FYI, ResourcePool is no longer in MG:U.
On 12/12/06, Jared Rypka-Hauer <[EMAIL PROTECTED]> wrote:
Actually... I wonder... There was an issue floating around where relocating ModelGlue event handlers would cause the system to use up it's memory. I wonder if, since the Reactor objects use more RAM (just because of everything they do), and the MG events relocate, and everything is on just the right F of MG, Reactor and the JVM, if that's not the cause of the issue. I'm basically wondering if the difference is using MG with the ResourcePool, relocating events, and Reactor Record objects as beans isn't basically the cause of most of the issue... Anyone fit that pattern? J ------------------------------------------------ Jared C. Rypka-Hauer Continuum Media Group LLC http://www.web-relevant.com Member, Team Macromedia - ColdFusion "That which does not kill me makes me stranger." - Yonah Schmeidler On Dec 12, 2006, at 4:06 PM, Lee McNeil wrote: Late to reply to this, but we've had such major memory leak issues with and around reactor (admittedly with model-glue too) in production that we unfortunately had to remove it and replace with much simpler CFC. With a JRun instance with 1.5GB we were dealing 25,000 CFM request per day and running out of memory and getting java out of memory errors by the end of the day. Removing Reactor reduced the problem significantly (memory peaked @ 40% but still ended @20% usage), and updating model-glue: unity beta1 to the latest subversion checkout solved the issue completely (now averaging @ 10% usage). Sorry that I can't offer constructive advice here, but we had to remove the issue ASAP, and building something very much like reactor but very very much simpler was the answer for us. Still using reactor for low impact sites, and really hope this gets solved.� However for higher load sites, I don't think it quite makes the mark.� Admittedly, I DO NOT THINK IT IS REACTORS FAULT.� I think that as pointed out by others it is a problem with CFCs and garbage collection is CFMX itself - which is a shame since reactor is a really a great product. Lee. On 12 Dec 2006, at 18:04, Daniel Gracia wrote: Hey guys, First of all, thanks for your incredible work for the framework. I just wanted to ask around if someone has also had the same problem with Reactor: I use� Mach-ii - reactor architecture and I have, every 12 hours, to restart my CF server (JRUN) because of the RAM level of the process has reach the maximum (600MB) and the site doesn t work anymore. Same kind of project using mach-ii but not reactor has not this kind of problems. �have the project isolated in his own Instance. Does anybody had same kind of problems? I don t� wish it off course but it could be cool if someone had experience the same and had solved or had any idea where to start fixing it. thanks -daniel. -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- Reactor for ColdFusion Mailing List [email protected] Archives at: http://www.mail-archive.com/reactor%40doughughes.net/ -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- Reactor for ColdFusion Mailing List [email protected] Archives at: http://www.mail-archive.com/reactor%40doughughes.net/ -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- Reactor for ColdFusion Mailing List [email protected] Archives at: http://www.mail-archive.com/reactor%40doughughes.net/ -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- --
-- Get Glued! The Model-Glue ColdFusion Framework http://www.model-glue.com -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- Reactor for ColdFusion Mailing List [email protected] Archives at: http://www.mail-archive.com/reactor%40doughughes.net/ -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- --
<<attachment: Community_Experts_Logo.jpg>>
