On Sunday, 7 December 2014 10:02:02 UTC-8, Alan Williamson wrote: > > Magnus, would like to see some code and a little more detail. > > - How many rows is returned? > - How many columns in each row? >
Basically, what I was was taking a cfdump at the very end of the page to check the performance of certain queries. Then I'd take the outputted SQL with variables included and put it into MySQL Workbench to play with it and see about tuning it to run faster. What I would see was that the query would run very fast already in Workbench already. The query with the biggest variance joined two tables and returned between 1 and 6 rows. It did have a lot of columns: about 25. None of the columns had a lot of data in them. But only about half of them were required and when I removed all of the unneeded ones, the query time as shown by OpenBD was cut by about 40 to 50%. What exactly is the OpenBD query time showing us. I assumed it got the query execution time from MySQL but perhaps, logically, it is timing the full round trip from CFML to MYSQL and back again. > > > > On 06/12/2014 12:28, Magnus wrote: > > For the query times I am basing the OpenBD supplies times on the number > provided in the query object dump. I would have thought thatthat that was > calculated before any rendering. > > Magnus > > On Saturday, 6 December 2014 01:32:46 UTC-8, Al Holden wrote: >> >> Both the CFDUMP and the Debug output can be resource intensive. Those >> two can inject a good amount of rendering script and css. >> >> As is the case with so many other experiments, the tools we use - to >> observe the outcome - can change the outcome itself. >> >> See if you can use other profiling methods, maybe CFTIMER ? >> >> I'd be interested to see if your numbers get any better. >> > > -- -- online documentation: http://openbd.org/manual/ http://groups.google.com/group/openbd?hl=en --- You received this message because you are subscribed to the Google Groups "Open BlueDragon" group. To unsubscribe from this group and stop receiving emails from it, send an email to [email protected]. For more options, visit https://groups.google.com/d/optout.
