This lab is intended to guide you installing Spark on a previously set-up Ambari cluster
- Having completed the lab 02 - Provision the environment
- MySQL JDBC driver must be setup on the node where Ambari server runs
If you have't installed MySQL JDBC driver on the Ambari server node, just jump on it and type:
$ sudo yum install mysql-connector-java -y
Then, let Ambari know about it
$ sudo ambari-server setup --jdbc-db=mysql --jdbc-driver=/usr/share/java/mysql-connector-java.jar
After having logged in into Ambari using the web UI, follow the visual instructions listed below:
Click on Add Service
Select Spark
Accept all the other components installation prompt
Distribute the new components on the cluster nodes (this image is just an example, you can spread them as you prefer)
Accept the default and click on NEXT
In order to proceed, type a password in the HIVE -> DATABASE section (not it down for later use)
Accept the proposal and click on PROCEED ANYWAY
Click on DEPLOY
After the installation, click on COMPLETE
Restart all the affected components as the warning message states
