How does openshift multiple pods loadbalancing work?

In openshift I have a project with one Spring Boot application.
When I send a HTTP request, the Spring Boot application logs a message. I can see this message in the log of the pod in openshift.

When I scale the pod to two and send the same HTTP request, only one of the pods receives this request and logs it. This also happens when I send a 1000 requests using JMeter: only one pod receives all requests.

Doesn't openshift balance the load?

I tried sending multiple requests and expected these requests to be balanced over the 2 pods.

Pod traffic load balancing, is not an OpenShift-specific mechanism. This is a core kubernetes functionality, handled by Service objects. Consider Services as your network path into pods. The Service will handle load balancing traffic to multiple pods. Services will balance network traffic within the cluster, and when you need to expose this network traffic externally you rely on Routes (OpenShift-specific) or Ingresses (kubernetes-generic).

Without knowing more specifics about how JMeter is accessing the application hosted in the pods, it's difficult to give a solid answer. However, there are several functionalities that Services and Routes expose to address pod network load balancing.

There are various Route-specific annotations that can control network "stickiness". Probably the two most relevant ones are haproxy.router.openshift.io/balance and haproxy.router.openshift.io/disable_cookies. There are also Service session affinity options as well.

All of these components work together to get network traffic into your application. And only you know the exact path that you're following and able to adjust any necessary configurations as such.