You are viewing a plain text version of this content. The canonical link for it is here.
Posted to issues@spark.apache.org by "Albert Chu (JIRA)" <ji...@apache.org> on 2017/07/29 00:24:00 UTC
[jira] [Commented] (SPARK-21570) File __spark_libs__XXX.zip does
not exist on networked file system w/ yarn
[ https://issues.apache.org/jira/browse/SPARK-21570?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16105900#comment-16105900 ]
Albert Chu commented on SPARK-21570:
------------------------------------
Oh, and because it will likely be asked and may likely be relevant. In this test setup, HDFS is not used at all.
{noformat}
<property>
<name>fs.defaultFS</name>
<value>file:///</value>
</property>
{noformat}
All temp dirs, staging dirs, etc. are configured to appropriate locations in /tmp or somewhere in the networked file system.
> File __spark_libs__XXX.zip does not exist on networked file system w/ yarn
> --------------------------------------------------------------------------
>
> Key: SPARK-21570
> URL: https://issues.apache.org/jira/browse/SPARK-21570
> Project: Spark
> Issue Type: Bug
> Components: YARN
> Affects Versions: 2.2.0
> Reporter: Albert Chu
>
> I have a set of scripts that run Spark with data in a networked file system. One of my unit tests to make sure things don't break between Spark releases is to simply run a word count (via org.apache.spark.examples.JavaWordCount) on a file in the networked file system. This test broke with Spark 2.2.0 when I use yarn to launch the job (using the spark standalone scheduler things still work). I'm currently using Hadoop 2.7.0. I get the following error:
> {noformat}
> Diagnostics: File file:/p/lcratery/achu/testing/rawnetworkfs/test/1181015/node-0/spark/node-0/spark-292938be-7ae3-460f-aca7-294083ebb790/__spark_libs__695301535722158702.zip does not exist
> java.io.FileNotFoundException: File file:/p/lcratery/achu/testing/rawnetworkfs/test/1181015/node-0/spark/node-0/spark-292938be-7ae3-460f-aca7-294083ebb790/__spark_libs__695301535722158702.zip does not exist
> at org.apache.hadoop.fs.RawLocalFileSystem.deprecatedGetFileStatus(RawLocalFileSystem.java:606)
> at org.apache.hadoop.fs.RawLocalFileSystem.getFileLinkStatusInternal(RawLocalFileSystem.java:819)
> at org.apache.hadoop.fs.RawLocalFileSystem.getFileStatus(RawLocalFileSystem.java:596)
> at org.apache.hadoop.fs.FilterFileSystem.getFileStatus(FilterFileSystem.java:421)
> at org.apache.hadoop.yarn.util.FSDownload.copy(FSDownload.java:253)
> at org.apache.hadoop.yarn.util.FSDownload.access$000(FSDownload.java:63)
> at org.apache.hadoop.yarn.util.FSDownload$2.run(FSDownload.java:361)
> at org.apache.hadoop.yarn.util.FSDownload$2.run(FSDownload.java:359)
> at java.security.AccessController.doPrivileged(Native Method)
> at javax.security.auth.Subject.doAs(Subject.java:422)
> at org.apache.hadoop.security.UserGroupInformation.doAs(UserGroupInformation.java:1657)
> at org.apache.hadoop.yarn.util.FSDownload.call(FSDownload.java:358)
> at org.apache.hadoop.yarn.util.FSDownload.call(FSDownload.java:62)
> at java.util.concurrent.FutureTask.run(FutureTask.java:266)
> at java.util.concurrent.Executors$RunnableAdapter.call(Executors.java:511)
> at java.util.concurrent.FutureTask.run(FutureTask.java:266)
> at java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1142)
> at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:617)
> at java.lang.Thread.run(Thread.java:748)
> {noformat}
> While debugging, I sat and watched the directory and did see that /p/lcratery/achu/testing/rawnetworkfs/test/1181015/node-0/spark/node-0/spark-292938be-7ae3-460f-aca7-294083ebb790/__spark_libs__695301535722158702.zip does show up at some point.
> Wondering if it's possible something racy was introduced. Nothing in the Spark 2.2.0 release notes suggests any type of configuration change that needs to be done.
> Thanks
--
This message was sent by Atlassian JIRA
(v6.4.14#64029)
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscribe@spark.apache.org
For additional commands, e-mail: issues-help@spark.apache.org