Github user rvesse commented on a diff in the pull request:
https://github.com/apache/spark/pull/22805#discussion_r229400780
--- Diff:
resource-managers/kubernetes/core/src/main/scala/org/apache/spark/deploy/k8s/SparkKubernetesClientFactory.scala
---
@@ -42,6 +42,9 @@ private[spark] object SparkKubernetesClientFactory {
sparkConf: SparkConf,
defaultServiceAccountToken: Option[File],
defaultServiceAccountCaCert: Option[File]): KubernetesClient = {
+
+ // TODO Support configurable context
+
--- End diff --
Note to reviewers. In testing this I noticed one issue is that when you
have multiple contexts defined in your `KUBECONFIG` file that while this PR
adds the ability for the integration test machinery to use an arbitrary context
the submission client will always use your current context. Therefore if your
current context does not match the requested context the integration tests will
fail/return false positives because the client and the test harness will use
different contexts.
I have filed SPARK-25887 for addressing this since the client change seemed
like it should be a separate issue and PR. I will provide a PR for that
tomorrow.
---
---------------------------------------------------------------------
To unsubscribe, e-mail: reviews-unsubscr...@spark.apache.org
For additional commands, e-mail: reviews-h...@spark.apache.org
