Le mer. 4 janv. 2017 à 06:55, Rajnish kamboj <rajnishk7.i...@gmail.com> a écrit :
> Might be you are right as we are running our system under load test so we > are firing 200 requests with JMeter (loop forever). This may be the cause > of 100% CPU usage with 4 vCPUs machine. > In this case should we manually restrict our Lucene search threads in our > application OR there is some other inbuild Lucene mechanism to restrict the > threads to search Lucene index. > Well, you could but that would not make sense, 100% CPU usage is really the best you can get. Why would you like to make things worse artificially? That said, you might still want to restrict the number of threads that have access to the Lucene index (typically by using a fixed threadpool) to a bit more than the number of CPUs you have. This will still allow for 100% CPU usage while reducing context switching and causing Lucene to have fewer thread-local objects, which means potentially better throughput and memory usage.