ScaleIO capabilities may be used for volumes that support traditional VMFS datastores provided that all VMFS volumes of the datastore are grouped into a single consistency group. In this scenario the entire datastore and all VMDKs in it will be snapped as a single unit.
Typically, a single VMFS datastore will occupy a single VMFS volume on a single ScaleIO volume. But if the datastore uses multiple extents to span volumes, multiple ScaleIO volumes may support a single datastore.
If multiple VMFS datastores are used to support multiple VMDKs for multiple ASM disks within an ASM diskgroup, then all volumes for all datastores for all ASM diskgroups for an Oracle database should be included in a consistency group.
ScaleIO supports Write Order Fidelity (sometimes called Write Ordering), meaning that any snapshot or clone used in this fashion will provide a crash consistent copy of the VMFS datastores.
pRDMs may utilize the full capabilities of ScaleIO array including initiating such features from inside the guest OS.