HA and Fencing

[raj]

Hi Guys,

I am trying to set up HA on a cluster of 3 ibm X3650 boxes.

I have done the "FENCE_JOIN=yes" in /etc/default/redhat-cluster-pve

and then modified the :"cluster.conf.new"

cp /etc/pve/cluster.conf /etc/pve/cluster.conf.new
nano /etc/pve/cluster.conf.new
ccs_config_validate -v -f /etc/pve/cluster.conf.new

Now when I try to do the following "fence_tool join", I get an error : "fence_tool: fenced not running, no lockfile"

I have tried to reboot all the 3 boxes in the cluster but still no joy.

Here is the config :

<?xml version="1.0"?>
<cluster name="bpscluster" config_version="5">
<cman keyfile="/var/lib/pve-cluster/corosync.authkey">
</cman>
<fencedevices>
<fencedevice agent="fence_RSA2" ipaddr="192.168.10.11" login="FENCING" name="Master" passwd="xxxxxxx" secure="1"/>
<fencedevice agent="fence_RSA2" ipaddr="192.168.10.12" login="FENCING" name="Slave01" passwd="xxxxxx" secure="1"/>
<fencedevice agent="fence_RSA2" ipaddr="192.168.10.13" login="FENCING" name="Slave02" passwd="xxxxxx" secure="1"/>
</fencedevices>
<clusternodes>
<clusternode name="master" nodeid="1" votes="1">
<fence>
<method name="1">
<device name="Master"/>
</method>
</fence>
</clusternode>
<clusternode name="slave01" nodeid="2" votes="1">
<fence>
<method name="1">
<device name="Slave01"/>
</method>
</fence>
</clusternode>
<clusternode name="slave02" nodeid="3" votes="1">
<fence>
<method name="1">
<device name="Slave02"/>
</method>
</fence>
</clusternode>
</clusternodes>
<rm>
<service autostart="1" exclusive="0" name="TestIP" recovery="relocate">
<ip address="192.168.7.180"/>
</service>
</rm>
</cluster>



Could someone help pls?

Cheers

 

[raj]

Sorry have also accepted the config from the gui before rebooting

 

[raj]

ok restarted the CMan service on all 3 nodes and that allowed me to join the fencing.

Now I would like to know what should be the ip address at the end of the config:

"<rm>
<service autostart="1" exclusive="0" name="TestIP" recovery="relocate">
<ip address="192.168.7.180"/>
</service>
</rm>
</cluster>"

Of what device should that be.

Also I get an error when I try to run :

"fence_RSA2 -x -l FENCING -p xxxxxxxx -a 192.168.10.11 -o status -n 1 -v"

-bash: fence_RSA2: command not found


I tried to start RGManager and then i get the error:

"Jan 9 10:17:34 master fenced[35049]: fencing node slave01 still retrying"

If I do a :

"# fence_node Master -vv"

I get :

"fence Master dev 0.0 agent fence_RSA2 result: error from agent
agent args: nodename=master agent=fence_RSA2 ipaddr=192.168.10.11 login=FENCING passwd=PASSW0RD secure=1
fence Master failed"


Can anyone assist please?

Cheers,

Raj

 

[tom]

I am not aware of such a fence agent (fencedevice agent="fence_RSA2").

maybe you are looking for fence_rsa?

NAME
       fence_rsa - Fence agent for IBM RSA

DESCRIPTION
       fence_rsa is an I/O Fencing agent which can be used with the IBM RSA II
       management interface. It logs into an RSA  II  device  via  telnet  and
       reboots  the  associated machine. Lengthy telnet connections to the RSA
       II device should be avoided while a GFS cluster is running because  the
       connection will block any necessary fencing actions.

       fence_rsa  accepts  options  on the command line as well as from stdin.
       Fenced  sends  parameters  through  stdin  when  it  execs  the  agent.
       fence_rsa can be run by itself with command line options.  This is useâ
       ful for testing and for turning outlets on or off from scripts.

the resource (IP 192.168.7.180) is just for testing. you do not need this.

 

[raj]

Cheers Tom that did the trick

 

[dietmar]

Now I would like to know what should be the ip address at the end of the config:

That is just an example for a failover IP. I suggest you simply remove this resource unless you want to configure a failover IP.