[Linux-cluster] GFS/cman related problems

[thomasie]

Dear ALL:

Hello, I hope I'm sending to the correct list.  I'm having problems
starting up gfs 6.1, and hopefully it's just something incorrect with my
configuration.

I used two standard machine . My system is Fedora Core 4. The kernel modules
load fine.

First of all,  I load all the needed kernel modules for the cluster. To do
this, execute the following commands (on both nodes):
    root at node1 # modprobe lock_dlm 

    root at node2 # modprobe lock_dlm 


Next, I start the cluster configuration service daemon on both nodes with
    root at node1 # ccsd 

    root at node2 # ccsd
    

Having started ccsd I need to create the cluster by starting the cluster
manager on both nodes:
    root at node1 # /sbin/cman_tool join 

    root at node2 # /sbin/cman_tool join    
    
when we are going to start both nodes we get that the first node without
problems, but the second one never get connected at cman and remain trying to
connect with the following error: 10.190.5.174 (node1) udp port 6809
unreachable.
I have not start firewall. And I use the command of  "netstat -a" , 
the result is 
udp        0      0 127.0.0.1:6809              0.0.0.0:*                     
         
udp        0      0 224.0.0.1:6809              0.0.0.0:*                     
         
but it receive package is 10.190.6.82:6809 (node2) -> 10.190.5.174 :6809
(node1)

Then, I run cman_tool with DEBUG defined and get this:
node1: sending HELLO
node2: sending membership request, but I get udp port 6809 unreachable.


My configuration file (cluster.xml) is 
<?xml version="1.0"?>
<cluster name="alpha" config_version="1">

<cman two_node="1" expected_votes="1">
  <multicast addr="224.0.0.1"/>
</cman>

<clusternodes>
  <clusternode name="node1" nodeid="1" votes="1">
    <multicast addr="224.0.0.1" interface="eth1"/>
    <fence>
      <method name="human">
        <device name="net_dev1"/>
      </method>
    </fence>
  </clusternode>

  <clusternode name="node2" nodeid="2" votes="1">
    <multicast addr="224.0.0.1" interface="eth0"/>
    <fence>
      <method name="human">
        <device name="net_dev2"/>
      </method>
    </fence>
  </clusternode>
</clusternodes>

<fencedevices>
  <fencedevice name="net_dev1" agent="fence_gnbd" server="node1" login="root"
passwd="111111"/>
  <fencedevice name="net_dev2" agent="fence_gnbd" server="node2" login="root"
passwd="111111"/>
</fencedevices>

</cluster>

any help is much appreciated.

Regards,
thomasie





--http://www.eyou.com
--Îȶ¨¿É¿¿µÄµç×ÓÐÅÏä  ÓïÒôÓʼþ  Òƶ¯ÊéÇ©  ÈÕÀú·þÎñ  ÍøÂç´æ´¢...ÒÚÓÊδ¾¡

--http://vip.eyou.com
--¿ì¿ìµÇ¼ÒÚÓÊVIPÐÅÏä  ×¢²áÄúÖÐÒâµÄÓû§Ãû

--http://sms.eyou.com
--ÎÞÓǶþ¶þ×å¡¢×ãÇò´ó¸»ÎÌ...¾¡ÔÚÒÚÓʶÌÐÅ