I've been facing this issue for the past few months too.
I always thought it was an infrastructure issue, but we were never able to
figure out what the infra issue was.
If others are facing this issue too - then maybe it's a valid bug.
Does anyone have any ideas on how we can debug this?
On Fri,
Hi!
We have dynamic allocation enabled for our regular jobs and sometimes they fail
with java.util.concurrent.TimeoutException: Futures timed out after [300
seconds]. Seems like spark driver starts broadcast just before the job has
received any executors from the YARN and if it takes more than
