Page 2 of 2

Re: Program execution statistics

PostPosted: Tue Jun 09, 2009 7:51 pm
by Bill Dennis
dick scherrer wrote: User ax713r cannot do anything that would prevent smf records for userid=ax713r from being written can they?
That's correct. I don't know of a method where a single user could stop their own records.

Re: Program execution statistics

PostPosted: Wed Jun 10, 2009 2:37 pm
by jeroc
Hi Dick,

Yes, some tasks are big and require more time than others. However, if a program contains poor performance constructs and if it does not take long time, then it is not necessary to spend time to optimize it. But if it takes long time, then it would be interesting to work on it. In conclusion, using execution statistics alone are not sufficient to say that a program should be improved. Same thing with identification of poor performance constructs. It would be better to correlate both.