Re: mustella performance

2012-10-28 Thread Frédéric THOMAS
sage d'origine- From: Frédéric THOMAS Sent: Sunday, October 28, 2012 1:36 PM To: flex-dev@incubator.apache.org Subject: Re: mustella performance As a hint, I set the process priority of my shell lower than normal and can use windows almost normaly while the tests are running. -Message

Re: mustella performance

2012-10-28 Thread Frédéric THOMAS
As a hint, I set the process priority of my shell lower than normal and can use windows almost normaly while the tests are running. -Message d'origine- From: Michael Schmalle Sent: Thursday, October 25, 2012 11:38 AM To: flex-dev@incubator.apache.org Subject: mustella performance Hi,

Re: mustella performance

2012-10-25 Thread Alex Harui
In the compile phase, mustella uses every available core for compilation. In an -all run on my quad core Windows laptop, the entire run takes around 8 hours, but mostly that is due to the runner phase which runs each test SWF serially. To test build-script changes, I would just run ant main checki

Re: mustella performance

2012-10-25 Thread Peter Ent
I don't know why that would happen, but I can share with you my experience running the -all tests. I use a MacBookPro and on several occasions found the machine's CPU meter at 100% and barely responsive, running the tests very, very slowly. This was 1 or 2 months ago now. What we (Alex, Carol, and

Re: mustella performance

2012-10-25 Thread Jarosław Szczepankiewicz
maybe running visualvm from jdk will give answer 2012/10/25 Michael Schmalle : > Hi, > > What is going on with the mustella test suite that would make all quad core > cpus run at 100%? > > I know these tests are broad but, my computer got really unstable trying to > run -all. > > Fact is I don't h