I'm trying to squeeze every thing i can from my ec2 account : I have a tomcat 7 installed on a ec2 large instance , i've been running some tests on it and right now i'm maxing out with apache benchmark at -c500 -n10000 at around 400ms , on my monitoring i see that at no time do i pass 50% on cpu usage , i've raised the maxthreads on the connector to 32000 and set the ulimit to unlimited , i'm also trying to run with visualvm and understand where i can improve my code but it seems to me that the machine can hold another tomcat listening on a different port with the same application and if i will just configure nginx to loadbalance between them i will get a serious lift in concurrent users, does that make any sense (all hits are stateless of course) ? the benchmark is not for static content , those number are for serve side logic i'm running a spring rest api application . thanks
|
feedback
|
