Hi there, To fix the port exhaustion problem in a high concurrency environment, We have setup multiple listening ports on each backend servers:
[...] server n1_0 10.0.0.1:80 check cookie n1 weight 10 server n1_1 10.0.0.1:800 check cookie n1 weight 10 server n1_2 10.0.0.1:801 check cookie n1 weight 10 ... [...] My question is: Could the logical servers such as "n1_1" and "n1_2" share the same check status as "n1_0"? This eliminates the need for these logical servers to send duplicate requests to the same physical server. e.g.: When "n1_0" down, HAProxy auto mark "n1_1" and or others logical servers using same the physical service to "down", vice versa. Is there something like: [...] server n1_0 10.0.0.1:80 check cookie n1 weight 10 server n1_1 10.0.0.1:800 check as n1_0 cookie n1 weight 10 server n1_2 10.0.0.1:801 check as n1_0 cookie n1 weight 10 ... [...] -- Best Regards BaiYang baiy...@gmail.com http://baiy.cn **** < END OF EMAIL > ****
