USER <--> Proxy <--> UMP Server
Problem: after develop a dashboard we published it to a customer. This customer has a NOC with a big screen with this dashboard so, for him this is a very important feature.
We have a proxy and all configurations were followed according CA's documentation. This proxy and all the environment worked well until this dashboard deployment. Since then we are having some problems with the close_wait connections in the UMP server. We have evidences about the access to this page matches the raise of close_wait connections.
You can see what's happening in the image below. The first image is the Google Analytics and pageviews, the second one simply a Performance Report with close_wait connections number in UMP server. Just a note: when the number of close_wait connections goes to 0 is the moment when we must to restart wasp probe.
After some searches I suppose that this could be a application problem. Am I right? Anyone else got/get this problem?
I have an opened case but I wonder if anyone has the same problem.