I have Apache Mesos 0.22.1 cluster (3 masters & 5 slaves), running Cloudera
HDFS (2.5.0-cdh5.3.1) in HA configuration and Spark 1.5.1 framework.
When I try to spark-submit compiled HdfsTest.scala example app (from Spark
1.5.1 sources) - it fails with "java.lang.IllegalArgumentException:
java.net.UnknownHostException: hdfs" error in executor logs. This error is
only observed when I pass HDFS HA Path as an argument ("hdfs://hdfs/"), when
I pass "hdfs://namenode1.hdfs.mesos:50071/tesfile" - everything works fine.
What I've found after enabling TRACE logging is that Spark driver actually
reads hdfs://hdfs URL correctly, but Spark executor - doesn't.
My Scala app code:
import org.apache.spark._
object HdfsTest {
def main(args: Array[String]) {
val sparkConf = new SparkConf().setAppName("HdfsTest")
val sc = new SparkContext(sparkConf)
val file = sc.textFile(args(0))
val mapped = file.map(s => s.length).cache()
for (iter <- 1 to 10) {
val start = System.currentTimeMillis()
for (x <- mapped) { x + 2 }
val end = System.currentTimeMillis()
println("Iteration " + iter + " took " + (end-start) + " ms")
}
sc.stop()
}
}
I compile this code and submit jar file to Spark in cluster mode:
/opt/spark/bin/spark-submit --deploy-mode cluster --class
com.cisco.hdfs.HdfsTest http://1.2.3.4/HdfsTest-0.0.1.jar
hdfs://hdfs/testfile
My spark-defaults.conf file:
spark.master spark://1.2.3.4:7077
spark.eventLog.enabled true
spark.driver.memory 1g
My spark-env.sh file:
export HADOOP_HOME=/opt/spark
export HADOOP_CONF_DIR=/opt/spark/conf
I have spark deployed on each slave in /opt/spark directory.
I can accesses HDFS using "hdfs dfs -ls hdfs://hdfs/" command in console,
without the need to specify active namenode address and port.
core-site.xml:
----------------------------------------------------------------------
<configuration>
<property>
<name>fs.default.name</name>
<value>hdfs://hdfs</value>
</property>
</configuration>
hdfs-site.xml:
----------------------------------------------------------------------
<configuration>
<property>
<name>dfs.ha.automatic-failover.enabled</name>
<value>true</value>
</property>
<property>
<name>dfs.nameservice.id</name>
<value>hdfs</value>
</property>
<property>
<name>dfs.nameservices</name>
<value>hdfs</value>
</property>
<property>
<name>dfs.ha.namenodes.hdfs</name>
<value>nn1,nn2</value>
</property>
<property>
<name>dfs.namenode.rpc-address.hdfs.nn1</name>
<value>namenode1.hdfs.mesos:50071</value>
</property>
<property>
<name>dfs.namenode.http-address.hdfs.nn1</name>
<value>namenode1.hdfs.mesos:50070</value>
</property>
<property>
<name>dfs.namenode.rpc-address.hdfs.nn2</name>
<value>namenode2.hdfs.mesos:50071</value>
</property>
<property>
<name>dfs.namenode.http-address.hdfs.nn2</name>
<value>namenode2.hdfs.mesos:50070</value>
</property>
<property>
<name>dfs.client.failover.proxy.provider.hdfs</name>
<value>org.apache.hadoop.hdfs.server.namenode.ha.ConfiguredFailoverProxyProvider
</value>
</property>
<property>
<name>dfs.namenode.shared.edits.dir</name>
<value>qjournal://journalnode1.hdfs.mesos:8485;journalnode2.hdfs.mesos:8485;journalnode3.hdfs.mesos:8485/hdfs</value>
</property>
<property>
<name>ha.zookeeper.quorum</name>
<value>master.mesos:2181</value>
</property>
<property>
<name>dfs.journalnode.edits.dir</name>
<value>/var/lib/hdfs/data/jn</value>
</property>
<property>
<name>dfs.namenode.name.dir</name>
<value>file:///var/lib/hdfs/data/name</value>
</property>
<property>
<name>dfs.datanode.data.dir</name>
<value>file:///var/lib/hdfs/data/data</value>
</property>
<property>
<name>dfs.ha.fencing.methods</name>
<value>shell(/bin/true)</value>
</property>
<property>
<name>dfs.permissions</name>
<value>false</value>
</property>
<property>
<name>dfs.datanode.du.reserved</name>
<value>10485760</value>
</property>
<property>
<name>dfs.datanode.balance.bandwidthPerSec</name>
<value>41943040</value>
</property>
<property>
<name>dfs.namenode.safemode.threshold-pct</name>
<value>0.90</value>
</property>
<property>
<name>dfs.namenode.heartbeat.recheck-interval</name>
<value>60000</value>
</property>
<property>
<name>dfs.datanode.handler.count</name>
<value>10</value>
</property>
<property>
<name>dfs.namenode.handler.count</name>
<value>20</value>
</property>
<property>
<name>dfs.image.compress</name>
<value>true</value>
</property>
<property>
<name>dfs.image.compression.codec</name>
<value>org.apache.hadoop.io.compress.SnappyCodec</value>
</property>
<property>
<name>dfs.namenode.invalidate.work.pct.per.iteration</name>
<value>0.35f</value>
</property>
<property>
<name>dfs.namenode.replication.work.multiplier.per.iteration</name>
<value>4</value>
</property>
<property>
<name>dfs.namenode.datanode.registration.ip-hostname-check</name>
<value>false</value>
</property>
<property>
<name>dfs.client.read.shortcircuit</name>
<value>true</value>
</property>
<property>
<name>dfs.client.read.shortcircuit.streams.cache.size</name>
<value>1000</value>
</property>
<property>
<name>dfs.client.read.shortcircuit.streams.cache.size.expiry.ms</name>
<value>1000</value>
</property>
<property>
<name>dfs.domain.socket.path</name>
<value>/var/run/hadoop-hdfs/dn._PORT</value>
</property>
</configuration>
--
View this message in context:
http://apache-spark-user-list.1001560.n3.nabble.com/Accessing-HDFS-HA-from-spark-job-UnknownHostException-error-tp25092.html
Sent from the Apache Spark User List mailing list archive at Nabble.com.
---------------------------------------------------------------------
To unsubscribe, e-mail: user-unsubscr...@spark.apache.org
For additional commands, e-mail: user-h...@spark.apache.org
