Stream Processing: filtering by the aggregation field, for example "HAVING count_QuestionType > 10"?

[canob]

Hi everyone,

Another simple question about Stream Processing: is it possible to filter by the defined aggregation field on the SQL Query?
For example:

[STREAM_TASK]
    Name dns_request_type_high_rate_NULL
    Exec CREATE STREAM dns_request_type_high_rate_NULL WITH (tag='dns_request_type_high_rate_NULL') AS SELECT ip AS ipaddress, qtype AS qtype, COUNT(qtype) AS count_qtype FROM TAG:'dns' WINDOW TUMBLING (60 SECOND) WHERE qtype = 10 HAVING count_qtype > 10 GROUP BY ip, qtype;

I want to obtain only the aggregated entries in 60 seconds with a COUNT greater than 10, but is not working.

Maybe I can do it with two STREAM TASKS:

[STREAM_TASK]
    Name dns_request_type_NULL
    Exec CREATE STREAM dns_request_type_NULL WITH (tag='dns_request_type_NULL') AS SELECT ip AS ipaddress, qtype AS qtype, COUNT(qtype) AS count_qtype FROM TAG:'dns' WINDOW TUMBLING (60 SECOND) WHERE qtype = 10 GROUP BY ip, qtype;
[STREAM_TASK]
    Name dns_request_type_high_rate_NULL
    Exec CREATE STREAM dns_request_type_high_rate_NULL WITH (tag='dns_request_type_high_rate_NULL') AS SELECT * FROM TAG:'dns_request_type_NULL' WHERE count_qtype > 10;

But I want to know if I can do it with only one, to get better performance.

Thanks in advance four your help!

[koleini]

Hi @canob, currently, our SQL syntax does not support the HAVING clause, and as you said, using two Stream Tasks is the correct way to achieve the same result.