quick set up history for submitting a spark job
just a simple history to remind myself.
1 ls
2 tar xzvf spark-1.2.1-bin-cdh4.tgz
3 cd spark-1.2.1-bin-cdh4
4 ls
5 cd conf
6 ls
7 cp spark-defaults.conf.template spark-defaults.conf
8 vim spark-defaults.conf
9 vi spark-defaults.conf
10 sudo yum install -y httpd && sudo systemctl start httpd
11 cd ../..
12 ls
13 cp spark-1.2.1-bin-cdh4.tgz /var/www/html/
14 sudo cp spark-1.2.1-bin-cdh4.tgz /var/www/html/
15 cd spark-1.2.1-bin-cdh4/conf/
16 vi spark-defaults.conf
17 ls
18 cp spark-env.sh.template spark-env.sh
19 vi spark-env.sh
20 cd ../
21 ./bin/spark-submit --class org.apache.spark.examples.SparkPi lib/spark-examples*.jar 10
22 history
only three config settings were added. inside spark-default.conf
:
spark.master mesos://zk://zookeeper0:2181/mesos
spark.executor.uri http://submitter/spark-1.2.1-bin-cdh4.tgz
inside spark-env.sh
:
#!/bin/sh
export MESOS_NATIVE_LIBRARY=/usr/lib/libmesos.so
happy day.