[SPARK-28992] Support update dependencies from hdfs when task run on executor pods

• Type: Improvement
• Status: In Progress
• Priority: Major
• Resolution: Unresolved
• Affects Version/s: 3.1.0
• Component/s: Kubernetes, Spark Core
• Assignee: Unassigned
• Reporter: Kent Yao 2
• Created: 05/Sep/19 11:33
• Updated: 04/Jan/21 08:03

Description

Here is a case: 

bin/spark-submit  --class com.github.ehiggs.spark.terasort.TeraSort hdfs://hz-cluster10/user/kyuubi/udf/spark-terasort-1.1-SNAPSHOT-jar-with-dependencies.jar hdfs://hz-cluster10/user/kyuubi/terasort/1000g hdfs://hz-cluster10/user/kyuubi/terasort/1000g-out1

Spark supports add jar logic and application-jar from hdfs - -  http://spark.apache.org/docs/latest/submitting-applications.html#launching-applications-with-spark-submit

Take spark on yarn for example, it creates a _spark_hadoop_conf_.xml file and upload the hadoop distribute cache, the executor processes can use this to identify where their dependencies located.

But on k8s, i tried and failed to update dependencies.

19/09/04 08:08:52 INFO scheduler.DAGScheduler: ShuffleMapStage 0 (newAPIHadoopFile at TeraSort.scala:60) failed in 1.058 s due to Job aborted due to stage failure: Task 0 in stage 0.0 failed 4 times, most recent failure: Lost task 0.3 in stage 0.0 (TID 9, 100.66.0.75, executor 2): java.lang.IllegalArgumentException: java.net.UnknownHostException: hz-cluster10
19/09/04 08:08:52 INFO scheduler.DAGScheduler: ShuffleMapStage 0 (newAPIHadoopFile at TeraSort.scala:60) failed in 1.058 s due to Job aborted due to stage failure: Task 0 in stage 0.0 failed 4 times, most recent failure: Lost task 0.3 in stage 0.0 (TID 9, 100.66.0.75, executor 2): java.lang.IllegalArgumentException: java.net.UnknownHostException: hz-cluster10 at org.apache.hadoop.security.SecurityUtil.buildTokenService(SecurityUtil.java:378) at org.apache.hadoop.hdfs.NameNodeProxies.createNonHAProxy(NameNodeProxies.java:310) at org.apache.hadoop.hdfs.NameNodeProxies.createProxy(NameNodeProxies.java:176) at org.apache.hadoop.hdfs.DFSClient.<init>(DFSClient.java:678) at org.apache.hadoop.hdfs.DFSClient.<init>(DFSClient.java:619) at org.apache.hadoop.hdfs.DistributedFileSystem.initialize(DistributedFileSystem.java:149) at org.apache.hadoop.fs.FileSystem.createFileSystem(FileSystem.java:2669) at org.apache.hadoop.fs.FileSystem.access$200(FileSystem.java:94) at org.apache.hadoop.fs.FileSystem$Cache.getInternal(FileSystem.java:2703) at org.apache.hadoop.fs.FileSystem$Cache.get(FileSystem.java:2685) at org.apache.hadoop.fs.FileSystem.get(FileSystem.java:373) at org.apache.spark.util.Utils$.getHadoopFileSystem(Utils.scala:1881) at org.apache.spark.util.Utils$.doFetchFile(Utils.scala:737) at org.apache.spark.util.Utils$.fetchFile(Utils.scala:522) at org.apache.spark.executor.Executor.$anonfun$updateDependencies$7(Executor.scala:869) at org.apache.spark.executor.Executor.$anonfun$updateDependencies$7$adapted(Executor.scala:860) at scala.collection.TraversableLike$WithFilter.$anonfun$foreach$1(TraversableLike.scala:792) at scala.collection.mutable.HashMap.$anonfun$foreach$1(HashMap.scala:149) at scala.collection.mutable.HashTable.foreachEntry(HashTable.scala:237) at scala.collection.mutable.HashTable.foreachEntry$(HashTable.scala:230) at scala.collection.mutable.HashMap.foreachEntry(HashMap.scala:44) at scala.collection.mutable.HashMap.foreach(HashMap.scala:149) at scala.collection.TraversableLike$WithFilter.foreach(TraversableLike.scala:791) at org.apache.spark.executor.Executor.org$apache$spark$executor$Executor$$updateDependencies(Executor.scala:860) at org.apache.spark.executor.Executor$TaskRunner.run(Executor.scala:409) at java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1149) at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:624) at java.lang.Thread.run(Thread.java:748)

 

Issue Links

• is duplicated by: SPARK-29974 Submitting with application jar on HA HDFS
• links to: GitHub Pull Request #25695