We have two separate clusters (one for Liferay portal, another - for Tableau BI). They are independent of each other. Now I want to use load balancer in high-availability manner (2 servers with heartbeat). The question is the following: is it possible to configure apache mod_jk (or haproxy) so that it do load balancing separately for both clusters? Or we should have independent load balancer for each cluster?
Thanks
Both approaches are possible and It would depend lastly in your own preferences. Thus, I can think just a only advantage in using two load balancers, which is the ability to manage two load balancers separately using status worker.
This way, you could stop one of the workers in just one loadbalancer so one of your applications will not receive requests in one worker node whilst the other application would continue handling incoming requests. I see this approach like more production grade.