# US 2021/0279188 A1 · Client input/output (i/o) access rate variation compensation

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

- Number: US 2021/0279188 A1
- Document ID: US20210279188A1
- Status: Published application
- Kind: A1
- Application number: 16/810,357
- Filed: 2020-03-05
- Published: 2021-09-09
- Earliest claimed: 2020-03-05
- Assignees: Seagate Technology LLC
- Applicants: Seagate Technology, LLC
- Inventors: Ryan James Goss; David W. Claude; Graham David Ferris; Daniel John Benjamin; Ryan Charles Weidemann
- Web page: https://patentgrep.com/patent/US20210279188A1

## CPC

- G06F 13/1668: {Details of memory controller}
- G06F 11/3034: {where the computing system component is a storage system, e.g. DASD based or network based (digital input from or digital output to record carriers G06F3/06; digital recording or reproducing G11B20/18; for distributed storage of data in networks, e.g. transport arrangements for network file system [NFS], storage area networks [SAN] or network attached storage [NAS], H04L67/1097)}
- 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 11/3041: {where the computing system component is an input/output interface (interconnection of, or transfer of information or other signals between, memories, input/output devices or central processing units G06F13/00)}
- G06F 11/3423: {where the assessed time is active or idle time}
- G06F 11/3433: {for load management (allocation of a server based on load conditions G06F9/505; load rebalancing G06F9/5083; redistributing the load in a network by a load balancer H04L67/1029)}
- G06F 11/3485: {for I/O devices}
- G06F 11/3495: {for systems}
- G06F 12/0246: {in block erasable memory, e.g. flash memory}
- G06F 12/0253: {Garbage collection, i.e. reclamation of unreferenced memory}
- G06F 11/1048: {using arrangements adapted for a specific error detection or correction feature}
- G06F 2201/81: Threshold
- G06F 2212/1016: Performance improvement
- G06F 2212/1032: Reliability improvement, data loss prevention, degraded operation etc
- G06F 2212/1044: Space efficiency improvement
- G06F 2212/7205: Cleaning, compaction, garbage collection, erase control
- 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

Method and apparatus for enhancing performance of a storage device, such as a solid-state drive (SSD). In some embodiments, the storage device monitors a rate at which client I/O access commands are received from a client to transfer data with a non-volatile memory (NVM) of the storage device. A ratio of background access commands to the client I/O access commands is adjusted to maintain completion rates of the client I/O access commands at a predetermined level. The background access commands transfer data internally with the NVM to prepare the storage device to service the client I/O access commands, and can include internal reads and writes to carry out garbage collection and metadata map updates. The ratio may be adjusted by identifying a workload type subjected to the storage device by the client.

## Claims

### Claim 1

A method, comprising:
monitoring a rate at which client I/O access commands are provided from a client to transfer data between the client and a non-volatile memory (NVM) of a storage device; and
adjusting a ratio of controller-generated background access commands to the client-generated I/O access commands to maintain completion rates of the client-generated I/O access commands at a predetermined level, the controller-generated background access commands including garbage collection (GC) read commands to retrieve current version data blocks from a first garbage collection unit (GCU) and GC write commands to subsequently store the retrieved current version data blocks in a second GCU while the first GCU is erased, wherein a rate at which the GC write commands are carried out is selected as a function of the GC read commands, and a rate at which the GC read commands are carried out is selected as a function of a detected type of workload presented to the storage device by the client, to prepare the storage device to service the client-generated I/O access commands.

### Claim 2 (depends on 1)

The method of claim 1, wherein the adjusting step comprises using an incrementing mechanism to adjust a first number of the background access commands and a second number of the client I/O access commands to execute over each of a succession of time intervals to maintain the completion times at the predetermined level.

### Claim 3 (depends on 1)

The method of claim 1, wherein the background access commands comprise internal read and write commands to carry out garbage collection operations, wherein the client I/O access commands comprise client read and write commands to transfer data from and to the NVM, and wherein the adjusting step comprises selecting a first ratio of the internal read commands to the client read commands, a second ratio of the internal write commands to the client write commands, and a third ratio of total read commands to total write commands.

### Claim 4 (depends on 3)

The method of claim 3, wherein the background access commands further comprise internal read and write commands to carry out metadata map updates to describe client data stored to the NVM.

### Claim 5 (depends on 1)

The method of claim 1, further comprising accumulating history data associated with the client I/O access commands issued to the storage device from the client, identifying a workload type in relation to the accumulated history data, and adjusting the ratio in relation to the identified workload type.

### Claim 6 (depends on 5)

The method of claim 5, wherein the detected workload is at least a selected one of a write dominated environment, a read dominated environment, a sequential access environment or a random access environment.

### Claim 7 (depends on 1)

The method of claim 1, wherein the storage device is characterized as a solid-state drive (SSD) and the NVM is characterized as NAND flash memory.

### Claim 8 (depends on 1)

The method of claim 1, wherein the predetermined level of the completion rates maintained by the adjustment of the ratio of background access commands to the client I/O access commands comprises an upper specified threshold and a lower specified threshold, and the completion rates are maintained between the upper and lower specified thresholds.

### Claim 9

(canceled)

### Claim 10 (depends on 1)

The method of claim 1, wherein a first ratio is selected for a first interval during which first access commands are executed to transfer data with the NVM, and wherein an incrementing mechanism adjusts the first ratio to use a different, second ratio during an immediately successive second interval during which second access commands are executed to transfer data with the NVM, the incrementing mechanism limiting a difference interval between the first ratio and the second ratio by a maximum predetermined amount to apply a smoothing function to the completion rate.

### Claim 11 (depends on 1)

The method of claim 1, further comprising disabling the adjusting of the ratio to operate in an uncompensated mode in which variation in the completion rate exceeds the predetermined level.

### Claim 12

A storage device, comprising:
a non-volatile memory (NVM);
a controller configured to transfer data between the NVM and a client responsive to client-generated I/O access commands supplied to the storage device by the client; and
a command scheduler circuit configured to monitor a rate at which the client-generated I/O access commands are provided from the client, to identify a workload type responsive to the client-generated I/O access commands, and to adjust a ratio of controller-generated garbage collection commands to the client-generated I/O access commands, by setting a rate of garbage collection read commands as a function of the workload type, to maintain completion rates of the client-generated I/O access commands at a predetermined level.

### Claim 13 (depends on 12)

The storage device of claim 12, wherein the client I/O access commands comprise client read and write commands to transfer data between the client and the NVM, and wherein the background access commands comprise internal reads and writes to transfer data internally with the NVM to prepare the storage device to service the client read and write commands.

### Claim 14 (depends on 13)

The storage device of claim 13, wherein the command scheduler circuit selects a first ratio of the internal read commands to the client read commands, a second ratio of the internal write commands to the client write commands, and a third ratio of total read commands to total write commands.

### Claim 15 (depends on 13)

The storage device of claim 13, wherein the internal read and write commands are carried out for garbage collection operations to relocate current version data blocks in the NVM and for metadata map updates to describe client data stored to the NVM.

### Claim 16 (depends on 12)

The storage device of claim 12, wherein the detected workload is at least a selected one of a write dominated environment, a read dominated environment, a sequential access environment or a random access environment.

### Claim 17 (depends on 12)

The storage device of claim 12, wherein the predetermined level of the completion rates maintained by the adjustment of the ratio of background access commands to the client I/O access commands comprises an upper specified threshold and a lower specified threshold, and the command scheduler circuit further successively adjusts the ratio to maintain the completion rates between the upper and lower specified thresholds.

### Claim 18 (depends on 12)

The storage device of claim 12, wherein the command scheduler circuit further operates to add a delay at the completion of at least some of the client I/O accesses prior to providing a notification to the client that the associated at least some of the client I/O accesses have been completed to further maintain the completion rate within the predetermined level.

### Claim 19

A system comprising a client device and a storage device coupled to the client device, the storage device comprising a controller circuit, a non-volatile memory (NVM) and a command scheduler circuit, the command scheduler circuit configured to maintain a completion rate of client-generated I/O access commands issued by the client device to the storage device within a predetermined range by monitoring a rate at which the client-generated I/O access commands are provided from the client, identifying a workload type responsive to the client-generated I/O access commands, and adjusting a ratio of controller-generated garbage collection commands to the client-generated I/O access commands, by setting a rate of the garbage collection commands as a function of the workload type.

### Claim 20 (depends on 19)

The system of claim 19, wherein the storage device comprises a selected one of a solid-state drive (SSD), a hard disc drive (HDD) or a hybrid data storage device (HDSD).
