Data storage device including a solid state drive

Patent metadata

Field	Value
Publication number	US-12379995-B2
Application number	US-202318349074-A
Country	US
Kind code	B2
Filing date	Jul 7, 2023
Priority date	Feb 27, 2023
Publication date	Aug 5, 2025
Grant date	Aug 5, 2025

Abstract

A data storage device comprises a disk array including a plurality of disks; and a redundant array of independent disks (RAID) controller configured to manage the disk array according to a RAID mode, wherein each of the plurality of disks includes: a nonvolatile memory device having a plurality of zones of which supports a sequential write operation; and a random write buffer associated to one of the plurality of zones, the random write buffer having a start address that is indicated by a write pointer, the write pointer indicating a location where a next write operation is to be performed.

First claim

What is claimed is:

1. A data storage device comprising:
a disk array including a plurality of disks; and
a redundant array of independent disks (RAID) controller configured to manage the disk array according to a RAID mode,
wherein each of the plurality of disks includes:
a nonvolatile memory device having a plurality of zones each of which supports a sequential write operation; and
a random write buffer associated to one of the plurality of zones, the random write buffer having a start address that is indicated by a write pointer, the write pointer indicating a location where a next write operation is to be performed,
wherein, when the RAID controller generates and stores a parity generated for a data chunk that is currently stored in one disk among the plurality of disks, the RAID controller retains a parity generated and stored for a data chunk that was previously stored in another disk among the plurality of disks.

2. The data storage device of claim 1 , wherein the random write buffer has a size equal to or greater than two times of a size of a data chunk which is a unit of a write operation.

3. The data storage device of claim 1 , wherein the plurality of disks includes first, second, third, and fourth disks, and the RAID controller stores a first data chunk in a random write buffer of the first disk, stores a parity generated corresponding to the first data chunk in a random write buffer of the third disk, and updates a write pointer of the first disk.

4. The data storage device of claim 3 , wherein the RAID controller stores a second data chunk in a random write buffer of the second disk, stores a parity generated corresponding to the second data chunk in a random write buffer of the fourth disk, and updates write pointers of the second disk.

5. The data storage device of claim 4 , wherein when the RAID controller stores the parity generated corresponding to the second data chunk in the random write buffer of the fourth disk, the RAID controller keeps the parity generated corresponding to the first data chunk valid.

6. The data storage device of claim 1 , wherein the plurality of disks includes first, second, third, fourth, and fifth disks, and wherein the RAID controller stores a first data chunk in a random write buffer of the first disk, stores a first parity generated corresponding to the first data chunk in a random write buffer of the third disk, stores a second parity generated corresponding to the first data chunk in a random write buffer of the fourth disk, and update a write pointer of the first disk.

7. The data storage device of claim 6 , wherein the RAID controller stores a second data chunk in a random write buffer of the second disk, stores a first parity generated corresponding to the second data chunk in the random write buffer of the fourth disk, stores a second parity generated corresponding to the second data chunk in a random write buffer of the fifth disk, and updates a write pointer of the second disk.

8. The data storage device of claim 7 , wherein when the RAID controller stores the first parity generated corresponding to the second data chunk in the random write buffer of the fourth disk and stores the second parity generated corresponding to the second data chunk in the random write buffer of the fifth disk, the RAID controller keeps the first parity and second parity generated corresponding to the first data chunk valid.

9. The data storage device of claim 1 , wherein the RAID controller stores a first data chunk in a random write buffer of a first disk, stores a parity generated corresponding to the first data chunk on one of the plurality of disks, advances a write pointer of the first disk by 0.5 unit, stores a second data chunk in a random write buffer of a second disk, stores a parity generated corresponding to the second data chunk on one of the plurality of disks, and further advances a write pointer of the first disk by 0.5 unit after storing the second data chunk and the parity generated corresponding to the second data chunk, and
wherein the 0.5 unit corresponds to a half of a size of a data chunk, the data chunk being a unit of data to perform a write operation by the RAID controller.

10. The data storage device of claim 9 , wherein the RAID controller manages a write pointer corresponding to an entire address space of the disk array, finds a last disk where a data chunk is lastly stored with reference to the write pointer corresponding to the entire address space after the data storage device is recovered from an error, and determines whether the last disk should be recovered or not with reference to a write pointer of a disk in which a data chunk is stored prior to the last disk when a failure occurs in the last disk.

11. The data storage device of claim 1 , wherein each of the plurality of disks includes a disk controller,
wherein when the disk controller receives a request for updating the write pointer, the disk controller controls data, corresponding to the random write buffer, to be sequentially written in an unused area of a zone corresponding to the random write buffer and updates the write pointer so that the write pointer indicates a start address of the random write buffer.

12. The data storage device of claim 1 , wherein the random write buffer includes a volatile memory device or a nonvolatile memory device supporting a random write operation and an overwrite operation.

13. The data storage device of claim 12 , wherein the random write buffer includes a volatile memory device, and the data storage device further includes an emergency power supply device to provide a power to the volatile memory device during power failure.

14. A method of operating a data storage device including a plurality of disks, the method comprising:
writing a data chunk in a random write area of a disk corresponding to a disk variable K, K being an integer;
determining whether the disk variable K is smaller than N−1, N being a positive integer;
when the disk variable K is smaller than N−1, writing a partial parity corresponding to written data chunks in a random write area of a disk corresponding to a disk variable (K+2), advancing a write pointer of the disk corresponding to the disk variable K, increasing the disk variable K by 1, and repeating the writing of the data chunk and the determining until the disk variable K becomes equal to N−1; and
when the disk variable K is equal to N−1, writing a full parity corresponding to the written data chunks in a random write area of a disk corresponding to a disk variable (K+1) and advancing the write pointer of the disk corresponding to the disk variable K, the write pointer indicating a location of the disk corresponding to the disk variable K where a next write operation is to be performed,
wherein the partial parity represents a parity generated when an available space to store additional data chunk is left in a stripe constituted by the plurality of disks, and the full parity represents a parity generated when no available space is left in the stripe.

15. The method of claim 14 , wherein a partial parity generated corresponding to a first data chunk is kept valid when a partial parity generated corresponding to the first data chunk and a second data chunk is stored, the second data chunk being written after the second data chunk is written.

16. A method of operating a data storage device including a plurality of disks, the method comprising:
writing a data chunk in a random write area of a disk corresponding to a disk variable K, K being an integer;
determining whether the disk variable K is smaller than N−3, N b

Assignees

• Sk Hynix Inc
• Seoul Nat Univ R&Db Foundation

Inventors

• Kim Minwook
• Kim Jinsoo
• Son Ikjoon
• Lee Euidong

Classifications

• G06F11/1084 Primary

  Degraded mode, e.g. caused by single or multiple storage removals or disk failures · CPC title

• G06F13/1673

  using buffers · CPC title

• G06F2212/214

  Solid state disk · CPC title

• G06F2212/1032

  Reliability improvement, data loss prevention, degraded operation etc · CPC title

• G06F2212/1016

  Performance improvement · CPC title