Dave, this is such a helpful reply for anyone looking to apply Performance Tuning to ColdFusion apps. I know you neither asked for nor indeed need my endorsement but I just wanted to amplify to others the worth of your response, having gone through these exercises many, many times myself.
Kind Regards - Mike Brunt, CTO Webapper http://www.webapper.com Downey CA Office 562.243.6255 AIM - webappermb "Webapper - Making the NET work" -----Original Message----- From: Dave Watts [mailto:[EMAIL PROTECTED]] Sent: Tuesday, July 30, 2002 6:07 PM To: CF-Talk Subject: RE: CFMX Taking all CPU Resources? > Code Code Code.... I actually explained in detail what the > code was doing...earlier Well, you certainly didn't provide anything as clear as what you've provided here. Even here, you'd be better off posting the actual code, I think. So, yes, "Code Code Code". If you want answers to programming questions, code is helpful. This is starting to remind me of "Car Talk", the radio show in which people call up, and say that their car is making this noise or that noise, and the hosts try to guess what's wrong with it, but they really have no idea of course. You know, you may well be right about your suspicions of CF MX. However, you're really going about this the wrong way, I think. Instead of taking what appears to many to be a belligerent, uncommunicative attitude, you might be better served by taking a more cooperative approach, and providing more information about the specifics of your problem. There are plenty of people here who will be willing to help you out, if you do so. > 1.<cfquery name="sQry" ...> > returns... ID,Title,Subject,Body,Published > date,Version,Published_by etc. > 2.Now Serialize this and write it to database in ONE FIELD/COLUMN. > 3.Create a dynamic page... that gets the content from the database. > 4. Deserialize ie WDDX2CFML > 5.Display the content... <cfoutput>#OutString.Title#</cfoutput> > 6.LOAD TEST with MS STRESS Tool/Load Runner or some else > 30-50 users.... in > CFMX...hitting > the same pages for different Content.(Url?cid=10,Url?cid=12 .....) > 7.Do the same for CF5.0... OK. Here's a possible plan for you. You've divided up your script into the above parts. You might wrap timers around each part, using GetTickCount, then output each part's elapsed time at the bottom of your page - a technique straight from "CF Performance Tuning 101". Then, while running your app under your maximum test load - the maximum number of users that still returns a tolerable response, according to your response time ceiling - view the application with a browser and look at those values. Then, do the same for CF 5. Then, compare which parts were slower than which other parts, if any, or whether everything was proportionally slower. For example, you might find that the part that's much slower in CF MX is the WDDX2CFML part, or you might find it's your database query. Once you've found that, you can more easily create a reproducible test case, like Jochem did earlier. You can then post that, and others can verify your results. But if you just keep complaining that CF MX is slower, then you're unlikely to get any help with your real problem. This would be a shame for you, since you'd have wasted so much time, and it would be a shame for others, if there's a real problem to be found. Good luck! 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/cf-talk@houseoffusion.com/ Unsubscribe: http://www.houseoffusion.com/index.cfm?sidebar=lists
