2 Replies Latest reply on Apr 20, 2015 3:58 AM by fbodin

Error using nodes with jboss.bind.address:127.0.0.1

fbodin Apr 17, 2015 6:06 AM

I want to use the jboss.bind.address:127.0.0.1 on all my JBoss 8.2 nodes connecting to the modcluster.

But when a take down a node in the modcluster, its stops working and the following error occurs.

Note: Everything works when I am using the "hostip" as jboss.bind.address (164.135.92.65, 164.135.92.66)

Fri Apr 17 11:21:39 2015] [debug] mod_manager.c(1910): manager_trans STATUS

[Fri Apr 17 11:21:39 2015] [debug] mod_manager.c(2618): manager_handler STATUS processing: "JVMRoute=L7700706-wildfly02&Load=100"

[Fri Apr 17 11:21:39 2015] [debug] mod_manager.c(1625): Processing STATUS

[Fri Apr 17 11:21:39 2015] [debug] proxy_util.c(1964): proxy: ajp: retrying the worker for (127.0.0.1)

[Fri Apr 17 11:21:39 2015] [debug] proxy_util.c(1970): proxy: ajp: worker for (127.0.0.1) has been marked for retry

[Fri Apr 17 11:21:39 2015] [debug] proxy_util.c(2026): proxy: ajp: has acquired connection for (127.0.0.1)

[Fri Apr 17 11:21:39 2015] [debug] proxy_util.c(2082): proxy: connecting ajp://127.0.0.1:8209/ to 127.0.0.1:8209

[Fri Apr 17 11:21:39 2015] [debug] proxy_util.c(2209): proxy: connected / to 127.0.0.1:8209

[Fri Apr 17 11:21:39 2015] [debug] proxy_util.c(2460): proxy: ajp: fam 2 socket created to connect to 127.0.0.1

[Fri Apr 17 11:21:39 2015] [error] (111)Connection refused: proxy: ajp: attempt to connect to 127.0.0.1:8209 (127.0.0.1) failed

[Fri Apr 17 11:21:39 2015] [error] ap_proxy_connect_backend disabling worker for (127.0.0.1)

[Fri Apr 17 11:21:39 2015] [debug] mod_proxy_cluster.c(1361): proxy_cluster_try_pingpong: can't connect to backend

[Fri Apr 17 11:21:39 2015] [debug] proxy_util.c(2044): proxy: ajp: has released connection for (127.0.0.1)

[Fri Apr 17 11:21:39 2015] [debug] mod_proxy_cluster.c(2331): proxy_cluster_isup: pingpong ajp://127.0.0.1:8209/ failed

[Fri Apr 17 11:21:39 2015] [debug] mod_manager.c(2667): manager_handler STATUS OK

[Fri Apr 17 11:21:39 2015] [error] proxy: CLUSTER: (balancer://wildfly02). All workers are in error state

mod_cluster_manager

LBGroup wildfly02: Enable Nodes Disable Nodes

Node L7700705-wildfly02 (ajp://127.0.0.1:8209):

Enable Contexts Disable Contexts

Balancer: wildfly02,LBGroup: wildfly02,Flushpackets: Off,Flushwait: 10000,Ping: 10000000,Smax: 100,Ttl: 600000000,Status: OK,Elected: 822,Read: 77907649,Transferred: 370722,Connected: 0,Load: 9

Virtual Host 1:

Contexts:

/app/web, Status: ENABLED Request: 0 Disable

/app/ws, Status: ENABLED Request: -1 Disable

Aliases:

localhost

default-host

Node L7700706-wildfly02 (ajp://127.0.0.1:8209):

Enable Contexts Disable Contexts

Balancer: wildfly02,LBGroup: wildfly02,Flushpackets: Off,Flushwait: 10000,Ping: 10000000,Smax: 100,Ttl: 600000000,Status: OK,Elected: 4954,Read: 468942147,Transferred: 2234254,Connected: 1,Load: 73

Virtual Host 1:

Contexts:

/app/ws, Status: ENABLED Request: 1 Disable

/app/web, Status: ENABLED Request: 0 Disable

Aliases:

localhost

default-host

Quick config summary

Two nodes, where each node has a httpd front-end and a JBoss Wildfly 8.2 server with the application deployed.

Installed RPM:s

httpd-tools-2.2.15-39.el6.x86_64

httpd_mod_cluster-1.2.6-1.noarch

httpd-2.2.15-39.el6.x86_64

/etc/httpd/conf.d/mod_cluster.conf

# Load mod_cluster modules

LoadModule slotmem_module /etc/httpd/modules/mod_slotmem.so

LoadModule manager_module /etc/httpd/modules/mod_manager.so

LoadModule proxy_cluster_module /etc/httpd/modules/mod_proxy_cluster.so

LoadModule advertise_module /etc/httpd/modules/mod_advertise.so

MemManagerFile /var/cache/mod_cluster

# Virtualhost for Wildfly mod_cluster

Listen 164.135.92.65:6666

Order deny,allow

Deny from all

Allow from all

</Directory>

LogLevel debug

ServerAdvertise Off

EnableMCPMReceive

KeepAliveTimeout 60

MaxKeepAliveRequests 0

</VirtualHost>

/etc/httpd/conf.d/app.conf

ServerName app.abc.se

KeepAliveTimeout 60

MaxKeepAliveRequests 0

AllowDisplay On

SetHandler mod_cluster-manager

Order deny,allow

Deny from all

Allow from all

</Location>

</VirtualHost>

JBoss 8.2 (two nodes L7700705 & L7700706)

<mod-cluster-config proxy-list="164.135.92.65:6666,164.135.92.66:6666" balancer="wildfly02" advertise="false" excluded-contexts="ROOT,invoker,jbossws,juddi,console,jolokia" auto-enable-contexts="true" sticky-session="true" sticky-session-remove="true" worker-timeout="1" max-attempts="3" smax="100" ttl="600" node-timeout="10" load-balancing-group="wildfly02" connector="ajp">

<dynamic-load-provider>

<load-metric type="busyness"/>

</dynamic-load-provider>

</mod-cluster-config>

</subsystem>

Message was edited by: Fredrik Bodin

1. Re: Error using nodes with jboss.bind.address:127.0.0.1

mbabacek Apr 17, 2015 6:24 AM (in response to fbodin)

Well, how could the balancer contact the worker node if it says it resides on localhost -- which is apparently false.
Unless you have all the nodes on a single system, you must start the workers with interfaces reachable from the balancer.

Cheers
-K-
1 of 1 people found this helpful
Actions
2. Re: Error using nodes with jboss.bind.address:127.0.0.1

fbodin Apr 20, 2015 3:58 AM (in response to mbabacek)

Thanks for the quick answer.. just wanted to be 100% sure.

---
I have a follow-up questions, which some might help me with.

Is it possible to configure an ACL using remote ip in undertow for the http-listner and ajp-listner? exapmle..
My problem now is that my apache front has a mod-plugin for SSO, which you now can by pass.
Or what possible options do I have, to solve this.. want to handle it as simple as possible.

Best regards Fredrik
Actions

Go to original post