I'm using apache http server as reverse proxy and balancer for a cluster of JBOSS EAP 6 server with Liferay 6.2 deployed on it.

Previously, mod_jk was the balancer. This leaded us to a problem that, during jboss startup, as soon as the AJP 8009 port was available, apache proceeded with forwarding request to the starting member. Since of course Liferay was not yet completely deployed, we used to get a 503 error on the browser.

I'm trying to replace it with mod_proxy_ajp that can use the mod_proxy_hcheck module. So I was trying to use the hcmethod=get and hcuri=/portale

<Proxy "balancer://cluster">
  BalancerMember "ajp://" loadfactor=1 route=node1 keepalive=On ttl=60 hcmethod=GET hcuri=/portale hcinterval=5 hcpasses=2 hcfails=2
  BalancerMember "ajp://" loadfactor=2 route=node2 keepalive=On ttl=60 hcmethod=GET hcuri=/portale hcinterval=5 hcpasses=2 hcfails=2
  ProxySet lbmethod=bytraffic

ProxyPass        / balancer://cluster/  stickysession=JSESSIONID lbmethod=bytraffic nofailover=Off
ProxyPassReverse / balancer://cluster/  stickysession=JSESSIONID

but this puts all the members in error state, I guess because that check cannot be done on 8009. Is there another way to do this (healthcheck based on the real availability of Liferay) or to do the hcuri healthcheck on that page but using the classic 8080 http port, without touching the AJP balancing on 8009?

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.