The Data Studio

Hive Insert Statement Is Non-standard

Tested Using Hortonworks Data Platform (HDP) Sandbox, Release 2.5 with Hive 1.2.1 and Hive 2.1.0

The standard SQL insert statement is of the form:

      insert into <table> (<column list>) values (<value_list>);
    

There are several Hive SQL variants. The one most like the one shown above is:

      insert into <table> values (<value_list>);
    

You cannot specify a column list in Hive, so you have to provide values for all the columns, and in the order in which they are defined in the table. The definition order should not be significant, but it is in Hive SQL.

Of course, individual inserts are not really suited to the Big Data tools because they are not transactional systems. The Big Data tools are basically for Data Warehousing, handling high volume data loads and then queries. They are designed to handle text and other "unstructured" data. That would all be OK if the vendors were not pushing them as a general replacement for relational databases.