This is the 12th day of my participation in the August More Text Challenge. For details, see “August More Text Challenge”.

In today’s IT world, the data storage challenges faced by enterprises and management are often the need for capacity, performance, and price. On the one hand, it is difficult to accurately estimate the demand for subsequent performance growth when an enterprise purchases a storage system. On the other hand, as the service volume increases, it is difficult to adjust services after adding disks to existing systems. How to solve the above problems, is the enterprise development process must think about the problem, especially in the early stage of IT system construction, to do a good job in long-term performance requirements planning.

The principle is introduced

During system expansion, the SmartMotion feature automatically migrates data and balances data on all disks. Customers do not need to make a long-term estimate of performance requirements when purchasing, but only need to meet the recent performance requirements, effectively reducing the initial purchase cost, thus reducing TCO. When the service volume increases, you need only to add disks to the system. SmartMotion can evenly distribute the data of original services to all disks through data migration to improve the performance of original services.

In Huawei RAID2.0+, all disks in a disk domain are divided into chunks of the same size. To allocate CKG, disks are selected using the pseudo-random algorithm, and then chunks are allocated from the selected disks to form CKG based on the RAID algorithm. After the random selection is performed using the pseudo-random algorithm, all chunks are evenly distributed on each disk.

Intelligent data migration process

SmartMotion is based on RAID2.0+ technology. After a disk is added to a disk domain, the system automatically starts SmartMotion. The working process is as follows:

  1. The first CKG without equalization was selected;
  2. The hard disk is selected again for CKG by pseudo random algorithm.
  3. If the selected disk is the same as the existing disk list, skip this CKG and go to Step 1.
  4. Compare the original CKG disk with the newly selected disk list, calculate the mapping between the disks to be migrated based on the difference in the list, and select the original CKG disk to be migrated and the corresponding target disk.
  5. The system traverses each disk to be migrated in the CKG, allocates new chunks from the target disk, migrates data from the original disk to the new disk, and releases the chunks on the original disk.
  6. If all CKG in the system have been traversed, the SmartMotion is complete. Otherwise, go to Step 1 to process the next CKG.

After SmartMotion is complete, all CKGS select new disks using the pseudo-random algorithm and complete the required data migration. In this case, all chunks are evenly distributed on all disks (including newly added disks).

Thank you very much