Created on 01-11-2018 02:54 PM - edited 08-18-2019 12:57 AM
I need to create checksums in my ETL process. Therefore I need to sum up a numeric column in my JDBC-connected table(s). Copying the data works fine so far, but I was not able to find the correct NiFi processor for this aggregation task.
How can I calculate the sum of a table column? Which processor to use (is there something similar for the SelectHiveQL processor, which works fine for the same aggregation task)? And how to compare these two results (from two different tables) to check if the sums are equal?
I already tried to use a ExecuteSQL processor but there I get an SchemaParseException:
I'm using NiFi 1.4.0 and Hive 1.2.1.
Created 01-11-2018 04:26 PM
Can you create alias name for the aggregate column sum(menge) as menge then the processor will work.
Change the SQL Query as
Select sum(menge)menge from <table>
Created 01-11-2018 04:26 PM
Can you create alias name for the aggregate column sum(menge) as menge then the processor will work.
Change the SQL Query as
Select sum(menge)menge from <table>
Created 01-12-2018 10:54 AM
Yes, that was the solution! Thank you very much!
Everything works fine with the following statement now:
SELECT SUM(menge) menge FROM mytable