Github user yifeih commented on a diff in the pull request:
https://github.com/apache/spark/pull/22146#discussion_r212636275
--- Diff:
resource-managers/kubernetes/core/src/main/scala/org/apache/spark/deploy/k8s/Config.scala
---
@@ -225,6 +225,18 @@ private[spark] object Config extends Logging {
"Ensure that major Python version is either Python2 or Python3")
.createWithDefault("2")
+ val KUBERNETES_DRIVER_CONTAINER_NAME =
+ ConfigBuilder("spark.kubernetes.driver.containerName")
--- End diff --
I generally prefer being more explicit about selecting the container name.
Selecting the first container has the danger of silent failures. Especially
since we currently hardcode the container name right now, somebody might write
their pod templates with a driver container name `spark-kubernetes-driver` but
not list it as the first, and then be confused why their spark applications are
failing? But I can see how this could be confusing because it assumes a strong
relationship between the spark conf and the template.
---
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
