DO NOT REPLY TO THIS EMAIL, BUT PLEASE POST YOUR BUG· RELATED COMMENTS THROUGH THE WEB INTERFACE AVAILABLE AT <http://issues.apache.org/bugzilla/show_bug.cgi?id=17762>. ANY REPLY MADE TO THIS MESSAGE WILL NOT BE COLLECTED AND· INSERTED IN THE BUG DATABASE.
http://issues.apache.org/bugzilla/show_bug.cgi?id=17762 [EMAIL PROTECTED] changed: What |Removed |Added ---------------------------------------------------------------------------- CC| |[EMAIL PROTECTED] ------- Additional Comments From [EMAIL PROTECTED] 2005-06-17 19:47 ------- (In reply to comment #8) > I think people should just avoid JNI for now. It doesn't give any performance > boost anyway. Huhh?!? Sorry, I just have to comment.. Please compare the performance (should I say penalty) of *not* using JNI... The last time I tested (and also just now), shoveling long byte streams through ServletResponse.getOutputStream() __completely___ brings the native apache server to a grinding hault (full CPU usage, e.g. D-O-S). My test on a 2GHZ box of a file download via FileInputStream -> getOutputStream() (4096 buffer) *pinned* the CPU (the Apache process, specifically), yet could only sustain a mere 1-2 MBytes/second. When I compared jni performance a year or two ago (when I could get it to work), this performance hit was marginal or negligible as I recall. And other JNI based servers (right now) also have no such performace hit. I concede that this isn't neccsarily the socket channels architecture's fault (perhaps it just need to be buffered/optimized on the mod_jk side?), but still... -- Configure bugmail: http://issues.apache.org/bugzilla/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are the assignee for the bug, or are watching the assignee. --------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
