We have a situation where a COGNOS user is submitting many queries to our live service and TASM is, correctly, placing them in the delay queue, once our concurrency throttle has been reached but the user is then "hogging" the delay queue with upto 50 queries waiting to be serviced. This is clearly unfair to other responsible users.
My question is, not being an expert in COGNOS, is if we were to introduce an additional throttle to reject any query submitted in excess of say, 10, would any subsequent query be lost and cause the COGNOS application to fail? or would the query retry?
I suspect the latter is the case and may have to be enabled but any confirmation would be appreciated.