OID 11g how to recover from a bad realm configuration in a clustered environment

Node 1 is our main node, Node2 is our secondary node. We are using IDM 11.1.1.6.
(1) used "setup.exe -deinstall" to remove OID installation.  Following the following instructions:
http://www.idmworks.com/blog/entry/uninstalling-oracle-fusion-middleware-products

We kept the admin server running.
We then deinstalled ASInstance on node 1 using "setup.exe -deinstall".
We then deinstalled Oracle IDM_home on node 1 using "setup.exe -deinstall".
We did not deinstall the remote instance on node2.
We then went to file system to remove idmdomain manually. It did not allow.

We disabled node manager  and AS instance control in Windows Servers. We restarted node 1, then we are able to successfully remove idmdomain.
We then deleted the remaining of AsInstance folder in the file system.

(2) We revert the 11g RAC database back to before we started the 11g OID install and configuration.

(3)We run "setup.exe" to Install and Configure node1.
   Thing went well. Admin server up, ... odsm works.

(4) On Node2, we tried to reconfigure by running "config.exe" but it is complaining about AS instance being existed.

we disabled both node manager and Oracle Instance Control (I think this is opmn control) in windows service.
 We then removed Oracle Instance Control windows service by issuing the following in an administrative command prompt:
      sc delete "Oracle ...Instance Control".
rerun "config.exe" succeeded.

(5) We want to http:<node1>:port/em to ensure that both oid1 and oid2 are up.
We then stopped the existing  wls_ods1 on node1. and restarted wls_ods1 so that it is aware of the 2nd node in the cluster.