0
votes

I have an application (Node.JS) deployed on OpenShift (bronze plan) with the Web Load Balancer activated, the minimum gears active are 3 and the max are 16. Sometimes in the main gear I can see more than one HAProxy instance running, for example now I have:

> ps -ef|grep /usr/sbin/haproxy
3505      37488      1  1 08:46 ?        00:00:01 /usr/sbin/haproxy -f /var/lib/openshift/<APP_ID>/haproxy//conf/haproxy.cfg -sf 37237
3505     149643      1  1 May28 ?        00:09:08 /usr/sbin/haproxy -f /var/lib/openshift/<APP_ID>/haproxy//conf/haproxy.cfg -sf 114873

looking the logs I can't any error. Any explanation about this? Thanks!

1
Are you experiencing some kind of error from this or???user2879327
@developercorey yes, accessing the app is very slow or impossible...Luca Morettoni

1 Answers

2
votes

This could be a consequence of executing Haproxy reload script (/etc/init.d/haproxy). This will usually create a new haproxy process to accept new connections. It will also keep the old process alive until there are still open connections to it. Once they are closed, old haproxy process will be terminated.