Taking down a CLUSTERIP manually adds extra entries into iptables

I've been playing the the failover capabilities of the resources-agents by taking down the CLUSTERIP address manually:

```
ip addr del 10.0.0.80/24 dev eth1
```

After a few seconds it automatically comes back up (good), but I get extra unwanted entries inside the iptables:

```
Chain INPUT (policy ACCEPT)
target prot opt source destination
CLUSTERIP all -- 0.0.0.0/0 10.0.0.80 CLUSTERIP hashmode=sourceip-sourceport clustermac=35:FC:51:8E:57:45 total_nodes=2 local_node=1 hash_init=0
CLUSTERIP all -- 0.0.0.0/0 10.0.0.80 CLUSTERIP hashmode=sourceip-sourceport clustermac=35:FC:51:8E:57:45 total_nodes=2 local_node=1 hash_init=0
CLUSTERIP all -- 0.0.0.0/0 10.0.0.80 CLUSTERIP hashmode=sourceip-sourceport clustermac=35:FC:51:8E:57:45 total_nodes=2 local_node=1 hash_init=0
```

(one extra line for every manual takedown of the ip address).

Probably need to add a check and not to add additional entries (or remove previous ones).

Start by reproducing the failover with `ip addr del 10.0.0.80/24 dev eth1`, then follow the resource-agent entry point that restores the CLUSTERIP address and inspect the resulting iptables rules. Done means repeated manual takedowns restore the address without adding duplicate CLUSTERIP entries.

Written by the indexing model from the issue text.