2

The relevant part of my /etc/haproxy/haproxy.cfg is:

global
    maxconn 30000
    ...

defaults
    ...

frontend frontend_for_all_sites
    maxconn 22000
    mode http
    bind *:80

    acl acl_hostname_www hdr_dom(host)    www.example.com
    acl acl_hostname_static hdr_dom(host) static.example.com
    use_backend www_backend      if acl_hostname_www
    use_backend static_backend   if acl_hostname_static

backend www_backend
    server www 127.0.0.1:9090 maxconn 500

backend static_backend
    server s 127.0.0.1:8080 maxconn 5000

So, I have 2 backends with only 1 backend each, without load balancing, but just proxying requests to different backends based on the request's hostname.

On the www backend I've set maxconn 500 and on the static backend I've set it to 5000.

The stats page now looks like this:

enter image description here

My question is what does the "Backend Limit" of 2200 for both backends represent? I understand that this value is 10% of the 22000 maxxconn on frontend_for_all_sites. Does this 2200 mean that the maximum amount of connections that the backends will process is 2200 and above that haproxy will return a 503?

2

My question is what does the "Backend Limit" of 2200 for both backends represent? I understand that this value is 10% of the 22000 maxxconn on frontend_for_all_sites.

2200 in Backend Limit is the value of fullconn which is by default 10% of maxconn of frontend.

Does this 2200 mean that the maximum amount of connections that the backends will process is 2200 and above that haproxy will return a 503?

No, the maximum amount of connections is sum of maxconn

Your Answer

By clicking "Post Your Answer", you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.