No increase in Performance after horizontal scaling !!
I am doing performance for one of company application. With one node, we achieve xxx TPS and then decided to scale the application.
Injector tool is Grinder, Web App Server is WebSphere, IBM HTTP Server used as reverse proxy. The Cluster nodes are on same server but different VM's with 16GB RAM & 16vCPU's.
DB, Grinder are all separate VM's.
Even after doing tuning of WebSphere and worker module parameters of proxy, I am not able to go beyond the xxx TPS that I got with single Node.
Any ideas, will be helpful.
PS : Network Latency and VM resrouces are not the bottleneck.