Raid storage device and method of management thereof
AssigneeSamsung Electronics Co., Ltd.
InventorHYUNSIK PARK
Life of this patent
Estimating the term…
Abstract
A method of managing a plurality of storage devices controlled in a RAID scheme includes detecting a failure disk among the storage devices, selecting one of a plurality of stripes of the storage devices according to a result of the detection, identifying a valid page of data included in a normal chunk of the selected stripe and a valid page of data of a lost chunk mapped to the failure disk with reference to address mapping information, recovering the valid page of the lost chunk among chunks included in the selected stripe with reference to information on the identified valid page, and copying the recovered valid page of the lost chunk and the valid page of the normal chunk to a new stripe of the storage devices.
Claims23
A method of managing a plurality of storage devices controlled in a redundant array of inexpensive disks (RAID) scheme, the method comprising:
- detecting a failure disk among the storage devices;
- selecting one of a plurality of stripes of the storage devices according to a result of the detection;
- identifying a valid page of data included in a normal chunk of data of the selected stripe, and a valid page of data of a lost chunk mapped to the failure disk with reference to address mapping information;
- recovering the valid page of the lost chunk among chunks included in the selected stripe with reference to information on the identified valid page;
- and
- copying the recovered valid page of the lost chunk and the valid page of the normal chunk to a new stripe of the storage devices.
A redundant array of inexpensive disks (RAID) storage device, comprising:
- a plurality of solid state drives;
- and
- a host interface configured to connect the solid state drives to a host,
- wherein among pages of a lost chunk included in a failure disk among the solid state drives, only valid pages are recovered during a RAID rebuild operation to recover the lost chunk using a parity chunk.
18-19. (canceled)
A method of managing a plurality of solid state drives, the method comprising:
- selecting one of a plurality of stripes allocated to manage the solid state drives in a log-structured redundant array of inexpensive disks (RAID) scheme;
- identifying a valid page included in a normal chunk of the selected stripe and a valid page of a lost chunk mapped to a failure disk among the plurality of solid state drives with reference to:
- a first map page which maps a logical address of a page included in the selected stripe to a physical address, and a reverse map page which maps the physical address of the page included in the selected stripes to the logical address;
- recovering a valid page included in the lost chunk among chunks included in the selected stripe with reference to information on the identified valid page;
- and
- copying the recovered valid page of the lost chunk and the valid page of the normal chunk to a new stripe.
(canceled)
24-34. (canceled)
Dependencies are inferred from the claim text.