You are viewing a plain text version of this content. The canonical link for it is here.
Posted to issues@spark.apache.org by "Marcelo Vanzin (JIRA)" <ji...@apache.org> on 2014/11/21 00:56:34 UTC
[jira] [Commented] (SPARK-4503) The history server is not
compatible with HDFS HA
[ https://issues.apache.org/jira/browse/SPARK-4503?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14220258#comment-14220258 ]
Marcelo Vanzin commented on SPARK-4503:
---------------------------------------
Hi there,
I just tried this with my Spark build and it seems to work fine. I haven't tried the official 1.1.0 tarball, though. Could you share the complete exception from your logs? It would be nice to know the call stack up to {{SecurityUtil.buildTokenService}}.
> The history server is not compatible with HDFS HA
> -------------------------------------------------
>
> Key: SPARK-4503
> URL: https://issues.apache.org/jira/browse/SPARK-4503
> Project: Spark
> Issue Type: Bug
> Affects Versions: 1.1.0
> Reporter: MarsXu
> Priority: Minor
>
> I use a high availability of HDFS to store the history server data.
> Can be written eventlog to HDFS , but history server cannot be started.
>
> Error log when execute "sbin/start-history-server.sh":
> {quote}
> ....
> 14/11/20 10:25:04 INFO SecurityManager: SecurityManager: authentication disabled; ui acls disabled; users with view permissions: Set(root, ); users with modify permissions: Set(root, )
> 14/11/20 10:25:04 WARN NativeCodeLoader: Unable to load native-hadoop library for your platform... using builtin-java classes where applicable
> Exception in thread "main" java.lang.reflect.InvocationTargetException
> at sun.reflect.NativeConstructorAccessorImpl.newInstance0(Native Method)
> at sun.reflect.NativeConstructorAccessorImpl.newInstance(NativeConstructorAccessorImpl.java:39)
> at sun.reflect.DelegatingConstructorAccessorImpl.newInstance(DelegatingConstructorAccessorImpl.java:27)
> at java.lang.reflect.Constructor.newInstance(Constructor.java:513)
> at org.apache.spark.deploy.history.HistoryServer$.main(HistoryServer.scala:187)
> at org.apache.spark.deploy.history.HistoryServer.main(HistoryServer.scala)
> Caused by: java.lang.IllegalArgumentException: java.net.UnknownHostException: appcluster
> at org.apache.hadoop.security.SecurityUtil.buildTokenService(SecurityUtil.java:377)
> ....
> {quote}
> When I set <export SPARK_HISTORY_OPTS="-Dspark.history.fs.logDirectory=hdfs://s161.zw.db.d:53310/spark_history"> in spark-evn.sh, can start, but no high availability.
> Environment
> {quote}
> spark-1.1.0-bin-hadoop2.4
> hadoop-2.5.1
> zookeeper-3.4.6
> {quote}
> The config file is as follows:
> {quote}
> !### spark-defaults.conf ###
> spark.eventLog.dir hdfs://appcluster/history_server/
> spark.yarn.historyServer.address s161.zw.db.d:18080
> !### spark-env.sh ###
> export SPARK_HISTORY_OPTS="-Dspark.history.fs.logDirectory=hdfs://appcluster/history_server"
> !### core-site.xml ###
> <property>
> <name>fs.defaultFS</name>
> <value>hdfs://appcluster</value>
> </property>
> !### hdfs-site.xml ###
> <property>
> <name>dfs.nameservices</name>
> <value>appcluster</value>
> </property>
> <property>
> <name>dfs.ha.namenodes.appcluster</name>
> <value>nn1,nn2</value>
> </property>
> <property>
> <name>dfs.namenode.rpc-address.appcluster.nn1</name>
> <value>s161.zw.db.d:8020</value>
> </property>
> <property>
> <name>dfs.namenode.rpc-address.appcluster.nn2</name>
> <value>s162.zw.db.d:8020</value>
> </property>
> <property>
> <name>dfs.namenode.servicerpc-address.appcluster.nn1</name>
> <value>s161.zw.db.d:53310</value>
> </property>
> <property>
> <name>dfs.namenode.servicerpc-address.appcluster.nn2</name>
> <value>s162.zw.db.d:53310</value>
> </property>
> {quote}
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscribe@spark.apache.org
For additional commands, e-mail: issues-help@spark.apache.org