Support Questions
Find answers, ask questions, and share your expertise
Alert: Welcome to the Unified Cloudera Community. Former HCC members be sure to read and learn how to activate your account here.

Unable to connect Spark SQL via JDBC driver

Unable to connect Spark SQL via JDBC driver

New Contributor

We are not able to connect Spark SQL through JDBC driver. Usually we use 10005 port to connect spark SQL in previous Horton works ( normal). But now we are using Hortonworks data cloud from AWS market place. ( HDP 2.5 and Apache Spark 2.0)

Your doumentation only suggests to connect via spark shell :

but for hive, there are multiple options, like jdbc and odbc :

Please help me to connect spark via jdbc from work bench for hortonworks data cloud.


Re: Unable to connect Spark SQL via JDBC driver

New Contributor

Hello @Aravind J L

With Hortonworks Data Cloud the Spark Thrift Server is not a pre-configured component. You can manually set it up using the following steps:

1. Login to Ambari on the Spark cluster type using your admin credentials

2. Make the config changes required if any to ensure the Spark Thrift Server is sized correctluy and you have adequate capacity on the cluster

3. Select the node on which you want to run the Spark Thrift Server and add it as a component

4. Either create an ssh tunnel or modify the security group settings for that node to allow access to port 10005 from the machine where your jdbc client is running.

Please let us know if you need further instructions on the steps.

Thanks! ram

cc: @Dominika Bialek


Re: Unable to connect Spark SQL via JDBC driver

New Contributor

Hi @Ram Venkatesh

Thanks for your update. I believe it is not that easy to connect to spark as you said, since I tried the same for hive. The steps you mentioned may work for the normal hortonworks setup not the cloud one. Finally realised, I have to follow the documentation by using SSL port 8443 and it perfectly worked for me to connect hive.

Please have a look:

I guess I have to connect spark the same way I did for hive . On ambari the spark port is 10016 ( my bad I mentioned 10005) and I opened the same port on security group as per your steps. But still unable to connect. The master node is already added as a component when I checked.

Please help me.

Don't have an account?
Coming from Hortonworks? Activate your account here