Which Storage Policy Changes Require a Full Rebuild of All Components for an Object?

CE.

The correct answers are A and E.

Storage policies in VMware vSAN are used to define the service levels and availability requirements of virtual machines and their associated virtual disks. These policies consist of a set of rules that govern various aspects of storage, such as the number of disk stripes, the level of redundancy, the IOPS limit, and so on.

When a storage policy is changed, vSAN must update the metadata associated with all the components that make up the object. Depending on the nature of the change, this may require a full rebuild of all components for the object. Here are the two storage policy changes that require a full rebuild of all components for an object:

A. Increase the number of disk stripes for an object from 2 to 3: This change requires a full rebuild of all components for the object because the number of stripes affects the way data is distributed across the disks in the cluster. When the number of stripes is increased, vSAN must redistribute the data to ensure that it is evenly spread across the new set of disks.

E. Change the failure tolerance method of an object from RAID-1 (Mirroring) to RAID-5 (Erasure Coding): This change requires a full rebuild of all components for the object because the failure tolerance method determines how the data is protected against hardware failures. When the failure tolerance method is changed, vSAN must rebuild all components to ensure that they conform to the new method. In this case, changing from RAID-1 to RAID-5 requires the data to be encoded and distributed across the disks in a different way.

The other three options do not require a full rebuild of all components for the object:

B. Increase the IOPS limit for an object: This change does not require a full rebuild of all components for the object because it only affects the way that the object is accessed and does not require any changes to the underlying data.

C. Reduce the Primary Failures to Tolerate from 2 to 1 for an object using a failure tolerance method of RAID-1 (Mirroring): This change does not require a full rebuild of all components for the object because it only affects the level of redundancy and does not require any changes to the underlying data.

D. Disable object checksum on an object: This change does not require a full rebuild of all components for the object because it only affects the way that the object is verified and does not require any changes to the underlying data.