Exact Extract — Huawei V3.0 Training Material, PDF Page 153
Common reconstruction: When a disk is faulty, the system uses an idle CK to replace the faulty one and restores data to the idle CK.
Huawei RAID 2.0+ uses block virtualization to improve storage-resource utilization, performance distribution, and reconstruction efficiency.
Instead of constructing traditional RAID groups exclusively from entire physical disks, RAID 2.0+ divides disks into smaller fixed-size physical units called chunks (CKs) .
Multiple chunks from different disks are combined to create a Chunk Group (CKG).
The CKG provides the logical structure on which the configured RAID redundancy is implemented.
For example, chunks can be organized into RAID 5, RAID 6, or RAID-TP layouts according to the selected protection policy.
This design distributes data and parity across available storage devices.
When a disk fails, the system reconstructs the affected chunks using available redundant information and distributes reconstruction operations across suitable storage resources.
This differs from conventional RAID implementations in which reconstruction may be restricted to a small number of disks.
Option B , CKG, identifies a group of chunks rather than the smallest physical building unit among the listed options.
Option C , disk, is the larger physical device.
Option A , cache, provides temporary high-speed storage rather than the underlying RAID allocation unit.
Therefore, D — Block (chunk) is correct.
Source qualification: The V3.0 slide uses the CK terminology, while Huawei's official storage configuration documentation supplies the explicit CK and CKG definitions.