GROUP BY not recognizing SELECT Elements?

[kdrakon]

Maybe I've been looking at my screen too long, but doesn't my group by contain the necessary elements to perform a grouping?

ksql> create stream rejected_orders with (kafka_topic='output.rejected_orders', value_format='json') as select * from orders where current_status = 'Rejected';

 Message
----------------------------
 Stream created and running
ksql> describe rejected_orders;

 Field          | Type
----------------------------------
 ROWTIME        | BIGINT
 ROWKEY         | VARCHAR(STRING)
 REQUEST_ID     | BIGINT
 ORDER_ID       | VARCHAR(STRING)
 PLATFORM       | VARCHAR(STRING)
 ORDER_TYPE     | VARCHAR(STRING)
ksql> select request_id, current_status from rejected_orders group by request_id;
Group by elements should match the SELECT expressions.
ksql>

This is running on the Docker image of ksql-cli, version 0.2 (080bbe2bfdbf).

[kdrakon]

I was actually trying to initially perform the following:

create table rejected_orders_table with (kafka_topic='output.rejected_orders_table', 
value_format='json') as select request_id, current_status from rejected_orders 
group by request_id;

[hjafarpour]

@kdrakon currently KSQL only supports GROUP BY clause with aggregate functions so you need to use at least one aggregate function such as count to use GROUP BY.

[kdrakon]

Thanks for clarifying @hjafarpour

[kdrakon]

Would
select request_id, count(request_id), current_status from rejected_orders group by request_id;
be expected to work then, because I'm getting the same error message?

[kdrakon]

Ah, I see, the above looks to be working if I remove the current_status field.