Adaptive redundancy control system, method, device and storage medium for memory

By introducing redundant bus design and adaptive redundant control in solid-state memory systems, the problems of low resource utilization and insufficient reliability in the existing technology are solved, component-level redundant control and switching management are realized, and the stars are improved The robustness and service life of NVMe-carrying solid state memory.

CN118689716BActive Publication Date: 2024-11-29WUHAN ZHUOMU TECH CO LTD
View PDF 1 Cites 0 Cited by

Patent Information

Application Number
CN202411176555.1
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2024-08-26
Publication Date
2024-11-29
Estimated Expiration
2044-08-26

AI Technical Summary

Technical Problem

In the prior art, redundant control is achieved by performing complete copying and backup of the entire solid-state memory system, and the resource utilization rate is low. When one component in the main and backup subsystem fails, the entire storage system cannot be used, resulting in solid-state memory. The system is less reliable.

Method used

Adopting a hierarchical solid-state memory architecture design, adding redundant bus design to the driver layer, realizing component-level adaptive redundancy control, monitoring component status through the control layer and issuing bus gate control instructions, the driver layer performs uplink and downlink bus selection Through operation, realize multi-component redundant combination mode to improve resource utilization and system stability.

Benefits of technology

It improves the robustness and service life of the on-site NVMe solid-state memory, ensures that the system can still be used normally when one component in the main and standby subsystem fails, and improves resource utilization and system stability.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118689716B_ABST
    Figure CN118689716B_ABST
Patent Text Reader

Abstract

The present invention proposes an adaptive redundant control system for a memory, including a main memory system and a backup memory system, both of which include a control layer, a drive layer and a storage medium layer, the control layer and the drive layer are connected via a control bus, and the drive layer and the storage medium layer are connected via a data bus; the drive layer of the main memory system and the drive layer of the backup memory system are connected via a redundant interconnection bus; the control layer is used to monitor the operating status of each component, and issue a bus gating control instruction according to the operating status of each component, and the drive layer performs uplink and downlink bus gating actions according to the bus gating control instruction. The present invention adopts a hierarchical solid-state memory architecture design, adds a redundant bus design in the drive layer, can realize adaptive redundant control at the component level, make redundant control and switching management more refined, realize a multi-component redundant combination mode, and improve resource utilization and system stability.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the field of memory control technology, and in particular to an adaptive redundant control system, method, device and storage medium of a memory. Background Art

[0002] With the rapid development of communication technology and aerospace technology, onboard memory technology has also been improved, and onboard solid-state memory based on NVMe technology has gradually developed and been applied. Onboard NVMe solid-state memory has the characteristics of high speed and large capacity, and its system composition is relatively complex and has high reliability requirements. At the same time, due to the influence of particle radiation in the satellite working environment, it is necessary to design redundant and fault-tolerant onboard memory to improve its reliability and service life.

[0003] Onboard NVMe solid-state storage uses CPU or SOC for storage control. When the main control chip or storage medium fails, the entire storage system corresponding to the main control chip will not work. Generally, redundancy design can be performed through hardware or software backup solutions to improve reliability.

[0004] A common redundant design scheme is to make a complete copy and backup of the entire solid-state storage system. When a component in the primary or backup storage subsystem fails, the storage subsystem is switched. At this time, the entire replaced storage subsystem and its supporting components cannot be used, and resource utilization is low. In more serious cases, when one component in each of the primary and backup subsystems fails, the entire storage system cannot be used, and the reliability of the solid-state storage system is low. Summary of the invention

[0005] The present invention proposes an adaptive redundant control system, method, device and storage medium for a memory, which solves the problems in the prior art of realizing redundant control by completely copying and backing up the entire solid-state memory system, resulting in low resource utilization, and when one component in each of the primary and standby subsystems fails, the entire storage system cannot be used, resulting in low reliability of the solid-state memory system.

[0006] The technical solution of the present invention is achieved in this way:

[0007] A first aspect of the present invention provides an adaptive redundant control system of a memory, comprising a main memory system and a backup memory system, wherein the main memory system and the backup memory system both comprise a control layer, a drive layer and a storage medium layer connected in sequence, wherein the control layer and the drive layer are connected via a control bus, and the drive layer and the storage medium layer are connected via a data bus; the drive layer of the main memory system and the drive layer of the backup memory system are connected via a redundant interconnection bus; the storage medium layer is used to provide a large-capacity storage space; the control layer is used to monitor the operating status of each component, and issue a bus selection control instruction according to the operating status of each component, and the drive layer performs an uplink and downlink bus selection action according to the bus selection control instruction.

[0008] The present invention adopts a hierarchical solid-state memory architecture design and adds a redundant bus design in the driver layer, which can realize adaptive redundant control at the component level, make redundant control and switching management more refined, realize a multi-component redundant combination mode, improve resource utilization and system stability, and greatly enhance the robustness and service life of the onboard NVMe solid-state memory.

[0009] Specifically, the driver layer includes a bus processing unit, and the bus processing unit includes:

[0010] A first uplink gating unit selects whether the data channel of the uplink bus is from the main memory system or the standby memory system according to the bus gating control instruction;

[0011] A second uplink gating unit selects whether the data channel of the uplink bus is to be sent to the main memory system or the backup memory system according to the bus gating control instruction;

[0012] A first downstream gating unit selects whether the data channel of the downstream bus is from the main memory system or the backup memory system according to the bus gating control instruction;

[0013] The second downstream gating unit selects whether the data channel of the downstream bus is directed to the main memory system or the backup memory system according to the bus gating control instruction.

[0014] Further, in the main memory system or the backup memory system: the downstream of the first uplink gating unit is connected to the storage medium layer, the upstream of the first uplink gating unit is connected to the second uplink gating unit, and the upstream of the second uplink gating unit is connected to the control layer; the upstream of the first downlink gating unit is connected to the control layer, the downstream of the first downlink gating unit is connected to the second downlink gating unit, and the downstream of the second downlink gating unit is connected to the storage medium layer;

[0015] Between the main memory system and the standby memory system: the downstream of the first upstream gating unit of the main memory system is connected to the second upstream gating unit of the standby memory system, and the upstream of the second upstream gating unit of the main memory system is connected to the first upstream gating unit of the standby memory system; the upstream of the first downstream gating unit of the main memory system is connected to the second downstream gating unit of the standby memory system, and the downstream of the second downstream gating unit of the main memory system is connected to the first downstream gating unit of the standby memory system.

[0016] Furthermore, the driver layer includes a bus protection unit, and the bus protection unit is used to refresh the program when an abnormality occurs in the bus processing unit.

[0017] Specifically, the driver layer includes a load and driver processing unit, and the load and driver processing unit is used for access processing, cache processing, and storage read and write driver processing of load data.

[0018] A second aspect of the present invention provides a method for adaptive redundancy control of a memory, comprising the following steps:

[0019] The control layer monitors the operating status of each component in real time and issues bus gating control instructions according to the operating status of each component;

[0020] Execute uplink and downlink bus gating actions according to bus gating control instructions to link the bus gating paths of the main memory system and the backup memory system;

[0021] The processing flow of the main memory system and the standby memory system is started according to the bus selection path.

[0022] Specifically, when abnormalities are detected in different components of the main memory system and / or the backup memory system, the bus processing unit of the driver layer cuts off the upstream and downstream bus selection paths of the abnormal components, and connects the normal components corresponding to the abnormal components in the main memory system or the backup memory system.

[0023] A third aspect of the present invention provides an electronic device, comprising a memory and a processor, wherein the memory stores a computer program that can be run on the processor, and the processor implements the steps of the above-mentioned adaptive redundancy control method when executing the computer program.

[0024] A fourth aspect of the present invention provides a computer-readable storage medium, wherein the storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the above-mentioned adaptive redundancy control method are implemented. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0026] Figure 1 A schematic diagram of a hierarchical architecture of an adaptive redundant control system of a memory according to the present invention;

[0027] Figure 2 It is a schematic diagram of the internal modules of an adaptive redundant control system of a memory of the present invention;

[0028] FIG3 (a) is a schematic diagram of bus connection of a redundant control system when an abnormality occurs in the standby control layer;

[0029] FIG3 (b) is a schematic diagram of bus connection of a redundant control system when a component of both the standby control layer and the standby drive layer is abnormal;

[0030] FIG3 (c) is a schematic diagram of bus connection of a redundant control system when a component of both the standby control layer and the main drive layer is abnormal;

[0031] FIG3 (d) is a schematic diagram of bus connection of a redundant control system when a component of the backup control layer, the backup storage medium layer and the main drive layer all fails;

[0032] Figure 4 Schematic diagram of a bus gating path when an abnormality occurs in a memory system according to an embodiment of the present invention. DETAILED DESCRIPTION

[0033] The technical solution of the present invention will be clearly and completely described below in conjunction with the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0034] Reference Figures 1 to 4In a first aspect, an embodiment of the present invention provides an adaptive redundant control system of a memory, comprising a main memory system and a backup memory system, wherein the main memory system and the backup memory system both comprise a control layer, a drive layer and a storage medium layer connected in sequence, wherein the control layer and the drive layer are connected via a control bus, and the drive layer and the storage medium layer are connected via a data bus; the drive layer of the main memory system and the drive layer of the backup memory system are connected via a redundant interconnection bus; the storage medium layer is used to provide a large-capacity storage space; the control layer is used to monitor the operating status of each component, and issue a bus gating control instruction according to the operating status of each component, and the drive layer performs uplink and downlink bus gating actions according to the bus gating control instruction.

[0035] The present invention adopts a hierarchical solid-state memory architecture design and adds a redundant bus design in the driver layer, which can realize adaptive redundant control at the component level, make redundant control and switching management more refined, realize a multi-component redundant combination mode, improve resource utilization and system stability, and greatly enhance the robustness and service life of the onboard NVMe solid-state memory.

[0036] like Figure 2 As shown, in a specific implementation manner, the management of the control layer is performed by the CPU, and the CPU is connected to the driver layer through an AXI4 bus; the bus redundancy and gating processing of the driver layer are completed in the FPGA, and the AXI4 bus is used to realize high-speed data bus connection, and the AXI-lite bus is used to realize low-speed control bus interconnection, and the AXI bus arbiter is used to realize bus gating; the driver layer also includes the interconnection between the FPGA and the DDR4 cache, and the interconnection bus is AXI4, and also includes the interconnection between the FPGA and the storage medium, and the interconnection bus is PCIE.

[0037] The driver layer is composed of load access, cache control, storage read and write, bus processing, bus protection, DDR4 and other units. Among them, the bus processing and bus protection units are implemented in FPGA. The bus processing unit is simple in composition and is a key unit for redundant protection. The bus processing unit is composed of registers and lookup table units in the FPGA chip, and can be protected using the FPGA refresh function to prevent the redundant backup function from failing due to bus abnormalities. The bus protection unit uses the built-in refresh unit of the FPGA. When an abnormality occurs in the bus processing unit, the program can be refreshed to ensure that the bus processing unit will not fail due to space particles and radiation factors. It can be seen that since the bus unit is configured with the FPGA self-refresh mechanism, the probability of an abnormality in the bus unit is extremely low. In this way, the method can complete the flexible switching of the main and standby systems based on bus switching. The storage medium layer is composed of SSDs based on the NVMe protocol, and this embodiment uses M.2 SSDs.

[0038] Specifically, the driver layer includes a bus processing unit, which completes functions such as control bus and data bus redundancy design, gating management, and status reporting; Figure 4 As shown, the bus processing unit includes:

[0039] A first uplink gating unit selects whether the data channel of the uplink bus is from the main memory system or the standby memory system according to the bus gating control instruction;

[0040] A second uplink gating unit selects whether the data channel of the uplink bus is to be sent to the main memory system or the backup memory system according to the bus gating control instruction;

[0041] A first downstream gating unit selects whether the data channel of the downstream bus is from the main memory system or the backup memory system according to the bus gating control instruction;

[0042] The second downstream gating unit selects whether the data channel of the downstream bus is directed to the main memory system or the backup memory system according to the bus gating control instruction.

[0043] like Figure 2 As shown, further, the driving layer includes a bus protection unit, and the bus protection unit is used to refresh the program when an abnormality occurs in the bus processing unit to ensure that the bus processing unit will not fail due to space particles and radiation factors.

[0044] Specifically, the driver layer includes a load and driver processing unit, and the load and driver processing unit is used for access processing, cache processing, and storage read and write driver processing of load data.

[0045] Specifically, the control layer of the solid-state memory system is composed of system software, which can monitor the functional status of each solid-state memory component and the strobe connection status of the driver layer bus. The system control software monitors the status of each layer in real time, so the redundant control of components and buses can be completed automatically without additional switching instructions.

[0046] A second aspect of an embodiment of the present invention provides a method for adaptive redundancy control of a memory, comprising the following steps:

[0047] The control layer monitors the operating status of each component in real time and issues bus gating control instructions according to the operating status of each component;

[0048] Execute uplink and downlink bus gating actions according to bus gating control instructions to link the bus gating paths of the main memory system and the backup memory system;

[0049] The processing flow of the main memory system and the standby memory system is started according to the bus selection path.

[0050] Specifically, when abnormalities are detected in different components of the main memory system and / or the backup memory system, the bus processing unit of the driver layer cuts off the upstream and downstream bus selection paths of the abnormal components, and connects the normal components corresponding to the abnormal components in the main memory system or the backup memory system.

[0051] like Figure 4 As shown, specifically, in the main memory system, the downstream of the first upstream gating unit (i.e., the upstream gating unit #1 in the figure) can be connected to the main storage medium layer, or can be switched to the second upstream gating unit (i.e., the upstream gating unit #2 in the figure) connected to the standby memory system; the upstream of the first upstream gating unit is connected to the second upstream gating unit; the upstream of the second upstream gating unit can be connected to the main control layer, or can be switched to the first upstream gating unit connected to the standby memory system; the downstream of the second downstream gating unit (i.e., the downstream gating unit #2 in the figure) can be connected to the main storage medium layer, or can be switched to the first downstream gating unit (i.e., the downstream gating unit #1 in the figure) connected to the standby memory system; the upstream of the second downstream gating unit is connected to the first downstream gating unit; the upstream of the first downstream gating unit can be connected to the main control layer, or can be switched to the second downstream gating unit connected to the standby memory system.

[0052] Similarly, in the backup memory system, the downstream of the first upstream gating unit can be connected to the backup storage medium layer, or can be switched to the second upstream gating unit connected to the main memory system; the upstream of the first upstream gating unit is connected to the second upstream gating unit; the upstream of the second upstream gating unit can be connected to the backup control layer, or can be switched to the first upstream gating unit connected to the main memory system; the downstream of the second downstream gating unit can be connected to the backup storage medium layer, or can be switched to the first downstream gating unit connected to the main memory system; the upstream of the second downstream gating unit is connected to the first downstream gating unit; the upstream of the first downstream gating unit can be connected to the backup control layer, or can be switched to the second downstream gating unit connected to the main memory system.

[0053] like Figure 4 As shown, this embodiment takes the case where both the primary storage medium layer and the standby control layer are abnormal as an example to illustrate the bus gating configuration scheme as follows:

[0054] In the main memory system, the upstream of the second uplink gating unit is connected to the main control layer, and the downstream is connected to the first uplink gating unit; the upstream of the first downlink gating unit is connected to the main control layer, and the downstream is connected to the second downlink gating unit;

[0055] In the backup memory system, the downstream of the second downlink gating unit is connected to the backup storage medium layer, and the upstream is connected to the first downlink gating unit; the downstream of the first uplink gating unit is connected to the backup storage medium layer, and the upstream is connected to the second uplink gating unit;

[0056] Between the main memory system and the standby memory system, the downstream of the first uplink gating unit of the main memory system is connected to the second uplink gating unit of the standby memory system; the downstream of the second downlink gating unit of the main memory system is connected to the first downlink gating unit of the standby memory system.

[0057] In this embodiment, the working process of the solid-state memory system is as follows:

[0058] In the first step, the main CPU completes initialization, and the standby CPU completes initialization;

[0059] In the second step, the active and standby CPUs detect the status of each component and the driver layer bus and determine the bus working mode. If all components and bus status are normal, the full working mode is entered and the active and standby CPUs can work independently;

[0060] In the third step, the active and standby CPUs configure their respective driver layers and storage medium layers;

[0061] The fourth step is to complete the system initialization and function configuration, and then the solid-state memory system enters the normal working state. After receiving external instructions, it can enter the storage recording or playback process.

[0062] In a specific implementation manner, when an abnormality occurs in the solid-state memory system, the specific workflow is as follows:

[0063] As shown in (a) of FIG3 , when an abnormality occurs in the standby control layer, the main memory system is under normal control, the standby memory system cuts off the control bus between the control layer and the drive layer, and the bus processing unit of the standby drive layer is connected to the bus processing unit of the main drive layer through a redundant interconnection bus.

[0064] As shown in (b) of Figure 3, when a component of the standby control layer and the standby drive layer (such as the load and drive processing unit) is abnormal, the main memory system controls normally, the standby memory system cuts off the control bus between the control layer and the drive layer, the uplink and downlink buses of the load and drive processing units of the standby drive layer are cut off, and the bus processing unit of the standby drive layer is connected to the bus processing unit of the main drive layer through a redundant interconnection bus.

[0065] As shown in (c) of Figure 3, when an abnormality occurs in a component of the standby control layer and the main drive layer (such as the load and drive processing unit), the standby memory system cuts off the control bus of the control layer and the drive layer, cuts off the uplink and downlink buses of the load and drive processing unit of the main drive layer, and the bus processing unit of the standby drive layer is connected to the bus processing unit of the main drive layer through a redundant interconnection bus.

[0066] As shown in (d) of FIG3 , when a component of the backup control layer, the backup storage medium layer, and the main drive layer (such as the load and drive processing unit) is abnormal, the backup memory system cuts off the control bus / data bus between the drive layer and the control layer and the storage medium layer, cuts off the uplink and downlink buses of the load and drive processing unit of the main drive layer, and the bus processing unit of the backup drive layer is connected to the bus processing unit of the main drive layer through a redundant interconnection bus.

[0067] In a specific implementation, the abnormality reporting process of each component in the driver layer is as follows:

[0068] In the first step, each component of the driver layer puts the status timing into the status register of the bus processing unit;

[0069] In the second step, the control layer configures the gating path of the bus processing unit according to the default state;

[0070] In the third step, the bus processing unit completes the packaging of all status registers and uploads the status package to the corresponding control layer according to the bus selection path.

[0071] The present invention proposes a method of automatic redundant control and bus interconnection of the system, and adopts a hierarchical design to realize refined and adaptive redundant management functions. Compared with the traditional redundant backup method, the system reliability is improved. Compared with the traditional method of completely copying and backing up the entire solid-state memory system to achieve redundant control, the control method of the present invention has a higher resource utilization rate. When one component in each of the primary and standby subsystems fails, the entire storage system can still be used normally, and the reliability of the solid-state memory system is relatively high; as long as the same component in the primary and standby systems does not fail at the same time, the normal operation of the system can be guaranteed.

[0072] A third aspect of an embodiment of the present invention provides an electronic device, including a memory and a processor, wherein the memory stores a computer program that can be run on the processor, and the processor implements the steps of the above-mentioned adaptive redundancy control method when executing the computer program.

[0073] A fourth aspect of an embodiment of the present invention provides a computer-readable storage medium, wherein the storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the above-mentioned adaptive redundancy control method are implemented.

[0074] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. An adaptive redundant control system for a memory, characterized in that: It includes a main memory system and a backup memory system, each of which includes a control layer, a drive layer and a storage medium layer connected in sequence, the control layer and the drive layer are connected via a control bus, and the drive layer and the storage medium layer are connected via a data bus; the drive layer of the main memory system and the drive layer of the backup memory system are connected via a redundant interconnection bus; The storage medium layer is used to provide a large-capacity storage space; The control layer is used to monitor the operating status of each component and issue bus gating control instructions according to the operating status of each component, and the driving layer performs uplink and downlink bus gating actions according to the bus gating control instructions; The driver layer includes a bus processing unit, and the bus processing unit includes: A first uplink gating unit selects whether the data channel of the uplink bus is from the main memory system or the standby memory system according to the bus gating control instruction; A second uplink gating unit selects whether the data channel of the uplink bus is to be sent to the main memory system or the backup memory system according to the bus gating control instruction; A first downstream gating unit selects whether the data channel of the downstream bus is from the main memory system or the backup memory system according to the bus gating control instruction; The second downstream gating unit selects whether the data channel of the downstream bus is directed to the main memory system or the backup memory system according to the bus gating control instruction.

2. The adaptive redundant control system of the memory as claimed in claim 1, characterized in that: In the main memory system or the backup memory system: the downstream of the first uplink gating unit is connected to the storage medium layer, the upstream of the first uplink gating unit is connected to the second uplink gating unit, and the upstream of the second uplink gating unit is connected to the control layer; the upstream of the first downlink gating unit is connected to the control layer, the downstream of the first downlink gating unit is connected to the second downlink gating unit, and the downstream of the second downlink gating unit is connected to the storage medium layer; Between the main memory system and the standby memory system: the downstream of the first upstream gating unit of the main memory system is connected to the second upstream gating unit of the standby memory system, and the upstream of the second upstream gating unit of the main memory system is connected to the first upstream gating unit of the standby memory system; the upstream of the first downstream gating unit of the main memory system is connected to the second downstream gating unit of the standby memory system, and the downstream of the second downstream gating unit of the main memory system is connected to the first downstream gating unit of the standby memory system.

3. The adaptive redundant control system of the memory as claimed in claim 1, characterized in that: The driver layer includes a bus protection unit, and the bus protection unit is used to refresh the program when an abnormality occurs in the bus processing unit.

4. The adaptive redundant control system of the memory as claimed in claim 1, characterized in that: The driver layer includes a load and driver processing unit, which is used for access processing, cache processing, and storage read and write driver processing of load data.

5. A method for adaptive redundancy control of a memory, based on the adaptive redundancy control system according to any one of claims 1 to 4, characterized in that: The following steps are involved: The control layer monitors the operating status of each component in real time and issues bus gating control instructions according to the operating status of each component; Execute uplink and downlink bus gating actions according to bus gating control instructions to link the bus gating paths of the main memory system and the backup memory system; The processing flow of the main memory system and the standby memory system is started according to the bus selection path.

6. The adaptive redundancy control method of a memory as claimed in claim 5, characterized in that: When abnormalities are detected in different components in the main memory system and / or the backup memory system, the bus processing unit of the driver layer cuts off the upstream and downstream bus selection paths of the abnormal components and connects the normal components corresponding to the abnormal components in the main memory system or the backup memory system.

7. An electronic device comprising a memory and a processor, wherein the memory stores a computer program that can be run on the processor, characterized in that: When the processor executes the computer program, the steps of the adaptive redundancy control method according to claim 5 or 6 are implemented.

8. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the steps of the adaptive redundancy control method according to claim 5 or 6 are implemented.

Citation Information

Patent Citations

  • Enabling communication between a single-port device and multiple storage system controllers

    WO2023130060A1