Description of problem:

If by accident in cluster.conf is for ip resource specified as 

<ip address="192.168.122.11/24" monitor_link="1"/>

then no error will be reported during cluster start, and this error in configuration will be possible to propagate to cluster nodes. 

Entering above is not possible via luci interface - as there is not space, this is only possible when manually editing cluster.conf. 

Version-Release number of selected component (if applicable):

cman-2.0.115-68.el5
rgmanager-2.0.52-9.el5


How reproducible:

Set up two node cluster, no qdisk. 

Configure service which needs ip address ( nfs/http ), edit cluster.conf per above. After changes services will be showed as running in clustat output but there will not be active ip address for them. Services will not be functional what is expected. 

In my testing I used nfs service, migrating it from one node to another was fine, mounting nfs share was also fine, but not error was reported that ip address has bad format for it. 

# ip a 

did not showed it. 


Steps to Reproduce:
1.
2.
3.
  
Actual results:


Expected results:

There should be mechanism to prevent obvious errors in cluster.conf to be propagated to cluster. In case there is unexpected value for resource specified in cluster.conf that should be reported back, and not allow such error to be send to cluster nodes.



Additional info: