Why does one nginx controller point to another via default backend?

Question

Nginx

there is a kubernetes server in the Amazon cloud (not eks) - I'm trying to understand and understand the nginx ingress controller configuration - 2 things are not entirely clear. Here is an excerpt from the nginx controller configuration (from nginx.conf)

{
server_name service1.cluster.ourcompany.ru
listen 80;
..........
..........


proxy_pass http://upstream-default-backend;
}

{
server_name service2.cluster.ourcompany.ru
listen 80;
..........
..........


proxy_pass http://upstream-default-backend;
}

upstream upstream-default-backend {
  # Load balance algorithm; empty for round robin, which is the default
    least_conn;
    keepalive 32;
    server 100.116.8.9:8080 max_fails=0 fail_timeout=0;
}

1. Why do you need to additionally create a default backend? I looked, in this case, another nginx controller is running on the IP address 100.116.8.9 (there is no console access to it - a greatly truncated version).
2. Do I understand correctly that routing in a kubernetes cluster first comes to the load balancer, then goes to the nginx ingress controller pod, and then cools down through the corresponding service to one of the pods?