How calculate HAProxy roundrobin weights

haproxyload balancing

Suppose I have two Redis database and want to distribute 70% of requests to DB1 and the other 30% to DB2. How can i calculate the weight parameter?

global
  maxconn     20000
  log         127.0.0.1 local0
  user        haproxy
  chroot      /usr/share/haproxy
  pidfile     /run/haproxy.pid
  daemon

defaults REDIS
  mode tcp
  timeout connect  4s
  timeout server  30s
  timeout client  30s

listen stats 
  bind :9000 # Listen on localhost:9000
  mode http
  stats enable  # Enable stats page
  stats hide-version  # Hide HAProxy version
  stats realm Haproxy\ Statistics  # Title text for popup window
  stats uri /haproxy_stats  # Stats URI
  stats auth admin:123  # Authentication credentials

frontend ft_redis
  bind 127.0.0.1:5000 name redis
  default_backend bk_redis
backend bk_redis
  balance roundrobin
  option tcp-check
  tcp-check connect
  tcp-check send PING\r\n
  tcp-check expect string +PONG
  tcp-check send info\ replication\r\n
  tcp-check expect string role:master
  tcp-check send QUIT\r\n
  tcp-check expect string +OK
  server redis_6379 localhost:7000 check inter 1s weight 179
  server redis_6380 localhost:7001 check inter 1s weight 77

Best Answer

weight has a valid range of 1 through 256 for each server, but you don't have to use 256 as a basis for the calculation.

The weight of each server is the ratio of that server's declared weight to the sum of all declared weights, so with 2 servers you can just use the values 30 and 70 and the distribution will be what you would expect: 30 ÷ (30 + 70) = 0.3 and 70 ÷ (30 + 70) = 0.7. The server that "weighs more" receives proportionally more requests. You could also use 3 and 7 or 33 and 77 or combination within the 1 - 256 range.

The values you are using, 77 and 179, should give similar results, since 77 ÷ (77 + 179) ≈ 0.3 and 179 ÷ (77 + 179) ≈ 0.7... it's just harder to make the mental calculation. Keeping your configuration so all of the weights add to a total sum of 100 is a more human-friendly solution.

Related Topic