Hi,
I’m using a standalone deployment (Nifi 1.19.1) that runs directly on the
server.
I’ve noticed something tricky for the first time…
When executing a query with Nifi processor ExecuteSQL the maximum decimal
precision that it returns is 6, anything above that returns “0E-7”.
Query:
“
select
0.0000000::decimal(38,7) as Test1 -- NOT OK -> Returns 0E-7
,0.0000000::decimal(38,6) as Test2 – OK -> Returns 0,000000
,0.0000000::decimal(12,7) as Test3 -- Needed -> NOT OK -> Returns 0E-7
,0.0000000::decimal(12,6) as Test4 -- OK -> Returns 0,000000
“
ExecuteSQL . List Queue . View content:
"
Obj___avro.schemaö_{"type":"record","name":"NiFi_ExecuteSQL_Record","namespace":"any.data","fields":[{"name":"TEST1","type":["null","string"]},{"name":"TEST2","type":["null","string"]},{"name":"TEST3","type":["null","string"]},{"name":"TEST4","type":["null","string"]}]}_avro.codec_nulléõ
žSø®²›_yá‚_„_@__0E-7__0.000000__0E-7__0.000000éõ
žSø®²›_yá‚_„
"
After ExecuteSQL I use a ConvertRecord with AvroReader and CSVRecordSetWriter
but since the content is already messed up it converts to:
“0E-7; 0.000000;0E-7;0.000000”
Any idea if this is a known issue or how to solve it?
----- ExecuteSQL Specs
Database Connection Pooling Service: DBCPConnectionPool-Netezza -> IBM Cloud
Pak for Data System (PostgreSQL)
setSQL: select 0.0000000::decimal(38,7) as Test1, 0.0000000::decimal(38,6) as
Test2, 0.0000000::decimal(12,7) as Test3, 0.0000000::decimal(12,6) as Test4
Normalize Table/Column Names: false
Use Avro Logical Types: false
Compression Format: NONE
Default Decimal Precision: 10
Default Decimal Scale: 0
-----
Thanks in advance.
Tiago Sebastião
