I have this simple query.

TABLE FILE CAR
  SUM CNT.DST.CAR
END

Returns 10

TABLE FILE CAR
  SUM CNT.DST.CAR/D6
END

Returns 1

SQL produced: ( when running on a real database, not CAR )

SELECT
 COUNT(DISTINCT  COUNT(DISTINCT T2."WTF_ID"))(INTEGER),
 SUM(T2."WTF_ID")(INTEGER)
 FROM
 INGRESS.P06;

With these warnings ( no errors, it just continues )
FOC2589 - AGGREGATION DONE BUT MAY PRODUCE INCONSISTENT RESULTS
FOC2612 - OVERFLOW MAY OCCUR WHEN SUMMING AN I2 FIELD WITH USAGE I, D OR P

A count distinct over a count distinct? Seems like something got lost in translation.


Is this adapter specific?

Greets,
Dave

A distinct count of distinct counts? Peculiar!

That looks a lot like the a bug I reported in 8.1.03 where the generated query contains a syntax error because it contains 2 distinct clauses.

Unfortunately, tech support seems to be offline at the moment, or I would have had more details.

I tried.
In both cases I get 10.
However with the format the TITLE is CAR,COUNT; without the format it is COUNT,DISTINCT,CAR.
Go figure!

I've never seen that with Oracle.

I just tested SUM CNT.DST.CAR/D6 in WF 8105 using both Oracle and SQL Server and got the following correct translation:

AGGREGATION DONE ...
SELECT
COUNT(DISTINCT T1."CARS")
FROM
HR.CAR_ORA T1;

AGGREGATION DONE ...
SELECT
COUNT(DISTINCT T1."CARS")
FROM
CAR_MSS T1;

Unfortunately, I don't have access Ingress to test.