Make driver aware of Autoscaling
Driver not aware of autoscaling (topology changes)
This is related to https://support.mongodb.com/case/01186544. We are experiencing exceptions when autoscaling is supposed to work flawlessly. The problem seems to be that the driver checks for a heartbeat every few seconds but fails to do so under heavy bursts of workloads, which we are currently facing.
When autoscaling kicks in, there should be a possibility to make the nodes aware that they have been scaled, so that when they communicate with the driver, they provide this information back to the driver. The driver would then explore the new topology first before throwing an exception on write failures.