Am I missing some "cibadmin -save" command or something?

I have 3 blades setup in an Active/Active/Standby configuration for an XMPP 
package.  It works fine, failing over processes to the standby blade when one 
of the active blades is rebooted, or starts having issues running our software.

But if I do a "service heartbeat stop" on all of the blades, when I start it 
all back up all of my resources and constraints are gone?

This isn't happening on any other of my HA clusters, but this is the first one 
that I have completely filled using cibadmin instead of the GUI.

Info:
    RedHat 5

    heartbeat.i386                           2.1.3-22.1             installed
    heartbeat-common.i386                    2.1.3-22.1             installed
    heartbeat-resources.i386                 2.1.3-22.1             installed
    pacemaker.i386                           0.6.4-1.1              installed
    pacemaker-pygui.i386                     1.3.1-1.3              installed

Cibs:
cat addXcp1:  (There are 7 more similar Jabber instances divided among the 2 
active blades)
     <resources>
       <group id="prodea-lorealm1-group">
         <primitive id="IP_prodea-lorealm1" class="ocf" type="IPaddr2" 
provider="heartbeat">
           <instance_attributes id="IP_prodea-lorealm1_instance_attrs">
             <attributes>
               <nvpair id="Ip_prodea-lorealm1_ip" name="ip" 
value="172.22.0.111"/>
               <nvpair id="Ip_prodea-lorealm1_nic" name="nic" value="eth0"/>
               <nvpair id="IP_prodea-lorealm1_target_role" name="target_role" 
value="started"/>
             </attributes>
           </instance_attributes>
         </primitive>
         <primitive class="ocf" type="Jabber" provider="heartbeat" 
id="router_prodea-lorealm1">
           <instance_attributes id="router_prodea-lorealm1_instance_attrs">
             <attributes>
               <nvpair id="router_prodea-lorealm1_xcp_id" name="xcp_id" 
value="prodea-lorealm1"/>
               <nvpair id="router_prodea-lorealm1_target_role" 
name="target_role" value="started"/>
             </attributes>
           </instance_attributes>
           <operations>
             <op id="id-monitor_prodea-lorealm1" name="monitor" 
description="monitor_prodea-lorealm1" interval="30" timeout="20" 
start_delay="20"/>
           </operations>
         </primitive>
         <instance_attributes id="prodea-lorealm1_instance_attrs">
           <attributes>
             <nvpair id="prodea-lorealm1_target_role" name="target_role" 
value="started"/>
           </attributes>
         </instance_attributes>
       </group>
     </resources>



_______________________________________________
Linux-HA mailing list
[email protected]
http://lists.linux-ha.org/mailman/listinfo/linux-ha
See also: http://linux-ha.org/ReportingProblems

Reply via email to