wildfly modcluster proxy can't acess my defult wildfly page with 80 port.

[jeryma]

hi,Stuart Douglas

1、my httpconf configuration is:

<IfModule manager_module>

  Listen 192.168.157.57:6666

  ManagerBalancerName mycluster

  <VirtualHost 192.168.157.57:6666>

    <Location />

     Order deny,allow

     Deny from all

     Allow from all

    </Location>

  KeepAliveTimeout 300

    MaxKeepAliveRequests 0

    ServerAdvertise on http://192.168.157.57:6666

    AdvertiseFrequency 5

    AdvertiseSecurityKey mySecurityKey

    AdvertiseGroup 224.0.1.105:23364

    EnableMCPMReceive

    AllowDisplay On

    LogLevel debug

   ErrorLog "logs/mycluster_log"

  <Location /mod_cluster-manager>

       SetHandler mod_cluster-manager

       Order deny,allow

       Deny from all

       Allow from all

    </Location>

  </VirtualHost>

</IfModule>

ProxyPreserveHost On

ProxyPass / balancer://mycluster/

ProxyPassReverse / balancer://mycluster

2、my modcluster domain.xml configuration(ha):

<mod-cluster-config advertise-socket="modcluster" proxy-list="192.168.157.57:6666" proxy-url="/" balancer="mycluster" advertise-security-key="mySecurityKey" load-balancing-group="main-server-group" connector="ajp">

3、Now point a browser to http://localhost/ (on port 80, where apache is listening). You should get the WildFly default welcome page, as expected. but  Connection is reset on the browser side.

and the domain start console apear this error ervery per second：

[Server:server-one] 11:09:13,825 ERROR [org.jboss.modcluster.mcmp.impl.DefaultMCMPHandler] (UndertowEventHandlerAdapter - 1) Error [null: null: {4}] sending command STATUS to proxy sitolftp.cnsuning.com/192.168.157.57:6666, configuration will be reset

[Server:server-three] 11:09:14,826 ERROR [org.jboss.modcluster.mcmp.impl.DefaultMCMPHandler] (UndertowEventHandlerAdapter - 1) Error [null: null: {4}] sending command STATUS to proxy sitolftp.cnsuning.com/192.168.157.57:6666, configuration will be reset

[Server:server-two] 11:09:18,829 ERROR [org.jboss.modcluster.mcmp.impl.DefaultMCMPHandler] (UndertowEventHandlerAdapter - 1) Error [null: null: {4}] sending command STATUS to proxy sitolftp.cnsuning.com/192.168.157.57:6666, configuration will be reset

[Server:server-four] 11:09:19,832 ERROR [org.jboss.modcluster.mcmp.impl.DefaultMCMPHandler] (UndertowEventHandlerAdapter - 1) Error [null: null: {4}] sending command STATUS to proxy sitolftp.cnsuning.com/192.168.157.57:6666, configuration will be reset

When I remove the proxypass,the error will disappear.but how can I acess my default wildfly page with 80 port?

@Stuart Douglas

what should I do? thanks for your guide.

[ctomc]

Hi,

what version of WildFly are you using?

can you try with nightly build that uses latest Undertow.

You can find it at https://community.jboss.org/thread/224262

there are many things that ware fixed in ajp protocol implementation since alpha4.

Also what is undertow subsystem configuration and can you post whole server.log.

[jeryma]

thanks for your reply.I used wildfly wildfly-8.0.0.Alpha4. And I had used 10.02 nightly build edtion with undertow is beta15,but this problem also exist.

my httpd error log :

[Thu Oct 03 11:28:07 2013] [notice] child pid 953 exit signal Segmentation fault (11)

[Thu Oct 03 11:28:07 2013] [debug] mod_proxy_cluster.c(678): update_workers_node starting

[Thu Oct 03 11:28:07 2013] [debug] mod_proxy_cluster.c(678): update_workers_node starting

[Thu Oct 03 11:28:08 2013] [debug] proxy_util.c(1818): proxy: grabbed scoreboard slot 0 in child 1040 for worker proxy:reverse

[Thu Oct 03 11:28:08 2013] [debug] proxy_util.c(1837): proxy: worker proxy:reverse already initialized

[Thu Oct 03 11:28:08 2013] [debug] proxy_util.c(1914): proxy: initialized worker 0 in child 1040 for (*) min=0 max=25 smax=25

[Thu Oct 03 11:28:08 2013] [notice] child pid 982 exit signal Segmentation fault (11)

[Thu Oct 03 11:28:08 2013] [notice] child pid 985 exit signal Segmentation fault (11)

[Thu Oct 03 11:28:09 2013] [debug] proxy_util.c(1818): proxy: grabbed scoreboard slot 0 in child 1069 for worker proxy:reverse

[Thu Oct 03 11:28:09 2013] [debug] proxy_util.c(1837): proxy: worker proxy:reverse already initialized

[Thu Oct 03 11:28:09 2013] [debug] proxy_util.c(1914): proxy: initialized worker 0 in child 1069 for (*) min=0 max=25 smax=25

[Thu Oct 03 11:28:09 2013] [debug] proxy_util.c(1818): proxy: grabbed scoreboard slot 0 in child 1072 for worker proxy:reverse

[Thu Oct 03 11:28:09 2013] [debug] proxy_util.c(1837): proxy: worker proxy:reverse already initialized

[Thu Oct 03 11:28:09 2013] [debug] proxy_util.c(1914): proxy: initialized worker 0 in child 1072 for (*) min=0 max=25 smax=25

[Thu Oct 03 11:28:09 2013] [debug] mod_proxy_cluster.c(678): update_workers_node starting

[Thu Oct 03 11:28:11 2013] [info] removed PID file /opt/jboss/httpd/httpd/logs/httpd.pid (pid=15803)

there is no error in my server.log and my httpd error log.I feel confused.

my undettow  is default configuration.

[ctomc]

What default configuration?

standalone.xml?

that one does not come with ajp listener configured in undertow subsystem...

server.log mainly booting log would be helpful.

[jeryma]

domain cluster with profile name="ha".

my undertow configuration is next:

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

                <buffer-caches>

                    <buffer-cache name="default" buffer-size="1024" buffers-per-region="1024" max-regions="10"/>

                </buffer-caches>

                <server name="default-server">

                    <ajp-listener name="ajp" socket-binding="ajp" max-post-size="10485760"/>

                    <http-listener name="default" socket-binding="http" max-post-size="10485760"/>

                    <host name="default-host" alias="localhost">

                        <location name="/" handler="welcome-content"/>

                    </host>

                </server>

                <servlet-container name="default" default-buffer-cache="default" stack-trace-on-error="local-only">

                    <jsp-config/>

                    <persistent-sessions path="persistent-web-sessions" relative-to="jboss.server.data.dir"/>

                </servlet-container>

                <handlers>

                    <file name="welcome-content" path="${jboss.home.dir}/welcome-content" directory-listing="true"/>

                </handlers>

            </subsystem>

            <subsystem xmlns="urn:jboss:domain:webservices:1.2">

                <modify-wsdl-address>true</modify-wsdl-address>

                <wsdl-host>${jboss.bind.address:192.168.157.57}</wsdl-host>

                <endpoint-config name="Standard-Endpoint-Config"/>

                <endpoint-config name="Recording-Endpoint-Config">

                    <pre-handler-chain name="recording-handlers" protocol-bindings="##SOAP11_HTTP ##SOAP11_HTTP_MTOM ##SOAP12_HTTP ##SOAP12_HTTP_MTOM">

                        <handler name="RecordingHandler" class="org.jboss.ws.common.invocation.RecordingServerHandler"/>

                    </pre-handler-chain>

                </endpoint-config>

                <client-config name="Standard-Client-Config"/>

            </subsystem>

            <subsystem xmlns="urn:jboss:domain:weld:1.0"/>

        </profile>

my server.log

11:04:15,560 INFO  [org.infinispan.remoting.transport.jgroups.JGroupsTransport] (Incoming-1,shared=udp) ISPN000094: Received new cluster view: [master:server-one/web|1] [master:server-one/web, master:server-two/web]

11:04:18,399 ERROR [org.jboss.modcluster.mcmp.impl.DefaultMCMPHandler] (UndertowEventHandlerAdapter - 1) Error [null: null: {4}] sending command STATUS to proxy sitolftp.cnsuning.com/192.168.157.57:6666, configuration will be reset

11:04:19,938 INFO  [org.infinispan.remoting.transport.jgroups.JGroupsTransport] (Incoming-3,shared=udp) ISPN000094: Received new cluster view: [master:server-one/web|2] [master:server-one/web, master:server-two/web, master:server-three/web]

11:04:23,211 INFO  [org.infinispan.remoting.transport.jgroups.JGroupsTransport] (Incoming-5,shared=udp) ISPN000094: Received new cluster view: [master:server-one/web|3] [master:server-one/web, master:server-two/web, master:server-three/web, master:server-four/web]

11:04:28,518 ERROR [org.jboss.modcluster.mcmp.impl.DefaultMCMPHandler] (UndertowEventHandlerAdapter - 1) Error [null: null: {4}] sending command STATUS to proxy sitolftp.cnsuning.com/192.168.157.57:6666, configuration will be reset

11:04:39,528 ERROR [org.jboss.modcluster.mcmp.impl.DefaultMCMPHandler] (UndertowEventHandlerAdapter - 1) Error [null: null: {4}] sending command STATUS to proxy sitolftp.cnsuning.com/192.168.157.57:6666, configuration will be reset

11:04:49,535 ERROR [org.jboss.modcluster.mcmp.impl.DefaultMCMPHandler] (UndertowEventHandlerAdapter - 1) Error [null: null: {4}] sending command STATUS to proxy sitolftp.cnsuning.com/192.168.157.57:6666, configuration will be reset

11:05:00,550 ERROR [org.jboss.modcluster.mcmp.impl.DefaultMCMPHandler] (UndertowEventHandlerAdapter - 1) Error [null: null: {4}] sending command STATUS to proxy sitolftp.cnsuning.com/192.168.157.57:6666, configuration will be reset

11:05:12,562 ERROR [org.jboss.modcluster.mcmp.impl.DefaultMCMPHandler] (UndertowEventHandlerAdapter - 1) Error [null: null: {4}] sending command STATUS to proxy sitolftp.cnsuning.com/192.168.157.57:6666, configuration will be reset

11:05:22,573 ERROR [org.jboss.modcluster.mcmp.impl.DefaultMCMPHandler] (UndertowEventHandlerAdapter - 1) Error [null: null: {4}] sending command STATUS to proxy sitolftp.cnsuning.com/192.168.157.57:6666, configuration will be reset

11:05:33,585 ERROR [org.jboss.modcluster.mcmp.impl.DefaultMCMPHandler] (UndertowEventHandlerAdapter - 1) Error [null: null: {4}] sending command STATUS to proxy sitolftp.cnsuning.com/192.168.157.57:6666, configuration will be reset

11:05:43,592 ERROR [org.jboss.modcluster.mcmp.impl.DefaultMCMPHandler] (UndertowEventHandlerAdapter - 1) Error [null: null: {4}] sending command STATUS to proxy sitolftp.cnsuning.com/192.168.157.57:6666, configuration will be reset

11:05:53,599 ERROR [org.jboss.modcluster.mcmp.impl.DefaultMCMPHandler] (UndertowEventHandlerAdapter - 1) Error [null: null: {4}] sending command STATUS to proxy sitolftp.cnsuning.com/192.168.157.57:6666, configuration will be reset

11:06:03,617 ERROR [org.jboss.modcluster.mcmp.impl.DefaultMCMPHandler] (UndertowEventHandlerAdapter - 1) Error [null: null: {4}] sending command STATUS to proxy sitolftp.cnsuning.com/192.168.157.57:6666, configuration will be reset

11:06:13,631 ERROR [org.jboss.modcluster.mcmp.impl.DefaultMCMPHandler] (UndertowEventHandlerAdapter - 1) Error [null: null: {4}] sending command STATUS to proxy sitolftp.cnsuning.com/192.168.157.57:6666, configuration will be reset

11:06:24,639 ERROR [org.jboss.modcluster.mcmp.impl.DefaultMCMPHandler] (UndertowEventHandlerAdapter - 1) Error [null: null: {4}] sending command STATUS to proxy sitolftp.cnsuning.com/192.168.157.57:6666, configuration will be reset

11:06:34,652 ERROR [org.jboss.modcluster.mcmp.impl.DefaultMCMPHandler] (UndertowEventHandlerAdapter - 1) Error [null: null: {4}] sending command STATUS to proxy sitolftp.cnsuning.com/192.168.157.57:6666, configuration will be reset

11:06:45,664 ERROR [org.jboss.modcluster.mcmp.impl.DefaultMCMPHandler] (UndertowEventHandlerAdapter - 1) Error [null: null: {4}] sending command STATUS to proxy sitolftp.cnsuning.com/192.168.157.57:6666, configuration will be reset

11:06:55,677 ERROR [org.jboss.modcluster.mcmp.impl.DefaultMCMPHandler] (UndertowEventHandlerAdapter - 1) Error [null: null: {4}] sending command STATUS to proxy sitolftp.cnsuning.com/192.168.157.57:6666, configuration will be reset

[ctomc]

this looks like your setup problem.

server (wildfly) cannot reach your apache proxy with mod_cluster, that is why it is saying that i cannot reach it on 192.168.157.57:6666 try telnet-ing to that address from machine that runs wildfly.

--

tomaz

[jeryma]

thanks for your help.

but my wildfly server (domain control)  and my apache modcluster in the same machine which address is 192.168.157.57.

actually,my domain cluster also have another problem which is modcluster can't identity other node in my domain control( three node in domain control,three node in my slave server) modcluster can only identify the ajp 8009 default port in my domain controller and other 8009 ajp port in my slave server.

however, other ajp 8109、8209 port in my domain controller and slave can‘t be identified by modcluster.(it seems look like bug in wildfly-8.0.0.Alpha4,We have responded to this problem to the community,community say they will fix this bug in next edition wildfly)

this is my http://192.168.157.57:6666/mod_cluster-manager accessing page.192.168.57.57 and 10.19.90.193 can ping through each other.

[ctomc]

xiaojie Ma wrote:

however, other ajp 8109、8209 port in my domain controller and slave can‘t be identified by modcluster.(it seems look like bug in wildfly-8.0.0.Alpha4,We have responded to this problem to the community,community say they will fix this bug in next edition wildfly)

Yes this was bug in Alpha4 and is already fix in current nightly builds.

you can try it out by downloading latest nightly build WildFly nightly builds available also Beta1 is just around the corner that will contain fix for this.

--

tomaz

[jeryma]

hi, tomaz,thank you very much for your support.

today,I use beta1 (10.04) edition wildfly for test, all server ajp port have already listened in starting console.(two node in domain controller)

master for example:

[Server:server-one] 11:28:51,952 INFO  [org.wildfly.extension.undertow] (MSC service thread 1-1) JBAS017520: Undertow AJP listener ajp was bound to /10.19.218.172:8009(node1)

[Server:server-two] 11:28:51,993 INFO  [org.wildfly.extension.undertow] (MSC service thread 1-2) JBAS017520: Undertow AJP listener ajp  was bound to /10.19.218.172:8159(node2)

slave is the same as domain controller.

But when I access the http://192.168.157.57:6666/mod_cluster-manager in my Browser,the problem is also exist.10.19.218.172 is domain control with 2 node,10.19.90.193 is slave with 2 node.

the question seem also exist as before.I confused if  my http.conf or domain.xml modcluster configuration have some problem or not.(configuration is on above )

If I remove the proxy-list="192.168.157.157:6666" from the domain.xml ,the slave does not identified by modcluster in http://192.168.157.57:6666/mod_cluster-manager page.

(10.05) latest nightly build wildfly-8.0.0.Beta2-SNAPSHOT is the same as beta1 (10.04).it sees look like  modcluster only identify the 8009 default port.

--------------------

By the way,If I want to set two http server connect with one jboss server,how should I configure in jboss ?

@Stuart Douglas ,Tomaz Cerar please help me ,thank your  very much!

[ctomc]

Mod cluster guys are looking into this and there are some fixes coming into wildfly codebase shortly.

maybe rhusar or jfclere can help bit more.

In any case i think this thread is bit all over the place, as original problem we "fixed" what you have now is different story.

So can you create new thread with description of you current problem so it will make it easier for mod cluster guys to help you

[jfclere]

Not sure about the 10.19.218.172:8159 no showing up.

For having more than one httpd just add it in the proxy-list.

[jeryma]

start console have this information:

[Server:server-two] 11:28:51,993 INFO  [org.wildfly.extension.undertow] (MSC service thread 1-2) JBAS017520: Undertow AJP listener ajp  was bound to /10.19.218.172:8159.I dont't understand what you mean.

And I have to add proxy-list in my domain.xml configuration so that modcluster works well with only default  8009 port. Add more than httpd just add it in the proxy-list .you say yes.

The problem is modcluster can't identify other nodes which I set in my host.xml. and I set httpd server and domain control in the same machine. It look lilke jgrop or broadcast having some problem?

[jfclere]

I have just tried master standalone with ajp on 8159 it works. Could you try what the latest snapshot or one using mod_cluster-1.3.0.Alpha1?

[jeryma]

what’s your  modcluster version and jboss version ？and what‘s your wildfly modcluster config? It show the 8159 node in http://*:6666/mod_cluster-manager?

[jfclere]

mod_cluster-1.2.6.Final. WildFly8 master.

+++

        <subsystem xmlns="urn:jboss:domain:modcluster:1.2">
            <mod-cluster-config advertise-socket="modcluster" advertise-security-key="secret" connector="ajp">
                <dynamic-load-provider>
                    <load-metric type="busyness"/>
                </dynamic-load-provider>
            </mod-cluster-config>
        </subsystem>
+++

And:

+++

<socket-binding name="ajp" port="${jboss.ajp.port:8159}"/>

+++

Yes it shows the node has excepted.