We have a new client, and currently we are performing stress test on the production like server with 100 concurrent http threads (using Jmeter). The problem is even though we ha
Comment by @JoseK triggers the whole bottleneck. our application is writing too many log. on one instance we were trying to reduce the amount of log in our log file, we notice the overall performance is better and higher CPU utilization.
We then try with log4k log level to "WARN", and wow! the application is utilizing 70% of all CPU cores.
We should check the Disk utilization before even asking the question. Thanks!