Good Morning,

We have an application which terminates their own SSL, therefore we utilize TLS 
passthrough in the route configuration. This is our preferred method of 
communicating with this particular application. This enforces haproxy to 
operate using tcp mode, which the balance method is hard coded to source [1].

The problems comes in that we’ve got a front door to openshift, so all traffic 
hits a load balancer external to openshift. Due to the source reading the IP of 
that external loadbalancer, all traffic gets routed to the same pod.

Looking through various PR’s and commits I cannot find the reason why source 
was chosen, but did see where last year, there was a glimpse of this being 
touched later on. Would anyone be able to share why this particular balance 
type was chosen? roundrobin seems like a better choice for us in our particular 
situation. I also feel like having an external load balancer to openshift is 
not uncommon and would love to see this be configurable.

  *   [1] 
https://github.com/openshift/origin/blob/master/images/router/haproxy/conf/haproxy-config.template#L237-L242
  *   [2] https://cbonte.github.io/haproxy-dconv/configuration–1.5.html#balance


--
John Skarbek
_______________________________________________
users mailing list
[email protected]
http://lists.openshift.redhat.com/openshiftmm/listinfo/users

Reply via email to