The thin provision integration functionality helps you to monitor the space usage on thin-provisioned LUNs and to avoid running out of space.
The following sample flow demonstrates how the ESXi host and the storage array interact to generate breach of space and out-of-space warnings for a datastore with underlying thin-provisioned LUN. The same mechanism applies when you use Storage vMotion to migrate virtual machines to the thin-provisioned LUN.
Using storage-specific tools, your storage administrator provisions a thin LUN and sets a soft threshold limit that, when reached, triggers an alert. This step is vendor-specific.
Using the vSphere Web Client, you create a VMFS datastore on the thin-provisioned LUN. The datastore spans the entire logical size that the LUN reports.
As the space used by the datastore increases and reaches the specified soft threshold, the following actions take place:
The storage array reports the breach to your host.
Your host triggers a warning alarm for the datastore.
You can contact the storage administrator to request more physical space or use Storage vMotion to evacuate your virtual machines before the LUN runs out of capacity.
If no space is left to allocate to the thin-provisioned LUN, the following actions take place:
The storage array reports out-of-space condition to your host.Caution:
In certain cases, when a LUN becomes full, it might go offline or get unmapped from the host.
The host pauses virtual machines and generates an out-of-space alarm.
You can resolve the permanent out-of-space condition by requesting more physical space from the storage administrator.