I have the following Spark SQL query: SELECT column_name, * from table_name;
I have multiple spark clusters, this query has been running daily on all of the clusters. After a recent zeppelin redeployment, it fails on just one of the clusters with the following exception: AnalysisException: cannot resolve '`column_name`' given input columns The columns listed after the exception does include the "column_name". The data and schema is the same on all clusters. The query does seem to work using beeline and has been only failing when executed through zeppelin. I should point out there are several zeppelin instances that are able to execute this query, only one that fails, which is why I reached out to the zeppelin community. Also, I reached out to Spark users but was unfortunately unable to get any help. I'd appreciate any insight you can provide.
