when recovery takes prolonged time due to waiting for all of the clients to reconnect, the recovery task triggers a the hangcheck timer.

There is likely a way to tell the kernel we are fine or otherwise have some sort of a ping happen to avoid this message that unnecessary scares users.