Github user mccheah commented on a diff in the pull request:
https://github.com/apache/spark/pull/20669#discussion_r173871228
--- Diff:
resource-managers/kubernetes/core/src/main/scala/org/apache/spark/deploy/k8s/submit/steps/DriverConfigPropertiesStep.scala
---
@@ -0,0 +1,85 @@
+/*
+ * Licensed to the Apache Software Foundation (ASF) under one or more
+ * contributor license agreements. See the NOTICE file distributed with
+ * this work for additional information regarding copyright ownership.
+ * The ASF licenses this file to You under the Apache License, Version 2.0
+ * (the "License"); you may not use this file except in compliance with
+ * the License. You may obtain a copy of the License at
+ *
+ * http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ */
+package org.apache.spark.deploy.k8s.submit.steps
+
+import java.io.StringWriter
+import java.util.Properties
+
+import io.fabric8.kubernetes.api.model._
+
+import org.apache.spark.SparkConf
+import org.apache.spark.deploy.k8s.Config._
+import org.apache.spark.deploy.k8s.Constants._
+import org.apache.spark.deploy.k8s.submit.KubernetesDriverSpec
+
+/**
+ * Create a config map with the driver configuration and attach it to the
pod. This needs to
+ * come at the end of the driver configuration so that all modifications
to the Spark config
+ * are reflected in the generated config map.
+ */
+private[spark] class DriverConfigPropertiesStep(resourceNamePrefix: String)
+ extends DriverConfigurationStep {
+
+ override def configureDriver(spec: KubernetesDriverSpec):
KubernetesDriverSpec = {
+ val configMapName = s"$resourceNamePrefix-driver-conf-map"
+ val configMap = buildConfigMap(configMapName, spec.driverSparkConf)
--- End diff --
I see - using the config map is a good idea then, but it still can't be in
a config step because it has to assume that all the properties its given are
the absolute final ones. We should keep the driver properties code in the same
_place_ as it was before, which was separate from the config steps and after
all the steps were processed. The implementation can still be with the config
map which allows spark-submit to pick up these properties.
Just to double check, will this logic need to be re-used anywhere? I don't
think the executor pods have to mount the same properties config map.
> I will also include --conf spark.driver.extraJavaOptions= in the
spark-submit launched in the driver
That won't work because by the time spark-submit is running, you'll want to
_already_ have these JVM options applied since spark-submit will be running the
driver JVM itself. You'll need to find a way to get the spark-submit JVM to run
with these JVM options.
---
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
