White Papers

VMware storage features
69 Dell EMC SC Series: Best Practices with VMware vSphere | 2060-M-BP-V
16.3.5 Automatic dead space reclamation
For automatic space reclamation to work, there are several requirements:
• The host must be running ESXi 6.5 and vCenter 6.5 or later.
• The datastore must be formatted with VMFS-6.
• The VMFS-6 datastore volumes must be stored within a 512k pagepool. VMware only supports
automatic unmap on array block sizes less than 1 MB.
• The SC Series array must be running a SCOS version that supports VAAI UNMAP. See the VMware
HCL for compatible versions.
Note: Array snapshots containing freed blocks will not release space until the snapshot expires.
16.3.6 In-guest space reclamation
For in-guest space reclamation to work, there are several requirements:
• The host must be running ESXi 6.0 and vCenter 6.0 or later. The datastore must be formatted with
VMFS-6 or VMFS-5. For VMFS-5 datastores, the advanced system setting
VMFS3.EnableBlockDelete must be set to 1.
• The datastore volumes must be stored within a 512k pagepool. VMware only supports automatic
space reclamation on array block sizes less than 1 MB.
• The SC Series array must be running a SCOS version that supports VAAI UNMAP. See the VMware
HCL for compatible versions.
• The virtual machine hardware must be version 11 or higher.
• The guest operating system must support UNMAP. For example: Windows 2012 R2 and higher.
• The virtual disks must be in the thin format.
Note: On datastore volumes stored with a 2 MB or greater pagepool, the volume can be manually unmapped
after in-guest space reclamation using the esxcli storage vmfs unmap --volume-label [datastore_name].
16.3.7 Thin provisioning stun
The thin provisioning stun primitive allows the SC Series array to send a special SCSI sense code back to the
ESXi host when there is an out of space (OOS) condition. ESXi will pause only the virtual machines that are
requesting additional pages until additional storage is added to remedy the situation.
For more information about VAAI primitives, see sections, “Array Thin Provisioning and VMFS Datastores”
and “Storage Hardware Acceleration” in the appropriate vSphere Storage Guide at VMware vSphere
documentation.
16.4 vStorage APIs for Storage Awareness (VASA)
The vStorage APIs for Storage Awareness (VASA) are a set of protocols, routines, and tools encapsulated
into an API that enable vCenter to detect the onboard capabilities of storage arrays. vCenter can obtain al of
the properties of a volume such as its RAID level, performance capabilities, if it is replicated, and if Data
Progression is enabled. In turn, vCenter uses these capabilities to enable profile-driven storage, allowing the
administrator to determine if a virtual machine meets the organization's compliance policies.
VASA 2.0 is also required for VMware vSphere Virtual Volumes™ (vVols). vVols use the VASA provider to
perform the out-of-band tasks required for the volume-to-container relationship management.