[mod_cluster-issues] [JBoss JIRA] (MODCLUSTER-356) BalancerMember directives cause httpd crash when using mod_cluster 1.2.x

Sat Sep 14 15:22:03 EDT 2013

    [ https://issues.jboss.org/browse/MODCLUSTER-356?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12804316#comment-12804316 ] 

Weinan Li commented on MODCLUSTER-356:
--------------------------------------

I've met the simliar problem but it seems doesn't related with settings in <Proxy ...> section:

I'm using JBoss EAP 6.1.0.Final + httpd 2.2.25 + mod_cluster 1.2.4.Final for testing. I'm working on a MacOS machine with JDK 7 installed:

{code}
jmin:~ weinanli$ java -version
java version "1.7.0_25"
Java(TM) SE Runtime Environment (build 1.7.0_25-b15)
Java HotSpot(TM) 64-Bit Server VM (build 23.25-b01, mixed mode)
min:~ weinanli$ uname -a
Darwin min 12.4.0 Darwin Kernel Version 12.4.0: Wed May  1 17:57:12 PDT 2013; root:xnu-2050.24.15~1/RELEASE_X86_64 x86_64
{code}

Here are the error logs from JBoss side:

{code}
[Server:server-three] 03:03:27,829 INFO  [org.jboss.modcluster] (ContainerBackgroundProcessor[StandardEngine[jboss.web]]) MODCLUSTER000011: jboss.web will use 30b18166-b3c4-3388-8fbe-50af073c7657 as jvm-route
[Server:server-three] 03:03:29,238 ERROR [org.jboss.modcluster] (ContainerBackgroundProcessor[StandardEngine[jboss.web]]) MODCLUSTER000042: Error null sending STATUS command to min/10.0.1.13:10001, configuration will be reset: null
{code}

After JBoss EAP started in domain mode then httpd start to throw segmentation fault periodically:

{code}
[Sun Sep 15 02:54:07 2013] [notice] child pid 84530 exit signal Segmentation fault (11)
{code}

Even if I kill JBoss EAP now, the seg fault keep thrown by httpd.

So I upgrade the mod_cluster of httpd side to 1.2.6.Final, and the problem disappeared:

{code}
[Sun Sep 15 03:07:47 2013] [notice] Balancer other-server-group changed
[Sun Sep 15 03:07:48 2013] [notice] Balancer other-server-group changed
[Sun Sep 15 03:07:48 2013] [notice] Balancer other-server-group changed
[Sun Sep 15 03:07:48 2013] [notice] Balancer other-server-group changed
[Sun Sep 15 03:07:48 2013] [notice] Balancer other-server-group changed
[Sun Sep 15 03:07:50 2013] [notice] Balancer other-server-group changed
{code}

And JBoss EAP side seems fine:

{code}
[Server:server-three] 03:07:47,302 INFO  [org.jboss.modcluster] (ContainerBackgroundProcessor[StandardEngine[jboss.web]]) MODCLUSTER000011: jboss.web will use 30b18166-b3c4-3388-8fbe-50af073c7657 as jvm-route
{code}

I could see the node is registered in management console:

{code}
Node 30b18166-b3c4-3388-8fbe-50af073c7657 (ajp://10.0.1.13:8009):

Enable Contexts Disable Contexts
Balancer: other-server-group,LBGroup: ,Flushpackets: Off,Flushwait: 10000,Ping: 10000000,Smax: 1,Ttl: 60000000,Status: OK,Elected: 0,Read: 0,Transferred: 0,Connected: 0,Load: 100
{code}

But the cluster still not working correctly and I get '503 Service Temporarily Unavailable':

{code}
min:~ weinanli$ telnet 10.0.1.13 80
Trying 10.0.1.13...
Connected to min.
Escape character is '^]'.
GET /
<!DOCTYPE HTML PUBLIC "-//IETF//DTD HTML 2.0//EN">
<html><head>
<title>503 Service Temporarily Unavailable</title>
</head><body>
<h1>Service Temporarily Unavailable</h1>
<p>The server is temporarily unable to service your
request due to maintenance downtime or capacity
problems. Please try again later.</p>
</body></html>
Connection closed by foreign host.
{code}

I'll try to upgrade the EAP side mod_cluster to see if it could solve the problem.

Here is my http config:

{code}
<VirtualHost *:80>
    ProxyPass / balancer://other-server-group
    ProxyPassReverse / balancer://other-server-group

    <Location />
        Order deny,allow
        Allow from All
    </Location>
   <Location /mod_cluster-manager>
      SetHandler mod_cluster-manager
      Order deny,allow
      Allow from 10.0.1
   </Location>

</VirtualHost>
{code}

{code}
<VirtualHost 10.0.1.13:10001>

  <Directory />
    Order deny,allow
    Deny from all
    Allow from 10.0.1.
  </Directory>

  KeepAliveTimeout 60
  MaxKeepAliveRequests 0
  ManagerBalancerName other-server-group
  EnableMCPMReceive
  AdvertiseGroup 224.0.1.105:23364
  AdvertiseFrequency 5
</VirtualHost>
{code}

Here is EAP config:

{code}
<subsystem xmlns="urn:jboss:domain:modcluster:1.1">
    <mod-cluster-config advertise-socket="modcluster" connector="ajp">
        <dynamic-load-provider>
            <load-metric type="busyness"/>
        </dynamic-load-provider>
    </mod-cluster-config>
</subsystem>
{code}

> BalancerMember directives cause httpd crash when using mod_cluster 1.2.x
> ------------------------------------------------------------------------
>
>                 Key: MODCLUSTER-356
>                 URL: https://issues.jboss.org/browse/MODCLUSTER-356
>             Project: mod_cluster
>          Issue Type: Bug
>    Affects Versions: 1.2.4.Final
>         Environment: SLES-11-SP1 x86_64 running httpd 2.2.24 or 2.4.6 with any mod_cluster-1.2.x version and JBoss-AS-7.2.0.Final
>            Reporter: Marco Danti
>            Assignee: Jean-Frederic Clere
>              Labels: balancerMember, crash, httpd, mod_cluster
>
> httpd dumps core in mod_proxy_cluster.c if the httpd.conf file contains BalancerMember directives.
> this happens at least with the following two setups:
>  1) httpd-2.4.6 and mod_cluster-1.2.x
>  2) httpd-2.2.24 and mod_cluster-1.2.3.Final

--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira