mod_cluster apache and 2 jboss_managers
gioppoluca Sep 30, 2013 11:22 AMI've this configuration:
1 Apache (on VM0)
Jboss_Master1 (domain master and server node on VM1) + Jboss_slave1 (just server node on VM2) each with server_group1 and server_group2 defined (each with just 1 node)
Jboss_Master2 (domain master and server node on VM3) + Jboss_slave2 (just server node on VM4) each with server_group3 defined
so in total I've got 3 couples of servers that has to be balanced between the two of them.
All nodes are different virtual machines.
I tryed to create two virtual hosts in Apache, but it came out that "mod_advertise: directive in more than one VirtualHost: not supported".
So I'm trying to figure out how to be sure that I end up having each server_group is balanced between its own nodes?
Can I define in apache
Listen 10.1.1.84:10001
<VirtualHost 10.1.1.84:10001>
## Vhost Location
<Location />
Order allow,deny
Allow from all
</Location>
# This directive allows you to view mod_cluster status at URL nvh_addr_port/mod_cluster-manager
<Location /mod_cluster-manager>
SetHandler mod_cluster-manager
Order deny,allow
Allow from 10.1.1.0/24
AllowDisplay on
</Location>
KeepAliveTimeout 60
MaxKeepAliveRequests 0
EnableMCPMReceive
ManagerBalancerName vdbcluster
AdvertiseFrequency 5
ServerAdvertise On
</VirtualHost>
And set on both the domains this?
[domain@10.1.1.43:9999 /] /profile=ha/subsystem=modcluster/mod-cluster-config=configuration:read-resource
{
"outcome" => "success",
"result" => {
"advertise" => true,
"advertise-socket" => "modcluster",
"auto-enable-contexts" => true,
"balancer" => "vdbcluster",
"dynamic-load-provider" => {"load-metric" => [{"type" => "busyness"}]},
"excluded-contexts" => "ROOT,invoker,jbossws,juddi,console",
"flush-packets" => false,
"flush-wait" => -1,
"max-attempts" => 1,
"node-timeout" => -1,
"ping" => 10,
"proxy-list" => "apache.internal:10001",
"proxy-url" => "/",
"smax" => -1,
"socket-timeout" => 20,
"ssl" => undefined,
"sticky-session" => true,
"sticky-session-force" => false,
"sticky-session-remove" => false,
"stop-context-timeout" => 10,
"ttl" => -1,
"worker-timeout" => -1
}
}
Will mod-cluster be smart enough to understand that there will be 6 nodes presenting but that he has to balance them in couples? The right couples?
Thanks