-
1. Re: mod_cluster is not working properly with jboss
lylewang Apr 30, 2015 1:18 AM (in response to bharat4212)Your error sounds like this ... https://velenux.wordpress.com/2012/10/11/solving-apache-503-service-temporarily-unavailable-error/
Are these 2 EAP nodes clustered ?
You didn't give details how you setup your mod_cluster and apache, but from what you said seems you didn't have EAP nodes clustered and just setup mod_cluster / apache to redirect requests ?
I'd suggest to cluster these 2 of your EAP nodes properly first , then setup single mod_cluster point for this EAP cluster
This article gives you an outline , followed by a step by step instruction showing how to setup domain mode JBoss7 cluster and then configure mod_cluster with it:
https://docs.jboss.org/author/display/AS72/AS7+Cluster+Howto
If you don't like domain mode and want to use standalone-ha profile (this can be easier) for your EAP instances, then you would want to try these 2 articles:
-
2. Re: mod_cluster is not working properly with jboss
bharat4212 Apr 30, 2015 3:39 AM (in response to bharat4212)Hi Wang,
I have already configured the both jboss in clustered mode, they are in domain mode. And its working fine.
If you could help me on the same, i will be really thankful to you.
-
3. Re: mod_cluster is not working properly with jboss
lylewang Apr 30, 2015 8:11 PM (in response to bharat4212)I have already configured the both jboss in clustered mode, they are in domain mode. And its working fine.
Then you're going to follow the article exactly: https://docs.jboss.org/author/display/AS72/AS7+Cluster+Howto
There shouldn't be much to do on mod_cluster side, if ServerAdvertise is enabled you just need to follow the "Cluster Configuration" section in the article above.
If you disabled server advertise then make sure you add the mod_cluster's IP:port in the " proxy-list" attribute in "modcluster" subsystem inside domain.xml, as described in doc: https://access.redhat.com/documentation/en-US/JBoss_Enterprise_Application_Platform/6.3/html/Administration_and_Configuration_Guide/Configure_the_mod_cluster_Subsystem_to_Use_TCP.html
If you followed all these article and docs , mod_cluster should be able to detect the status of your eap nodes and add them back automatically when it's up again.
Did you check your mod_cluster-manager status when having the error ?
In addition, mod_cluster 1.2.11 is tested configuration see if you could upgrade to this version: https://access.redhat.com/articles/111663