Check your /var/log/messages, you must be hitting a “nf_conntrack: table full, dropping packet” error
________________________________ From: jerry liu <[email protected]> To: [email protected] Sent: Tuesday, August 14, 2012 9:05 AM Subject: encounter a problem when testing HAProxy Hi, When I was doing a test with HAProxy, I bumped into a problem, I didn't know why it happened, I wish you could help me out. Here's the scenario, I set up HAProxy on server A and apache on server B, I used gatling as the test client that was running on server C to send 1000 requests per second to HAProxy, I noticed that, the current sessions is growing rapidly and when it hit around 9000, the gatling client got the "connection timed out" error. Below is the conf of the HAProxy I was using and attached is the screenshots that shows the status of HAProxy from the beginning of starting test client to the time when the error happened. I'm looking forward to hearing from you soon. global log 127.0.0.1 local0 maxconn 131086 daemon< div> stats socket /tmp/haproxy.sock defaults log global mode http option httplog option httpclose option dontlognull option abortonclose log 127.0.0.1 local3 retries 3 option redispatch maxconn 131086 timeout connect 20000 contimeout 50000 clitimeout 50000 srvtimeout 50000</ div> listen web_proxy ip address of Server A mode http balance roundrobin # cookie SERVERID insert indirect nocache # option httpchk HEAD /index.html HTTP/1.0 server webA ip address of Server B stats enable stats hide-version stats uri /haproxy-stats stats realm Haproxy\ statistics stats auth admin:admin stats refresh 3s stats admin if TRUE
