I am using haproxy to loadbalance between two webservers. I have a client generator which generates 100 threads per second and browses through the information in this webserver. When there is only one webserver behind the loadbalancer everything goes fine, but when there are two servers on the backend, on the client side i often get

Thread UserSession34: Unable to open URL http://10.76.2.15:80/PHP/index.html (Cannot assign requested address)<br>


What is the general cause of this "cannot assign requested address" issue ? and how to solve it ?

hi

let me guess, this haproxy assign something like cluster address or "floating" IP address? meaning something like a "general" IP address that "cover" real IP addresses behind it?

If so, then maybe something is wrong during the masking or translation...but I could be wrong though...

I see this error on the client side, But take a look at my haproxy config file :

global
log 127.0.0.1 local0
log 127.0.0.1 local1 notice
#log loghost local0 info
maxconn 4096
#debug
#quiet
user haproxy
group haproxy

defaults
log global
mode http
option httplog
option dontlognull
retries 3
option redispatch
maxconn 2000
contimeout 5000
clitimeout 50000
srvtimeout 50000

listen webfarm 10.76.2.15:80
mode http
stats enable
stats auth myuser:mypass
balance roundrobin
cookie JSESSIONID prefix
option httpclose
option forwardfor
option httpchk HEAD /check.txt HTTP/1.0
server webA 10.76.2.107:80 cookie A check
server webB 10.76.2.109:80 cookie B check

hi

Sorry, I never use haproxy before so I can't help much. However, I suggest to enable "debug" option and see if haproxy log provide a useful hint.

Just personal comment: I was suspecting that it has something to do with MAC address to IP cluster assignment. Somehow it "clashes", thus preventing the assignment of the cluster IP. Or maybe it has something to do with the internal load balancing mechanism itself.