mod_cluster 1.3 with AS7; Cannot register node...

[ja101]

Hello!

I am trying to get mod_cluster1.3 work with AS7.1.1 final but for some reason the node does not register with the  mod_cluster manager. Is there anything I am missing? Any help will be much appreciated...

The AS node gives the following error

INFO  [org.jboss.modcluster.mcmp.impl.DefaultMCMPHandler] (ContainerBackgroundProcessor[StandardEngine[jboss.web]]) IO error sending command CONFIG to proxy

However, the httpd access_log gives  "INFO / HTTP/1.1" 200 581  for this node and I get the following from error_log

(111)Connection refused: AH00957: http: attempt to connect to 127.0.0.1:8080 (127.0.0.1) failed

[proxy:error] [pid 1161:tid 140222859224832] AH00959: ap_proxy_connect_backend disabling worker for (127.0.0.1) for 60s

[proxy:error] [pid 1162:tid 140222859224832] AH00940: http: disabled connection for (127.0.0.1)

[proxy:error] [pid 1249:tid 140222859224832] AH00940: http: disabled connection for (127.0.0.1)

[proxy:error] [pid 1163:tid 140222859224832] AH00940: http: disabled connection for (127.0.0.1)

[proxy:error] [pid 1249:tid 140222859224832] AH00940: http: disabled connection for (127.0.0.1)

I should also mention that I have a perfectly fine working node with Tomcat within this setup.

I have also used the testhttpd.war tool provided with the jboss mod_cluster distro and it reports no errors.

I use the minimal config for AS7 as advised in the doc (mod_cluster Documentation)

My httpd.conf is as under:

<IfModule manager_module>

  #change port value according to your preferences

  Listen *:6666

  ManagerBalancerName mycluster

  <VirtualHost *:6666>

        <Location />

            #change according to the network, multiple Require entries are allowed

        Require ip 192.168.

        </Location>

        KeepAliveTimeout 300

        MaxKeepAliveRequests 0

        ServerAdvertise on http://192.168.152.132:6666

        AdvertiseFrequency 5

        AdvertiseGroup 224.0.1.106:23364

        EnableMCPMReceive

        AllowDisplay On

        <Location /mod_cluster_manager>

        SetHandler mod_cluster-manager

        Order deny,allow

        Allow from all

        </Location>

  </VirtualHost>

</IfModule>

[mbabacek]

The error is that you bind AS to localhost. Do not do that, preferably ever with any mod_cluster setup.

Furthermore, you forbade your 127.0.* worker from connecting to your balacner with this directive: Require ip 192.168.

Steps:

• start both balancer and workers on non-localhost ip addresses, e.g. 192.something
• make sure bidirectional communication is possible, i.e. that balancer can contact worker nodes on the host:port they report and that workers can contact your balancer on its MCPM enabled VirtualHost

Please, let me know it it helped.

Cheers

-K-

[ja101]

Hi Michal,

Thank you for your quick reply...

I am relatively inexperienced with AS so could you please elaborate a bit more on 

"The error is that you bind AS to localhost. Do not do that, preferably ever with any mod_cluster setup". and how can I bind it to the real ip of the node...

the bidirectional communication is verified and as the tomcat node does connect with the balancer, I am inclined to believe that there is an issue with my AS configuration...

Thanks!

junaid

[ja101]

UPDATE:

I have changed the config in AS7 for the address binding. Please see below for two snips from my AS7 config

<subsystem xmlns="urn:jboss:domain:modcluster:1.0">

            <mod-cluster-config advertise-socket="modcluster" proxy-list="192.168.152.132:6666">

                <dynamic-load-provider>

and binding:

</interface>

        <interface name="public">

            <any-address/>

        </interface>

Now, although I do NOT get the errors like

(111)Connection refused: AH00957: http: attempt to connect to 127.0.0.1:8080 (127.0.0.1) failed

But I still get the following on AS node

INFO  [org.jboss.modcluster.mcmp.impl.DefaultMCMPHandler] (ContainerBackgroundProcessor[StandardEngine[jboss.web]]) IO error sending command CONFIG to proxy c152-132.localcloud/192.168.152.132:6666: java.net.SocketTimeoutException: Read timed out

and

192.168.153.180 - - [06/Nov/2015:11:58:43 +0100] "CONFIG / HTTP/1.1" 200 -

192.168.153.180 - - [06/Nov/2015:11:57:03 +0100] "INFO / HTTP/1.1" 200 587

on the balancer node.

Its a bit strange that I can now see the AS node 192.168.153.180 in my mod_cluster manager GUI  outside Virtual Hosts but it says 'node undefined' for it and errors are still there....

192.168.153.180 - -

and 192.168.152.132 is the balancer node

I stumbled across this post  mod_cluster manager does not discover JBoss nodes which reports a similar error but has not been addressed...