Zeppelin HDP 2.4 SparkSql aggregate functions failing

[#1]

HDP 2.4 SparkSql on zeppelin aggregate functions are not working. I do a select * from table and it works. However when I do select count(*) from table or any another aggregate function (sum, etc) it fails. I am using Zeppelin version zeppelin-server-0.6.0.2.4.0.0-169.jar

[#2]

@Sunile Manjee Are you able to run select count from beeline?

[#3]

@Neeraj Sabharwal Yes.

[#4]

This seems to be a bug. I am opening a Jira.

[#5]

@Neeraj Sabharwal

I found the issue. I restarted the interpertur and rebinded the variables. works fine now.

[#6 · Accepted solution]

We had similar issues with the hive interpreter while trying to run aggregations and gouping by columns:

1. Hive interpreter cannot be declared directly in notebook by using %hive. Interpreter must already be set to hive

2. First line in editor must be blank and Hive QL statements must start on second line, otherwise a NullPointer exception will be thrown after you submit job. This threw us off. Somehow we started the statement on the second line and it executed without errors. Then, when we went back and put the statement on the first line, it failed again. Moved the statement to the second line, with the first line blank, and it executed without any errors. Strange.