Mailing-List: contact reviews-help@spark.apache.org; run by ezmlm
Precedence: bulk
Reply-To: reviews@spark.apache.org
From: berngp <git@git.apache.org>
To: reviews@spark.apache.org
Reply-To: reviews@spark.apache.org
References: <git-pr-433-spark@git.apache.org>
In-Reply-To: <git-pr-433-spark@git.apache.org>
Subject: [GitHub] spark pull request: [SPARK-1522] : YARN ClientBase throws a
 NPE if...
Content-Type: text/plain
Message-Id: <20140418033736.BF9CC99759E@tyr.zones.apache.org>
Date: Fri, 18 Apr 2014 03:37:36 +0000 (UTC)

Github user berngp commented on a diff in the pull request:

    https://github.com/apache/spark/pull/433#discussion_r11763060
  
    --- Diff: yarn/common/src/main/scala/org/apache/spark/deploy/yarn/ClientBase.scala ---
    @@ -378,55 +378,48 @@ object ClientBase {
       val APP_JAR: String = "app.jar"
       val LOG4J_PROP: String = "log4j.properties"
     
    -  // Based on code from org.apache.hadoop.mapreduce.v2.util.MRApps
    -  def populateHadoopClasspath(conf: Configuration, env: HashMap[String, String]) {
    -    val classpathEntries = Option(conf.getStrings(
    -      YarnConfiguration.YARN_APPLICATION_CLASSPATH)).getOrElse(
    -        getDefaultYarnApplicationClasspath())
    -    for (c <- classpathEntries) {
    -      YarnSparkHadoopUtil.addToEnvironment(env, Environment.CLASSPATH.name, c.trim,
    -        File.pathSeparator)
    -    }
    +  def populateHadoopClasspath(conf: Configuration, env: HashMap[String, String]) = {
    +    val classPathElementsToAdd = getYarnAppClasspath(conf) ++ getMRAppClasspath(conf)
    +    addToAppClasspath(env, classPathElementsToAdd)
    +    classPathElementsToAdd
    +  }
     
    -    val mrClasspathEntries = Option(conf.getStrings(
    -      "mapreduce.application.classpath")).getOrElse(
    -        getDefaultMRApplicationClasspath())
    -    if (mrClasspathEntries != null) {
    -      for (c <- mrClasspathEntries) {
    -        YarnSparkHadoopUtil.addToEnvironment(env, Environment.CLASSPATH.name, c.trim,
    -          File.pathSeparator)
    -      }
    -    }
    +  protected[yarn] def getYarnAppClasspath(conf: Configuration) = 
    +    getAppClasspathForKey(YarnConfiguration.YARN_APPLICATION_CLASSPATH, conf)(getDefaultYarnApplicationClasspath)
    +
    +  protected[yarn] def getMRAppClasspath(conf: Configuration) =
    +    getAppClasspathForKey("mapreduce.application.classpath", conf)(getDefaultMRApplicationClasspath)
    +
    +  protected[yarn] def addToAppClasspath(env: HashMap[String, String], elements : Iterable[String]) {
    +    for ( c <- elements ) 
    +      yield (YarnSparkHadoopUtil.addToEnvironment(env, Environment.CLASSPATH.name, c.trim, File.pathSeparator))
    --- End diff --
    
    I kept the `for (c <- classPathElementsToAdd.flatten)`, to be honest the `yield` was a remanence of a different approach.


---
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 infrastructure@apache.org or file a JIRA ticket
with INFRA.
---