We have a fully virtualized datacentre with a SAN synced to a 2nd location. Our co- sourcing partner has advised us to use clustering for High Availability and to avoid database corruptions. How would clustering avoid db corruptions?