[ https://issues.jboss.org/browse/MODCLUSTER-303?page=com.atlassian.jira.pl... ] Martin Kamp Jensen edited comment on MODCLUSTER-303 at 4/15/15 8:15 AM: ------------------------------------------------------------------------ If I understand this correctly, this is exactly what we would benefit from. Note that we cannot use multicast. Currently, we are testing the following setup and experiencing a mod_cluster discovery issue/challenge: - The first node (WildFly 8.2) is set up, mod_cluster and JGroups having proxy_list and initial_hosts, respectively, set up to point to the IP address of the node - The first node is started - Later, the second node is set up, mod_cluster and JGroups having proxy_list and initial_hosts, respectively, set up to point to the IP address of the new node in addition to the IP address of the first node - The second node is started - JGroups on boths nodes discovers both nodes - mod_cluster on the first node discovers both nodes (maybe because mod_cluster of the second node, which has the IP address of the first node in its proxy_list, talks to the first node) - mod_cluster on the second node only discovers itself - A restart of the first node (with proxy_list updated to include both nodes) fixes the problem was (Author: mkjensen): If I understand this correctly, this is exactly what we would benefit from. Note that we cannot use multicast. Currently, we are testing the following setup and experiencing a mod_cluster discovery issue/challenge: - The first node (WildFly 8.2) is set up, mod_cluster and JGroups having proxy_list and initial_hosts, respectively, set up to point to the IP address of the node - The first node is started - Later, the second node is setup, mod_cluster and JGroups having proxy_list and initial_hosts, respectively, set up to point to the IP address of the new node in addition to the IP address of the first node - The second node is started - JGroups on boths nodes discovers both nodes - mod_cluster on the first node discovers both nodes (maybe because mod_cluster of the second node, which has the IP address of the first node in its proxy_list, talks to the first node) - mod_cluster on the second node only discovers itself - A restart of the first node (with proxy_list updated to include both nodes) fixes the problem -- This message was sent by Atlassian JIRA (v6.3.11#6341)