Data updating technology

• US 10,725,662 B2
• Filed: 08/20/2018
• Issued: 07/28/2020
• Est. Priority Date: 02/18/2016
• Status: Active Grant

• Current Assignee: Huawei Technologies Co., Ltd. (Huawei Investment & Holding Co., Ltd.)
• Original Assignee: Huawei Technologies Co., Ltd. (Huawei Investment & Holding Co., Ltd.)
• Inventors: Yu, Qun, Xu, Jun, Wang, Yuangang
• Primary Examiner(s): Khan, Masud K
• Application Number:
  US16/105,315
• Publication Number:
  US 20180356979A1
• Time in Patent Office: 708 Days
• CPC Class Codes: G06F 12/0868 Data transfer between cache memory and other subsystems, e.g. storage devices or host systems
• CPC Class Codes: G06F 2212/1016 Performance improvement
• CPC Class Codes: G06F 2212/222 Non-volatile memory
• CPC Class Codes: G06F 2212/281 Single cache
• CPC Class Codes: G06F 3/06 Digital input from, or digital output to, record carriers , e.g. RAID, emulated record carriers or networked record carriers
• CPC Class Codes: G06F 3/061 Improving I/O performance
• CPC Class Codes: G06F 3/0614 Improving the reliability of storage systems
• CPC Class Codes: G06F 3/065 Replication mechanisms
• CPC Class Codes: G06F 3/0656 Data buffering arrangements
• CPC Class Codes: G06F 3/067 Distributed or networked storage systems, e.g. storage area networks [SAN], network attached storage [NAS]
• CPC Class Codes: G06F 3/0689 Disk arrays, e.g. RAID, JBOD

Abstract

A storage system includes a management node and a plurality of storage nodes forming a redundant array of independent disks (RAID). When the management node determines that not all data in an entire stripe is updated based on a received write request, the management node sends update data chunk obtained from to-be-written data to corresponding storage node. The storage node do not directly update, based on the received update data chunks, data block stored in storage device of the storage node, but store the update data chunk into non-volatile memories (NVM) cache of the storage node and send the update data chunk to another storage node to backup. According to the data updating method, write amplification problems caused in a stripe update process can be reduced, thereby improving update performance of the storage system.

Claims

1. A data updating method performed in a storage system having a plurality of storage nodes forming a redundant array of independent disks (RAID), comprising:

receiving, by a management node of the storage system, a write request carrying to-be-written data, wherein a size of the to-be-written data is smaller than a stripe size of the RAID;

obtaining, by the management node, an update data chunk from the to-be-written data;

sending, by the management node to a first storage node in the plurality of storage nodes, a first update request for updating a first data block of a stripe in the RAID, the first data block being stored in the first storage node, the first update request comprising the update data chunk and an update indicator having a first value selected according to the size of the to-be-written data being smaller than the stripe size;

in response to the update indicator in the first update request, storing, by the first storage node, information regarding the update data chunk into a new data node of a log chain of the first data block in a non-volatile memory (NVM) cache of the first storage node, wherein the information regarding the update data chunk comprises the update data chunk or a storage address of the update data chunk in the NVM cache, and wherein the NVM cache contains data representing the log chain of the first data block, the log chain comprises data nodes, each data node containing information regarding a data chunk to be written into the first data block in a corresponding data write event performed on the first data block; and

sending, by the first storage node, a backup request to a second storage node in the plurality of storage nodes, the backup request carrying the update data chunk for storing in the second storage node.

2. The method of claim 1, wherein the new data node is at an end or a head of the log chain, wherein the data nodes in the log chain are arranged in a sequential order according to a time sequence of corresponding data write events.

3. The method of claim 1, further comprising:

sending, by the management node to the plurality of storage nodes, a read request for reading data blocks in the stripe;

obtaining, by the first storage node, an updated first data block based on information regarding update data chunks in the log chain and the first data block stored in the first storage node;

sending, by each of the plurality of storage nodes to the management node, a corresponding data block in the stripe, including the first storage node sending the updated first data block as the first data block to the management node;

receiving, by the management node, data blocks in the stripe sent by the plurality of storage nodes;

calculating, by the management node, an updated parity block based on the data blocks sent by the plurality of storage nodes; and

storing, by the management node, the updated parity block into a storage node storing a parity block of the stripe.

4. The method of claim 3, further comprising:

sending, by the management node to the first storage node, a second update request for updating the first data block of the stripe, the second update request comprising an update indicator having a second value for instructing to store the updated first data block into the first storage node;

in response to the update indicator in the second update request, storing, by the first storage node, the updated first data block into storage space for the first data block in the first storage node.

5. The method of claim 4, further comprising:

releasing, by the first storage node, storage space occupied by the log chain of the first data block after storing the updated first data block into the storage space for the first data block.

6. A data updating method performed by a first storage node of a storage system having a management node and a plurality of storage nodes forming a redundant array of independent disks (RAID), comprising:

receiving from the management node a first update request for updating a first data block of a stripe in the RAID, the first data block being stored in the first storage node, the first update request comprising the update data chunk and an update indicator having a first value indicating that the update data chunk is to be cached by the first storage node;

storing information regarding the update data chunk into a new data node of a log chain of the first data block in a non-volatile memory (NVM) cache of the first storage node in response to the update indicator in the first update request, wherein the information regarding the update data chunk comprises the update data chunk or a storage address of the update data chunk in the NVM cache, and wherein the NVM cache contains data representing the log chain of the first data block, the log chain comprises data nodes, each data node containing information regarding a data chunk to be written into the first data block in a corresponding data write event performed on the first data block; and

sending a backup request to a second storage node in the plurality of storage nodes, the backup request carrying the update data chunk for storing in the second storage node.

7. The method of claim 6, wherein the new data node is at an end or a head of the log chain, the data nodes in the log chain being arranged in a sequential order according to a time sequence of corresponding data write events.

8. The method of claim 6, further comprising:

receiving from the management node a read request carrying an identifier of the stripe and an identifier of the first data block;

determining, based on the identifier of the stripe and the identifier of the first data block, that the log chain of the first data block is stored in the NVM cache of the first storage node;

obtaining an updated first data block based on information regarding data chunks in the data nodes of the log chain and the first data block stored in the first storage node; and

sending the updated first data block to the management node.

9. The method of claim 8, further comprising:

receiving from the management node a second update request for updating the first data block of the stripe, the second update request comprising an update indicator having a second value for instructing to store the updated first data block into the first storage node; and

in response to the update indicator in the second update request, storing the updated first data block into storage space for the first data block in the first storage node.

10. The method of claim 9, further comprising:

releasing storage space occupied by the log chain of the first data block after storing the updated first data block into the storage space for the first data block.

11. The method of claim 6, wherein the information regarding the data chunk of said each data node in the log chain further comprises:

an offset of the data chunk in the first data block, a length of the data chunk, and address information of an adjacent data node of said each data node.

12. A storage system, comprising:

a plurality of storage nodes forming a redundant array of independent disks (RAID), the plurality of storage nodes including a first storage node;

a management node coupled to the plurality of storage nodes and configured to;

receive a write request carrying to-be-written data, wherein a size of the to-be-written data is smaller than a stripe size of the RAID;

obtain an update data chunk from the to-be-written data;

send, to the first storage node in the plurality of storage nodes, a first update request for updating a first data block of a stripe in the RAID, the first data block being stored in the first storage node, the first update request comprising the update data chunk and an update indicator having a first value selected according to the size of the to-be-written data being smaller than the stripe size;

the first storage node being configured to;

store information regarding the update data chunk into a new data node of a log chain of the first data block in a non-volatile memory (NVM) cache of the first storage node in response to the update indicator in the first update request, wherein the information regarding the update data chunk comprises the update data chunk or a storage address of the update data chunk in the NVM cache, and wherein the NVM cache contains data representing the log chain of the first data block, the log chain comprises data nodes, each data node containing information regarding a data chunk to be written into the first data block in a corresponding data write event performed on the first data block; and

send a backup request to a second storage node in the plurality of storage nodes, the backup request carrying the update data chunk for storing in the second storage node.

13. The storage system of claim 12, wherein the new data node is at an end or a head of the log chain, wherein the data nodes in the log chain are arranged in a sequential order according to a time sequence of corresponding data write events.

14. The storage system of claim 12, wherein the management node is further configured to send to the plurality of storage nodes a read request for reading data blocks in the stripe;

and the first storage node is further configured to;

in response to the read request, obtain an updated first data block based on information regarding update data chunks in the log chain and the first data block stored in the first storage node; and

send the updated first data block as data of the first data block to the management node.

15. The storage system of claim 14, wherein the management node is further configured to send to the first storage node a second update request for updating the first data block of the stripe, the second update request comprising an update indicator having a second value for instructing to store the updated first data block into the first storage node;

and

wherein the first storage node is further configured to in response to the update indicator in the second update request, store the updated first data block into storage space for the first data block in the first storage node.

16. The storage system of claim 15, wherein the first storage node is further configured to release storage space occupied by the log chain of the first data block after storing the updated first data block into the storage space for the first data block.

17. The storage system of claim 12, wherein the information regarding the data chunk of said each data node in the log chain further comprises:

an offset of the data chunk in the first data block, a length of the data chunk, and address information of an adjacent data node of said each data node.