I’m running a Hive query over JDBC in a Java app that I wrote. I want to be able to turn off task conversion as I am looking to stream the data back. I thought I could do that by using the following JDBC URL: jdbc:hive2://192.168.132.128:10000/default?hive.fetch.task.conversion=none. My SQL statement has an ORDER BY in it, but other than that it is just a straight up “SELECT * FROM <table name>”. The task conversion is still occurring, and that causes the job to blow up, because the table has 30+ million rows in it. I just want to stream the data so I can take advantage of the fetch size and read the data in batches.
↧