# US 2026/0244542 A1 · Port-binding disaster recovery method and apparatus for storage system, and device and non-volatile readable storage medium

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

- Number: US 2026/0244542 A1
- Document ID: US20260244542A1
- Status: Published application
- Kind: A1
- Application number: 19/162,466
- Filed: 2024-09-29
- Published: 2026-08-20
- Earliest claimed: 2023-12-28
- Assignees: Suzhou Metabrain Intelligent Technology Co., Ltd.
- Applicants: Suzhou Metabrain Intelligent Technology Co., Ltd.
- Inventors: Yankai ZHANG; Ximeng ZHOU
- Web page: https://patentgrep.com/patent/US20260244542A1

## CPC

- G06F 11/201: {between storage system components}
- G06F 11/2005: {using redundant communication controllers}
- G06F: ELECTRIC DIGITAL DATA PROCESSING (computer systems based on specific computational models G06N)

## Abstract

Provided in the present application are a port-binding disaster recovery method and apparatus for a storage system, and a device and a non-volatile readable storage medium. The method comprises: respectively binding ports of a first device and ports of a second device, so as to form several communication paths, and calculating the performance of each communication path; grouping the communication paths on the basis of the performance of the communication paths, and performing data transmission on the basis of groups; and in response to an abnormality occurring in communication paths, adjusting the groups of the communication paths on the basis of the cause of the abnormality and the number of communication paths in each group. By means of the solution of the present application, the stability and disaster recovery capability of communication services can be improved in a scenario where a plurality of paths have faults; an enabled path can be switched and used to share service pressure in a scenario where links are congested and the performance pressure is large, and the storage performance can be improved at a port level, so as to ensure the smoothness and stability of client services; and the stable switching of a communication path can be realized, and risk early warning can be performed, so as to reduce the risk of a path having a fault or being missing.

## Claims

### Claim 1

A method for storage system port binding disaster recovery, comprising:
respectively binding a port of a first device and a port of a second device to form a plurality of communication paths, and calculating a performance of each communication path;
dividing the communication paths into groups based on the performance of the communication paths and performing data transmission based on the groups; and
in response to an abnormality in the communication path, adjusting a group-dividing way of the communication paths based on the abnormality cause and a quantity of communication paths in the groups;
wherein the step of binding a port of a first device and a port of a second device to form a communication path comprises:
binding a RoCE port of the first device with a RoCE port of the second device to form a communication path;
respectively creating a virtual network interface card of the RoCE port of the first device and a virtual network interface card of the RoCE port of the second device; and
determining whether the first device is capable of communicating with the second device.

### Claim 2 (depends on 1)

The method according to claim 1, wherein the step of dividing the communication paths into groups based on the performance of the communication paths and performing data transmission based on the groups comprises:
ranking the performances of communication paths from high to low;
selecting a threshold quantity of communication paths with a preceding performance ranking as members of an available group for data transmission; and
taking remaining communication paths as members of a backup group.

### Claim 3 (depends on 2)

The method according to claim 2, wherein,
selecting a threshold quantity of communication paths with a preceding performance ranking as members of an available group for data transmission comprises:
determining the communication paths with a first 50% of the performance ranking as members of the available group;
taking remaining communication paths as members of a backup group comprises:
determining the communication paths with a last 50% of the performance ranking as members of the backup group.

### Claim 4 (depends on 3)

The method according to claim 3, wherein
determining the communication paths with a first 50% of the performance ranking as members of the available group comprises:
setting the communication paths with the first 50% of the performance ranking as members of an active state, wherein the communication paths in the active state are used for data transmission;
and determining the communication paths with a last 50% of the performance ranking as members of the backup group comprises:
setting the communication paths with the last 50% of the performance ranking to an enabled state, wherein the communication paths in the enabled state are used for taking over abnormal communication paths in the available group for data transmission when an abnormity occurs to the communication paths in the available group.

### Claim 5 (depends on 2)

The method according to claim 2, wherein the step of in response to an abnormality in the communication path, adjusting a group-dividing way of the communication paths based on the abnormality cause and a quantity of communication paths in the groups comprises:
determining the abnormity cause of the communication path in response to an occurrence of an abnormity of the communication path;
executing a first preset strategy in response to the abnormity cause of the communication path being a path disconnection; and
executing a second preset strategy in response to the abnormity cause of the communication path being path congestion.

### Claim 6 (depends on 5)

The method according to claim 5, wherein the step of executing a first preset strategy in response to the abnormity cause of the communication path being a path disconnection comprises:
in response to the abnormity cause of the communication path being path disconnection, marking a disconnected communication path as an unavailable path, and determining a group-dividing way of the disconnected communication path;
in response to the disconnected communication path being grouped into the backup group, acquiring the quantity of communication paths of the backup group, and performing a first preset operation based on an acquired quantity; and
in response to the disconnected communication path being grouped into the available group, acquiring the quantity of communication paths of the backup group, and performing a second preset operation based on an acquired quantity.

### Claim 7 (depends on 6)

The method according to claim 6, wherein the step of in response to the disconnected communication path being grouped into the backup group, acquiring the quantity of communication paths of the backup group, and performing a first preset operation based on the acquired quantity comprises:
acquiring the quantity of communication paths of the backup group in response to the disconnected communication path being grouped into the backup group;
in response to the quantity of communication paths of the backup group being less than 1, sending a warning of no redundant path; and
in response to the quantity of communication paths of the backup group being equal to 1, clearing the warning of no redundant path, and sending a warning of insufficient redundant paths.

### Claim 8 (depends on 6)

The method according to claim 6, wherein the step of in response to the disconnected communication path being grouped into the available group, acquiring the quantity of communication paths of the backup group, and performing a second preset operation based on an acquired quantity comprises:
acquiring the quantity of communication paths of the backup group in response to the disconnected communication path being grouped into the available group;
in response to the quantity of communication paths of the backup group being equal to 1, switching the communication paths of the backup group to the available group and setting the same to active state and sending a warning of no redundant path;
in response to the quantity of communication paths of the backup group being greater than 1, calculating performances of all the communication paths in the backup group;
switching communication path with a highest performance to the available group and setting the same to the active state; and
in response to the quantity of communication paths of the backup group being less than 1, sending a warning of no redundant path.

### Claim 9 (depends on 5)

The method according to claim 5, wherein the step of executing a second preset strategy in response to the abnormity cause of the communication path being path congestion comprises:
in response to the abnormity cause of the communication path being path congestion, calculating the performances of the communication path in which congestion occurs and all communication paths in the backup group;
in response to the communication path with a highest performance being a communication path where congestion occurs, taking no action; and
in response to the communication path with the highest performance being a communication path in the backup group, switching the communication path with the highest performance in the backup group to the available group and setting the same to the active state.

### Claim 10

(canceled)

### Claim 11 (depends on 1)

The method according to claim 1, wherein the step of binding a RoCE port of the first device with a RoCE port of the second device to form a communication path comprises:
creating a one-to-one connection between each RoCE port of the first device and each RoCE port of the second device, respectively, to form several communication paths.

### Claim 12 (depends on 1)

The method according to claim 1, wherein the step of determining whether the first device is capable of communicating with the second device comprises:
respectively configuring an IP for a first virtual network interface card corresponding to a RoCE port of the first device and a second virtual network interface card corresponding to a RoCE port of the second device;
determining whether a network of the first device can be connected with a network of the second device; and
in response to the network of the first device being able to be connected with the network of the second device, sending information of the first device to the second device, thereby causing the second device to configure according to the information of the first device.

### Claim 13 (depends on 12)

The method according to claim 12, wherein the first device is a host, the second device is a storage node, a RoCE port of the host and a RoCE port of the storage node are connected via an optical fiber, and before respectively configuring an IP for a first virtual network interface card corresponding to a RoCE port of the first device and a second virtual network interface card corresponding to a RoCE port of the second device, the method further comprises:
configuring a BOND port binding through the RoCE port of the host to generate a first virtual network interface card of the host; and
configuring the BOND port binding through the RoCE port of the storage node to generate a second virtual network interface card of the storage node.

### Claim 14 (depends on 12)

The method according to claim 12, wherein the first device is a host and the second device is a storage node, and the step of sending information of the first device to the second device, thereby causing the second device to configure according to the information of the first device comprises:
the storage node using information of the host to configure host management; and
the host discovering the storage node using a first command and connecting to the storage node using a second command to complete a configuration of the host and the storage node.

### Claim 15 (depends on 14)

The method according to claim 14, wherein the storage node using information of the host to configure host management comprises:
the storage node acquiring a unique identification of the host and using the unique identification of the host to configure host management.

### Claim 16 (depends on 14)

The method according to claim 14, wherein the step of the host discovering the storage node using a first command and connecting to the storage node using a second command to complete a configuration of the host and the storage node comprises:
the host discovering the storage node using an NVMe discover command and connecting to the storage nodes using an NVMe connect command, wherein the first command comprises the NVMe discover command and the second command comprises the NVMe connect command.

### Claim 17 (depends on 1)

The method according to claim 1, wherein the first device is a first storage node and the second device is a second storage node, and the step of determining whether the first device is capable of communicating with the second device comprises:
respectively configuring an IP for a first virtual network interface card corresponding to a RoCE port of a first storage node and a second virtual network interface card corresponding to a RoCE port of a second storage node;
determining whether a network of the first storage node is able to connect with a network of the second storage node; and
in response to the network of the first storage node being able to connect with the network of the second storage node, creating a cluster based on the first storage node and the second storage node.

### Claim 18

(canceled)

### Claim 19 (depends on 1)

A computing device, comprising:
at least one processor; and
a memory storing thereon computer instructions executable on the processor, wherein the instructions, when executed by the processor, implement the steps of the method according to claim 1.

### Claim 20 (depends on 1)

A non-volatile readable computer storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the steps of the method according to claim 1.

### Claim 21 (depends on 9)

The method according to claim 9,
wherein the step of in response to the communication path with the highest performance being a communication path in the backup group, switching the communication path with the highest performance in the backup group to the available group and setting the same to the active state comprises:
in response to the communication path with the highest performance being a communication path in the backup group, switching the communication path with the highest performance in the backup group to the available group and setting the same to the active state, and adding the communication path in which congestion occurs to the backup group.
