# US 2021/0081273 A1 · Method and System for Host-Assisted Data Recovery Assurance for Data Center Storage Device Architectures

> Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt

- Number: US 2021/0081273 A1
- Document ID: US20210081273A1
- Status: Published application
- Kind: A1
- Application number: 16/739,098
- Filed: 2020-01-09
- Published: 2021-03-18
- Earliest claimed: 2019-09-17
- Assignees: Western Digital Technologies, Inc.
- Applicants: Western Digital Technologies, Inc.
- Inventors: Daniel Lee HELMICK; Cory James PETERSON; Jay SARKAR
- Web page: https://patentgrep.com/patent/US20210081273A1

## CPC

- G06F 11/108: {Parity data distribution in semiconductor storages, e.g. in SSD}
- G06F 9/30189: {according to execution mode, e.g. mode flag}
- G06F 9/4881: {Scheduling strategies for dispatcher, e.g. round robin, multi-level priority queues}
- G06F 11/0772: {Means for error signaling, e.g. using interrupts, exception flags, dedicated error registers}
- G06F 11/1068: {in sector programmable memories, e.g. flash disk (G06F11/1072 takes precedence)}
- G06F 11/3037: {where the computing system component is a memory, e.g. virtual memory, cache (accessing, addressing or allocating within memory systems or architectures G06F12/00; checking stores for correct operation G11C29/00)}
- G06F 12/0246: {in block erasable memory, e.g. flash memory}
- G06F 12/1408: {by using cryptography (for digital transmission H04L9/00)}
- G06F 11/2094: {Redundant storage or storage space (G06F11/2056 takes precedence)}
- G06F 2212/1032: Reliability improvement, data loss prevention, degraded operation etc
- G06F 2212/7201: Logical to physical mapping or translation of blocks or pages
- G06F 2212/7207: management of metadata or control data
- G06F 2212/7208: Multiple device management, e.g. distributing data over multiple flash devices
- G06F: ELECTRIC DIGITAL DATA PROCESSING (computer systems based on specific computational models G06N)

## Abstract

A method of error management includes, in response to a read request for first data from a first storage device of a plurality of storage devices under one or more common data protection schemes, receiving a read uncorrectable indication regarding the first data, obtaining uncorrected data and metadata of an LBA associated with the first data, and obtaining the same LBA from one or more other storage devices of the plurality. The method further includes comparing the uncorrected data with the data and metadata from the other storage devices, speculatively modifying the uncorrected data based, at least in part, on the other data to create a set of reconstructed first data codewords, and, in response to a determination that one of the reconstructed first data codewords has recovered the first data, issuing a write_raw command to rewrite the modified data and associated metadata to the first storage device.

## Claims

### Claim 1

A method of error management for data storage and solid-state storage devices (SSD), comprising:
in response to a read request for first data from a first storage device of a plurality of storage devices under one or more common data protection schemes, receiving a read uncorrectable indication regarding the first data;
issuing a read_raw command to obtain uncorrected data and metadata of the logical block address (LBA) associated with the first data;
obtaining the same LBA, or the same first data stored at another LBA, from one or more other storage devices of the plurality of storage devices;
comparing the uncorrected data with the data and metadata from the other storage devices; and
speculatively modifying the uncorrected data based at least in part on the other data to create a set of reconstructed first data codewords; and
in response to a determination that one of the reconstructed first data codewords has recovered the first data, issuing a write_raw command to rewrite the modified data and associated metadata to the first storage device.

### Claim 2 (depends on 1)

The method of claim 1, further comprising:
in response to a determination that the reconstructed first data codewords cannot recover the first data, obtaining the same LBA, or the same user data stored at another LBA, from one or more additional other storage devices of the plurality.

### Claim 3 (depends on 1)

The method of claim 1, further comprising reading the corrected data at the first storage device to verify that the corrected data is there.

### Claim 4 (depends on 3)

The method of claim 3, further comprising reading a buffer that uses a same read path as is used for a media read recovery.

### Claim 5 (depends on 1)

The method of claim 1, further comprising rewriting the modified data and metadata to at least one of:
one or more other storage devices of the plurality of storage devices; or
a RAM buffer.

### Claim 6 (depends on 1)

The method of claim 1, wherein comparing the first data with the data and metadata from the one or more other storage devices further includes deducing specific bits of the LBA that are in error.

### Claim 7 (depends on 1)

The method of claim 1, wherein the plurality of storage devices are disposed in a data center, and together comprise at least one of a RAID protected, erasure coded or replicated system.

### Claim 8 (depends on 1)

The method of claim 1, further comprising:
in response to a determination that the first data cannot be corrected at the LBA level, issuing an extended_read_raw command to obtain data and metadata having a size larger than a logical block that is associated with the first data, from one or more other storage devices of the plurality.

### Claim 9 (depends on 8)

The method of claim 8, further comprising:
in response to the extended read_raw command, obtaining at least one of a full redundant array of inexpensive drives (RAID) stripe of data or some or all of an entire erasure code block.

### Claim 10 (depends on 1)

The method of claim 1, wherein speculatively modifying the uncorrected data further comprises:
characterizing the error modes of the other storage devices and using that characterization to make a best guess as to the actual values of erroneous bits of the first data.

### Claim 11 (depends on 1)

The method of claim 1, wherein speculatively modifying the uncorrected data further comprises:
obtaining multiple versions of the LBA of the first data, and applying one of a voting scheme or a weighted voting scheme to the multiple versions.

### Claim 12 (depends on 1)

The method of claim 1, wherein speculatively modifying the uncorrected data further comprises:
determining if the data from the one or more other storage devices of the plurality of storage devices uses the same HLBAs as the first storage device; and
in response to a determination that they do, including reconstructed first data codewords where ECC bits of the uncorrected data are modified in addition to bits of user data.

### Claim 13

A non-volatile data storage device, comprising:
a storage device host configured to:
receive a read uncorrectable event response for a first codeword (CW) stored in the storage device;
issue a read_raw command for the CW and a read_raw extended command for each of other CWs protected in a data protection stripe that includes the first CW;
construct a pseudo_read_raw result for the first CW from each of the other CWs and the data protection stripe; and
using the pseudo_read_raw result, construct a set of possible replacement CWs for the first CW and push one or more elements of the set to the storage device.

### Claim 14 (depends on 13)

The non-volatile data storage device of claim 13, wherein the data protection stripe is an XOR stripe.

### Claim 15 (depends on 13)

The data storage device of claim 13, wherein the storage device host is further configured to:
determine if one of the one or more elements of the set of possible replacement CWs recovered original data of the first CW; and
in response to a determination that the original data of the first CW is recovered, issue a write_raw command to rewrite the corrected first CW to the storage device.

### Claim 16 (depends on 15)

The data storage device of claim 15, wherein the storage device host is further configured to:
in response to a determination that the original data of the first CW is not recovered:
identify a second CW of the other CWs protected in the data protection stripe that protects the first CW;
construct a pseudo_read_raw result for the second CW from each of the other CWs and the data protection stripe; and
using the pseudo_read_raw result for the second CW, construct a set of possible replacement CWs for the second CW and push one or more elements of the set to the storage device.

### Claim 17 (depends on 15)

The data storage device of claim 15, wherein the storage device host is further configured to:
in response to a determination that the original data of the second CW is also not recovered, proceed to even higher levels of a data hierarchy of which the data storage device is a part, leveraging layers of data protection at each level, to recover the first CW and the second CW.

### Claim 18

An apparatus, comprising:
means for receiving an indication of uncorrectable data in response to a read command regarding first data stored on a first SSD of a first data center;
means for issuing a read_raw command to obtain data and metadata of one or more LBAs in which the uncorrectable data is stored;
means for issuing a read_raw_extended command to obtain the data and metadata of the one or more LBAs, or the same first data stored at one or more other LBAs from one or more other SSDs (“other data”), the other SSDs being either in the first data center or in other data centers, wherein the first and the other data centers are part of at least one of a RAID protected, erasure coded, or replicated system;
means for speculatively correcting the first data using the other data; and
means for issuing a write_raw command to re-write the speculatively corrected data to the first SSD, in response to receiving an indication that the speculatively corrected data recovered the first data.

### Claim 19 (depends on 18)

The apparatus of claim 18, wherein the means for speculatively correcting further comprises:
means for characterizing the error modes of the first SSD, and
means for using an error mode characterization to make a best guess as to the actual values of the first data.

### Claim 20 (depends on 19)

The apparatus of claim 19, wherein the means for using an error mode characterization further comprises means for making a prioritized ordering of possible values of the best guess.
