sechevarria, take note: using different mount points for different servers does not prevent data corruption due to concurrency errors caused by split-brain. Even if the mountpoint was different on different servers, you'd still have the database or applicaiton running simultaneously on both hosts (not to mention your volume manager and the file system itself) and this in itself would cause data corruption if it is on shared storage.
The only way to completely ensure data safety from concurrency and split-brain is I/O Fencing being implemented in VCS. Otherwise, just do your best to prevent split-brain by using public network low-priority LLT heartbeats, etc. Another 'trick' people used to do was to put the IP resource at the bottom of the service group tree, since the IP resource does a ping before bringing up the IP (in case of duplicate IPs on the network). That's fine, but public network low-priority heartbeats are a better workaround.
