I try work on the R interpreter in zeppelin. I am using HDP 2.5

[#1]

I follow this link https://community.hortonworks.com/questions/55139/no-r-interpreter-in-zeppelin-hdp-25.html. to resolve this error but i am still getting the error "I/O error on POST request for "http://sandbox.hortonworks.com:8998/sessions": Connection refused (Connection refused); nested exception is java.net.ConnectException: Connection refused (Connection refused)" when I run these command using %livy.spark.r

%livy.sparkr
foo <- TRUE
print(foo)
 

[#2]

@Narasimma varman

For %livy.sparkr - this looks like a problem connecting to livy server. Try to restart livy from Spark service.

It is also useful to check the log files:

- interpreter log - usually located in /var/log/zeppelin folder

- application log - %livy.sparkr triggers Spark job in Yarn. So, check the application log for more details

I hope that gives you some idea of why the process fails.

[#3]

Whenever I try to set an environment variable for R interpreter by editing conf/zeppelin-env.sh in SPARK_HOME. Once I restart zeppelin service, SPARK_HOME value automatically change into its own default value(/usr/hdp/current/spark-client). How to retrieve this issue

[#4]

Do you manage your cluster through Ambari? If so, make the change from that site and restart services from there.