Iceberg Vacuum: Timeout waiting for connection from pool

ConnectionPoolTimeoutException: Timeout waiting for connection from pool
6512a0e6-4c06-4d20-bc95-6720bc7e7071.zip (33,2 КБ)

We are getting such error while running VACUUM command over highly updatable iceberg table in aws GLUE.
I already saw similar question advice to set fs.s3a.connection.maximum to some higher value.
where do we have to set it in our case(aws cluster)? in executor node instances/coordinator or Aws glue datasource settings?
In case of EC2 - when i stop/start engine, node will be terminated and started again so i will have to set this property again?

Also after getting this error all queries to Glue on that node are failing for some period(haven’t measured but smth around 30 min orm ore), later it is good again. There is no way to avoid such behavior?
7ceb07e2-5b5c-4dc5-b209-ee93efc9a2fd.zip (17,6 КБ)

@vladislav-stolyarov For a immediate workaround, are you able to add executors? If not try increasing fs.s3a.connection.maximum to something like 200, if that does not help, try increasing fs.s3a.connection.establish.timeout.

You might also want to consider upgrading to the latest version as there have been some fixes on this issue