Github user mateiz commented on a diff in the pull request:
https://github.com/apache/spark/pull/86#discussion_r10332780
--- Diff:
core/src/main/scala/org/apache/spark/deploy/SparkAppArguments.scala ---
@@ -0,0 +1,155 @@
+/*
+ * 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
+
+import scala.collection.mutable.ArrayBuffer
+
+private[spark] class SparkAppArguments(args: Array[String]) {
+ var master: String = null
+ var deployMode: String = null
+ var executorMemory: String = null
+ var executorCores: String = null
+ var driverMemory: String = null
+ var supervise: Boolean = false
+ var queue: String = null
+ var numExecutors: String = null
+ var files: String = null
+ var archives: String = null
+ var mainClass: String = null
+ var primaryResource: String = null
+ var name: String = null
+ var childArgs: ArrayBuffer[String] = new ArrayBuffer[String]()
+ var moreJars: String = null
+ var clientClasspath: String = null
+
+ loadEnvVars()
+ parseArgs(args.toList)
+
+ def loadEnvVars() {
+ master = System.getenv("MASTER")
+ deployMode = System.getenv("DEPLOY_MODE")
+ }
+
+ def parseArgs(args: List[String]) {
+ primaryResource = args(0)
+ parseOpts(args.tail)
+ }
+
+ def parseOpts(opts: List[String]): Unit = opts match {
+ case ("--name") :: value :: tail =>
+ name = value
+ parseOpts(tail)
+
+ case ("--master") :: value :: tail =>
+ master = value
+ parseOpts(tail)
+
+ case ("--class") :: value :: tail =>
+ mainClass = value
+ parseOpts(tail)
+
+ case ("--deploy-mode") :: value :: tail =>
+ if (value != "client" && value != "cluster") {
+ System.err.println("--deploy-mode must be either \"client\" or
\"cluster\"")
+ System.exit(1)
+ }
+ deployMode = value
+ parseOpts(tail)
+
+ case ("--num-executors") :: value :: tail =>
+ numExecutors = value
+ parseOpts(tail)
+
+ case ("--executor-cores") :: value :: tail =>
+ executorCores = value
+ parseOpts(tail)
+
+ case ("--executor-memory") :: value :: tail =>
+ executorMemory = value
+ parseOpts(tail)
+
+ case ("--driver-memory") :: value :: tail =>
+ driverMemory = value
+ parseOpts(tail)
+
+ case ("--supervise") :: tail =>
+ supervise = true
+ parseOpts(tail)
+
+ case ("--queue") :: value :: tail =>
+ queue = value
+ parseOpts(tail)
+
+ case ("--files") :: value :: tail =>
+ files = value
+ parseOpts(tail)
+
+ case ("--archives") :: value :: tail =>
+ archives = value
+ parseOpts(tail)
+
+ case ("--arg") :: value :: tail =>
+ childArgs += value
+ parseOpts(tail)
+
+ case ("--more-jars") :: value :: tail =>
+ moreJars = value
+ parseOpts(tail)
+
+ case ("--client-classpath") :: value :: tail =>
+ clientClasspath = value
+ parseOpts(tail)
+
+ case ("--help" | "-h") :: tail =>
+ printUsageAndExit(0)
+
+ case Nil =>
+
+ case _ =>
+ printUsageAndExit(1, opts)
+ }
+
+ def printUsageAndExit(exitCode: Int, unknownParam: Any = null) {
+ if (unknownParam != null) {
+ System.err.println("Unknown/unsupported param " + unknownParam)
+ }
+ System.err.println(
+ "Usage: spark-app <primary binary> [options] \n" +
+ "Options:\n" +
+ " --master MASTER_URL spark://host:port,
mesos://host:port, yarn, or local\n" +
+ " --deploy-mode DEPLOY_MODE Mode to deploy the app in, either
\"client\" or \"cluster\"\n" +
+ " --class CLASS_NAME Name of your application's main
class (required for Java apps)\n" +
+ " --arg ARG Argument to be passed to your
application's main class.\n" +
+ " Multiple invocations are possible,
each will be passed, in order.\n" +
+ " --num-executors NUM Number of executors to start
(Default: 2)\n" +
+ " --executor-cores NUM Number of cores per executor
(Default: 1)\n" +
+ " --executor-memory MEM Memory per executor (e.g. 1000M, 2G)
(Default: 1G)\n" +
+ " --driver-memory MEM Memory for driver (e.g. 1000M, 2G)
(Default: 512 Mb)\n" +
+ " --name NAME The name of your application
(Default: Spark)\n" +
+ " --queue QUEUE The YARN queue to use for allocation
requests (Default: 'default')\n" +
+ " --more-jars jars For \"cluster\" deploy mode, comma
separated list of local jars\n" +
+ " that you want SparkContext.addJar to
work with. Only works with YARN.\n" +
+ " --files files Comma separated list of files to be
placed next to all executors.\n" +
+ " Only works with YARN.\n" +
+ " --archives archives Comma separated list of archives to
be placed next to all executors.\n" +
+ " Only works with YARN.\n" +
+ " --client-classpath ENTRIES Entries to be placed on the client
JVM's classpath"
--- End diff --
Instead of environment vars, couldn't we do everything through system
properties? SparkConf already initializes from system properties by default,
and any environment vars that we might need to set could be given a system
property to do the same thing (some YARN settings might currently only use env
vars). For setting env vars on the workers it's possible to do it in all of the
cluster managers that Spark supports.
---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your project does not have this feature
enabled and wishes so, or if the feature is enabled but not working, please
contact infrastructure at infrastruct...@apache.org or file a JIRA ticket
with INFRA.
---
