[Linux-HA] Failback

[Zachár Balázs]

Hello!

I have an Active/Passive cluster with heartbeat v2. After I generate an
error to test (ifdown eth0 and there is an alias IPaddress whics is
managed with the standard OCF RA), the cluster is works and the
resources are start on b-node. After that I "repair" the a-node (ifup
eth0) and I would like to migrate the resources back to the a-node whith
this command on b-node:

#crm_resource -M -r group_1

After that the resources stop on b-node but not start on a-node... If i
do this on b-node:

#crm_resource -U -r group_1

They start again on b-node.

Now I do this on a-node:

#crm_resource -W -r group_1
resource group_1 not running

now i do thison a-node:
#cibadmin -S
#crm_resource -W -r group_1
resource group_1 is running on: b-node


OK the CIB is ok, but if i would like to migrate back again the
resources to the a-node, they still stop on b-node but not start on
a-node...

Why? I think a-node not active or something...


Sorry my poor english
Thanks,
Balázs