Thanks!
On Jun 24, 8:45 am, "Peter J. Farrell" <pe...@mach-ii.com> wrote:
> MercuryNewt said the following on 06/24/2009 07:12 AM:> Thanks Peter!
>
> > I have another question for you then; our application make extensive
> > use of CFCs and most of those were preloaded and stored in application
> > scope variables. However, when I move them over to Mach-II
> > properties, they caused that app to timeout after a server restart due
> > to the extreme amount of time it took to load all the CFCs into
> > memory. This wasn't the case with the previous Application scope
> > solution, but obviously due to the additional Mach-II logic to load
> > these CFCs into the framework, it just takes longer. We don't want to
> > deal with timeout issues in the production app, so what would you
> > recommend we do with these CFCs? Would you suggest we just call them
> > normally within the application as they are being used or do something
> > with a timeout setting?
>
> Well, it really shouldn't be an additional time for Mach-II to
> instantiate the CFCs into the property scope versus loading them in the
> application scope in the onApplicationStart. You might look at the
> MACHII_REQUEST_TIMEOUT variable in the bootstrapper (which your
> Application.cfc extends) so you can programatically increase the request
> timeout when the framework is loading. It's in the documentation on
> using the bootstrapper on the wiki.
>
> If you were lazy loading things into the application randomly over time
> (I'm sorta assuming based on your message) and not in the
> onApplicationStart at the get go, then yes it will take Mach-II longer
> to load at application start because you're taking the hit to load
> everything when the framework loads instead of over time. It's a better
> approach in my opinion because once the framework loads, everything is
> ready to go. So I would look at that bootstrapper setting.
>
> Also, if you are using Adobe CF8 on the Java JVM 1.6 series -- there is
> a serious class loader bug in the Java JVM that significant slows down
> CFC instantiation (up to 10 times slower if not more in certain cases).
> This is not a bug in CF, but a defect in the JVM -- just happens to be
> the CF8 ships with a 1.6 JVM that has this issue. Upgrading to a newer
> 1.6 JVM fixes the problem and speeds things up considerably. Here's a
> nice post about upgrading:
>
> http://www.coldfusionmuse.com/index.cfm/2009/6/11/upgrade.ColdFusion.JVM
>
> Extra suggestion: be sure to backup your jvm.config file first just in
> case you mess it up
>
> Here's a decent post on the class loader bug issue:
>
> http://corfield.org/blog/index.cfm/do/blog.entry/entry/Java_6_and_Col...
>
> (Example of the bug 6.5 minutes to load an application on a 1.6 JVM with
> the class loader bug and 50 seconds on a 1.6 JVM when the issue was fixed)
>
> Best,
> .Peter
--~--~---------~--~----~------------~-------~--~----~
You received this message because you are subscribed to Mach-II for CFML list.
To post to this group, send email to mach-ii-for-coldfusion@googlegroups.com
To unsubscribe from this group, send email to
mach-ii-for-coldfusion-unsubscr...@googlegroups.com
For more options, visit this group at
http://groups.google.com/group/mach-ii-for-coldfusion?hl=en
SVN: http://greatbiztoolsllc.svn.cvsdude.com/mach-ii/
Wiki / Documentation / Tickets:
http://greatbiztoolsllc.trac.cvsdude.com/mach-ii/
-~----------~----~----~----~------~----~------~--~---
