> we all know about the Enable Performance Monitoring flag > in the CF debug settings, it throws out some rough > information about the processing time of each template > used the the request.
Well, actually, that's a different setting. The "Show Detailed Processing Time Breakdown" setting shows the exploded times per file. The only reason I bring this up is that, to determine application performance, you typically need to look at both the exploded benchmarking information and the information that's exposed in the performance counters. As Matt said, you really need to load test the application to get useful performance information. Your response, that you should test it with a single request, may lead you into a false sense of security - very often, you won't see bottlenecks until your application is under load - whether that load comes from a load test tool or from real users. In addition, load testing makes the bottlenecks more readily apparent, even if those bottlenecks show up in a single user test. To summarize the process within an email, you run a load test against your application with varying numbers of users. The numbers that you get from this will make the bottlenecks readily apparent. At this point, you then run a load test, and while you're doing that you run a single user, and that single user receives debug information, so you can see which scripts contain the bottlenecks. Then, within a single script containing a bottleneck, you use the GetTickCount function to divide the script into sections, to determine the location of the bottleneck within the script. Note that this is a bit of an oversimplification, of course, but it's close enough. > I can see roughly what is taking the time in this > request, but it is very rough, it would be nice to > see that for template bar.cfm 10ms was spent processing > tag <CFFOO/> or even just return something useful instead > of 0 for those very quick templates. It would also be > better if multiple calls to the same template were not > lumped togethor, for example in the above all the > templates in \PBCORE\ get run multiple times (4 times > if I count correctly) but they each only get one time. You're right, in that this stuff isn't exposed. To find out which tags are taking longer, you end up adding diagnostic code using GetTickCount, as mentioned above. Dave Watts, CTO, Fig Leaf Software http://www.figleaf.com/ voice: (202) 797-5496 fax: (202) 797-5444 ______________________________________________________________________ Signup for the Fusion Authority news alert and keep up with the latest news in ColdFusion and related topics. http://www.fusionauthority.com/signup.cfm FAQ: http://www.thenetprofits.co.uk/coldfusion/faq Archives: http://www.mail-archive.com/[email protected]/ Unsubscribe: http://www.houseoffusion.com/index.cfm?sidebar=lists
