOverride for those commands (30s).