Hello all, I am surprised that it is not possible to define "spark.kubernetes.driver.request.cores" when submitting a spark job on kubernetes.
My understanding is that it would limit the CPU requests for the driver on the k8s cluster and we could still define how many cores (threads) we use in the driver with spark.driver.cores. In the 2.4 branch, the request is then defined by "spark.driver.cores", so we cannot throttle the driver. Is my understanding wrong or I missed something ? Thanks Christophe -- Sent from: http://apache-spark-user-list.1001560.n3.nabble.com/ --------------------------------------------------------------------- To unsubscribe e-mail: user-unsubscr...@spark.apache.org
