Can you provide a stack trace of a thread in such a state?
Also, what URLs are you loading in jMeter to reproduce.
Thanks!
~Brad
Can you provide a stack trace of a thread in such a state?
Also, what URLs are you loading in jMeter to reproduce.
Thanks!
~Brad
At the moment the url is just /blog and the users or threads is set at 100, so its very low.
It just hangs, now its the same as when I hit the pages on my own. It will eventually just hang. Now I did in one of my tests get a hibernate query timeout, but that was a once off.
I have been trying all different things, and I am not sure what to say. ColdFusion has 2gb for its GC it is using about 10% of this at any given time, even when I throw it 1000 concurrent users. But the requests in FusionReactor blow out to 300secs thats right each request gets longer and longer.
I noticed the same thing with FR if I hit fresh 10 times, the requests get longer, and eventually you can hit the jackpot and it will hang as well.
I know something is not right here, so I decided to put it up on a production server. And it runs better, but I can still get it to run out to 15 secs for a page refresh, but just hitting the browser refresh. However I was not able to get CF to hang, which is kinda good because my host would kill me
As for stack traces, from where? The logs report nothing when it hangs, but I can tell you that when you try to kill ColdFusion service in Windows, it can take about 15minutes to kill it.
Here is FR longeest request info