Livy Job Server Implementation

Livy Job Server Installation on Horton Works Data Platform (HDP 2.4):
Download the compressed file from
http://archive.cloudera.com/beta/livy/livy-server-0.2.0.zip

Unzip the file and work on configuration changes.

Configurations:

1	Add 'spark.master=yarn-cluster' in config file '/usr/hdp/current/spark-client/conf/spark-defaults.conf'

livy-env.sh
Add these entries

1 2 3 4 5 6 7 8

export SPARK_HOME=/usr/hdp/current/spark-client export HADOOP_HOME=/usr/hdp/current/hadoop-client/bin/ export HADOOP_CONF_DIR=/etc/hadoop/conf export SPARK_CONF_DIR=$SPARK_HOME/conf export LIVY_LOG_DIR=/jobserver-livy/logs export LIVY_PID_DIR=/jobserver-livy export LIVY_MAX_LOG_FILES=10 export HBASE_HOME=/usr/hdp/current/hbase-client/bin

log4j.properties
By default logs roll on console in INFO mode. It can be changed to file rolling with debug enabled with the below configuration:

1 2 3 4 5 6

log4j.rootCategory=DEBUG, NotConsole log4j.appender.NotConsole=org.apache.log4j.RollingFileAppender log4j.appender.NotConsole.File=/Analytics/livy-server/logs/livy.log log4j.appender.NotConsole.maxFileSize=20MB log4j.appender.NotConsole.layout=org.apache.log4j.PatternLayout log4j.appender.NotConsole.layout.ConversionPattern=%d{yy/MM/dd HH:mm:ss} %p %c{1}: %m%n

Copy the following libraries to the path '<LIVY SEVER INSTALL PATH>/rsc-jars'

1 2 3 4 5 6 7 8

hbase-common.jar hbase-server.jar hbase-client.jar hbase-rest.jar guava-11.0.2.jar protobuf-java.jar hbase-protocol.jar spark-assembly.jar

Integrating web Application with Livy job server:
Copy the following libraries to tomcat classpath (lib folder)

1 2	livy-api-0.2.0.jar livy-client-http-0.2.0.jar

Servlet invokes the LivyClientUtil

LivyClientUtil:

1 2 3 4 5

String livyUrl = "http://127.0.0.1:8998"; LivyClient client = new HttpClientFactory().createClient(new URL(livyUrl).toURI(), null); client.uploadJar(new File(jarPath)).get(); TimeUnit time = TimeUnit.SECONDS; String jsonString = client.submit(new SparkReaderJob("08/19/2010")).get(40,time);

SparkReaderJob:

1 2 3 4 5

public class SparkReaderJob implements Job<String> { @Override public String call(JobContext jc) throws Exception { JavaSparkContext jsc = jc.sc(); } }

References:
http://livy.io/
https://github.com/cloudera/livy