EMCVNX Virtual Provisioning provides pool-based storage provisioning by implementing pool LUNs that can be either thin or thick. Thin LUNs provide on-demand storage that maximizes the utilization of your storage by allocating storage as it is needed. Thick LUNs provide high and predictable performance for your applications mainly because all of the user capacity is reserved and allocated upon creation. Both types of LUNs benefit from the ease-of-use features of pool-based provisioning. Pools and pool LUNs are also the building blocks for advanced data services such as Fully Automated Storage Tiering for Virtual Pools (FAST™ VP), Compression, and Deduplication. Virtual Provisioning enables organizations to reduce storage costs by increasing capacity utilization, simplifying storage management, and reducing application downtime. Virtual Provisioning also helps companies to reduce power and cooling requirements and reduce capital expenditures.
Now that we understand the general idea behind EMC VNX Virtual Provisioning lets go over some of the highlighted points and improvements discussed at EMC World 2014.
All compressed and deduplicated LUNs are thin provisioned by default.
Pool LUN Metadata
3GB + 2% of allowed LUN capacity.
Example: 100GB thick LUN consumes 105GB of space.
Latest performance numbers
VNX 1st series Thick Pool LUNs equal to performance of VNX 2nd series Thin Pool LUNs
Due to advances in EMC VNX MCx architecture.
Default for VNX 2nd series LUN creation is thin.
Thin LUN metadata automatically moves to the highest tier of the pool unless the tiering policy pins the LUN to a specific tier.
Thin LUNs only.
Returns unused space to the Pool upon deletion.
Currently support is Windows 2012 NTFS, ESXi 5.x, Veritas Storage Foundation.
Replication and Migration
Thin to thin
Thin attributes are maintained by all replication products.
Replicator, RecoverPoint, MirrorView, SAN Copy, SnapView Clones, PowerPath Migration Enabler.
Thick to thin
RecoverPoint, remote SAN Copy pull, PowerPath Migration Enabler maintain thin attributes of the target.
Thick to thin replication is supported by all products, but thin attribute of target may not be maintained in all cases.
Thin to thick
Supported by all products, but thin attribute may not be maintained in the case of replication reversal.