How to understand the meaning of TCP data statistics?

Question

Computer networks

I configure virtual servers through Ansible - that is, all their configs are identical. But on some servers, as on the graph, the statistics data run into some kind of limit. How to understand where this limit comes from? How to fix so that there is no such shelf.

Answer 1

check first with netstat or ss to which software you have the vast majority of TIME_WAIT connections, and then you can think further.