Ugh, it looks like the yourkit probes were indeed the culprit. What a catch-22...sure makes it a lot harder to do find performance bottlenecks while under load.
In any case, thanks much for helping me find that which was right under my nose. On Wed, Dec 19, 2012 at 4:27 PM, Simone Bordet <[email protected]> wrote: > Hi, > > On Wed, Dec 19, 2012 at 10:14 PM, Jeff Segal <[email protected]> > wrote: > > Strangely, I'm seeing the same behavior in yourkit as with the previous > > tests, but these threads are blocking in BufferDataCache rather than > > ServletHolder. And just to confirm, what I'm seeing is not simply N > threads > > permanently blocked by one, but N threads spending most of their time > > blocked and briefly becoming runnable intermittently. > > Really strange. > Do you get bad behavior without any profiling tools ? > I ask because it would not be the first time that I see profiler > artifacts (i.e. code injected by profilers that causes performance > artifacts) that make profiling sessions completely useless. > > OS paging is also out of the picture ? > > Simon > -- > http://cometd.org > http://webtide.com > Developer advice, training, services and support > from the Jetty & CometD experts. > ---- > Finally, no matter how good the architecture and design are, > to deliver bug-free software with optimal performance and reliability, > the implementation technique must be flawless. Victoria Livschitz > _______________________________________________ > jetty-users mailing list > [email protected] > https://dev.eclipse.org/mailman/listinfo/jetty-users >
_______________________________________________ jetty-users mailing list [email protected] https://dev.eclipse.org/mailman/listinfo/jetty-users
