Power failure protection method and system for storage server equipment and computer device

Abstract

The invention relates to a power failure protection method and system for storage server equipment and a computer device, and the method comprises the steps: monitoring an output signal of a power supply unit through a CPLD, and generating a non-shielding interruption signal when the output signal of the power supply unit is recognized to be abnormal; the power supply switching control unit controls a power supply mode to be switched from the power supply unit to the battery backup unit, controls related necessary hardware to enter a core closing mode through the ACPI and hardware cooperation module, and interrupts power supply to unnecessary hardware; controlling a dirty data identification unit to read a dirty data bitmap marked by the storage server equipment, and positioning dirty data to be protected in a memory of the storage server equipment; the dirty data transferred and stored in the preset storage unit are verified through the cyclic redundancy check unit, when the dirty data passes verification, the storage server device is controlled to execute the shutdown process, and when the power-on signal is received again, the dirty data are read from the preset storage unit and recovered to the memory cache.

Description

Technical Field

[0001] This invention relates to the field of storage server technology, and in particular to a method, system, and computer device for power failure protection of storage server equipment. Background Technology

[0002] With the rapid development of information technology, storage servers play a crucial role in data centers and various computing environments. However, in actual operation, storage server devices may face abnormal power outages of the power supply unit (PSU), which can lead to serious consequences such as data loss, system crashes, and business interruptions.

[0003] In related technologies, when the power supply unit malfunctions, existing storage server devices may not be able to quickly switch to backup power mode, or may not be able to effectively control hardware power consumption and operating status during the switching process. This may lead to system instability during power outages, or even failure to resume normal operation. Furthermore, the data transfer and protection mechanisms at the moment of power failure are often inadequate, failing to ensure that all corrupted data is saved promptly and completely.

[0004] When the power supply unit experiences an abnormal power outage, existing storage server devices typically cannot precisely control hardware power consumption. This leads to excessive power consumption when powered by the battery backup unit, shortening battery life and increasing the risk of data loss. Furthermore, existing technologies lack effective verification mechanisms during data transfer, failing to ensure data integrity and consistency. Data transfer failures or errors can result in data inconsistencies within the storage pool, impacting normal business operations. Summary of the Invention

[0005] The present invention provides a method, system and computer device for power failure protection of storage server equipment, aiming to solve at least one of the technical problems existing in the prior art.

[0006] The technical solution of the present invention is a power failure protection method for a storage server device, which includes: The output signal of the power supply unit is monitored in real time by CPLD. When the abnormal output signal of the power supply unit is detected, it is determined that the power supply unit is in an abnormal power failure state and an unmaskable interrupt signal is generated. In response to the non-maskable interrupt signal, the power supply switching control unit controls the power supply mode to switch from the power supply unit to the battery backup unit, and controls the relevant necessary hardware to enter the core shutdown mode through ACPI and hardware coordination module, and interrupts the power supply to the non-necessary hardware. When an indication signal is received that the power consumption of the storage server device has decreased to a preset threshold, the dirty data identification unit is controlled to read the dirty data bitmap marked by the storage server device and locate the dirty data to be protected in the memory of the storage server device. The dirty data is transferred to a preset storage unit via the PCIe channel, and the dirty data transferred to the preset storage unit is verified by the cyclic redundancy check unit to verify the integrity of the transferred dirty data. When the dirty data verification passes, the storage server device is controlled to execute the shutdown process, and the battery backup unit is controlled to stop supplying power. When the power-on signal is received again, the dirty data is read from the preset storage unit and restored to the memory cache, so that the storage server device is restored to normal operation.

[0007] According to some embodiments of the present invention, the step of monitoring the output signal of the power supply unit in real time via CPLD, and determining that the power supply unit is in an abnormal power-off state when an abnormal output signal of the power supply unit is detected, includes: The POK signal, output voltage, and AC-OK signal output by the power supply unit are monitored in real time using a CPLD. When the POK signal output by the power supply unit is low, it is determined that the output signal of the power supply unit is abnormal, and the power supply unit is in an abnormal power-off state. When the output voltage of the power supply unit is lower than the preset value, it is determined that the output signal of the power supply unit is abnormal, and the power supply unit is in an abnormal power failure state. When the AC-OK signal output by the power supply unit is lost or is low, it is determined that the output signal of the power supply unit is abnormal, and the power supply unit is in an abnormal power-off state.

[0008] According to some embodiments of the present invention, the step of controlling the necessary hardware to enter a core-shutdown mode and interrupting the power supply to the non-essential hardware via the ACPI and hardware coordination module includes: The relevant necessary hardware and non-necessary hardware are accurately identified by a preset hardware priority list. The relevant necessary hardware includes a first CPU and a second CPU, and the non-necessary hardware includes memory associated with the first CPU, PCIe peripherals, and non-critical hard drives. By configuring ACPI parameters in the BIOS of the storage server device, the ACPI and hardware coordination module triggers the first CPU to enter the core shutdown mode based on the ACPI parameters, and controls the second CPU to reduce its frequency to the lowest operating frequency. The ACPI and hardware coordination module disconnects the power supply to the first CPU-associated memory, the PCIe peripherals, and the non-critical hard drive.

[0009] According to some embodiments of the present invention, the necessary hardware further includes the core memory of the storage server device, a storage unit, a controller with built-in redundant hard disks and fan groups, and the method further includes: The battery backup unit provides power to the second CPU, the core memory of the storage server device, the storage unit, and the redundant hard drive built into the controller. The storage server device's BMC performs parallel actions to either reduce the fan speed of the fan group or maintain only minimal cooling for the second CPU, so that the power consumption of the storage server device drops to a preset threshold within a preset time.

[0010] According to some embodiments of the present invention, the step of controlling the dirty data identification unit to read the dirty data bitmap marked by the storage server device and locate the dirty data to be protected in the memory of the storage server device when receiving an indication signal that the power consumption of the storage server device has decreased to a preset threshold includes: When an indication signal is received that the power consumption of the storage server device drops to a preset threshold within a preset time, the dirty data identification unit is controlled to read the dirty data bitmap marked by the distributed storage software in the storage server device. Accurately locate the non-persistent dirty data in the memory of the storage server device, and treat the non-persistent dirty data as dirty data to be protected; Calculate and store the original redundancy value of the dirty data to be protected.

[0011] According to some embodiments of the present invention, a cyclic redundancy check unit is used to verify the integrity of dirty data transferred to the preset storage unit, including: The dirty data after being transferred is checked by the cyclic redundancy check unit to obtain the redundancy check value of the dirty data. The original redundancy value of the dirty data to be protected and the redundancy check value of the dirty data are compared for consistency. When the original redundancy value of the dirty data and the redundancy check value of the dirty data are consistent, it is determined that the integrity check of the transferred dirty data has passed. If the original redundancy value of the dirty data and the redundancy check value of the dirty data are inconsistent, it is determined that the integrity check of the transferred dirty data has failed.

[0012] According to some embodiments of the present invention, it further includes: When the power supply unit is in an abnormal power failure state, the distributed storage software of the storage server device informs the protection status of the cluster synchronization node, and tells other nodes to suspend data writing to the node in order to suspend data interaction. When the power supply unit is detected to be powered on, the dirty data and its redundancy check value in the storage unit are read. If the dirty data transferred to the preset storage unit passes the verification, the dirty data is restored to the memory cache, asynchronously written to the storage pool, and the state is restored to the cluster synchronization node to restore data interaction. At the same time, the BMC restores power to the fan, non-critical hard drive, and PCIe peripheral, wakes up the first CPU and restores the frequency so that the storage server device returns to normal operation. If the verification of dirty data transferred to the preset storage unit fails, the cluster data reconstruction process is triggered to synchronize the dirty data from the replica node to ensure the consistency of the storage pool.

[0013] According to some embodiments of the present invention, it further includes: After the power supply switching control unit switches the power supply mode from the power supply unit to the battery backup unit, it monitors the power level of the battery backup unit in real time. When the battery backup unit's power level falls below a safety threshold, an alarm message is issued; When the battery backup unit's power level is lower than a first preset power level, the battery backup unit is charged. When the battery backup unit reaches the second preset power value, charging of the battery backup unit is stopped.

[0014] This invention also provides a power failure protection system for a storage server device, comprising: The CPLD is used to monitor the output signal of the power supply unit in real time. When the abnormal output signal of the power supply unit is detected, it determines that the power supply unit is in an abnormal power failure state and generates a non-maskable interrupt signal. A power supply switching control unit is used to switch the power supply mode from the power supply unit to the battery backup unit in response to the non-maskable interruption signal. ACPI and the hardware coordination module are used to control the necessary hardware to enter the core shutdown mode and interrupt the power supply to the non-essential hardware. The dirty data identification unit is used to read the dirty data bitmap marked by the storage server device and locate the dirty data to be protected in the memory of the storage server device when it receives an indication signal that the power consumption of the storage server device has dropped to a preset threshold. A cyclic redundancy check unit is used to transfer the dirty data to be protected to a preset storage unit through a PCIe channel, and to check the dirty data transferred to the preset storage unit to verify the integrity of the transferred dirty data. The storage coordination module is used to control the storage server device to perform a shutdown process and control the battery backup unit to stop supplying power when the dirty data verification passes. When a power-on signal is received again, the module reads the dirty data from the preset storage unit and restores it to the memory cache so that the storage server device can be restored to normal operation.

[0015] The present invention also relates to a computer device, including a memory and a processor, wherein the processor performs the above-described method when executing a computer program stored in the memory.

[0016] The present invention also relates to a computer-readable storage medium storing computer program instructions thereon, which, when executed by a processor, implement the above-described method.

[0017] The power-loss protection method, system, and computer device for storage server devices provided in this invention have at least one of the following advantages or beneficial effects: By monitoring the output signal of the power supply unit in real time using a CPLD, abnormal output signals of the power supply unit can be identified immediately, avoiding data loss or system crashes caused by abnormal power loss of the power supply unit. Upon detecting an abnormal power loss of the power supply unit, a non-maskable interrupt signal is generated. The power supply switching control unit can quickly switch the power supply mode to the battery backup unit and, through ACPI and a hardware coordination module, control the necessary hardware to enter a core-shutdown mode, interrupting power supply to non-essential hardware. This coordinated control mechanism not only ensures stable system operation when the power supply unit is abnormal, but also reduces power consumption and extends the power supply time of the battery backup unit.

[0018] When the power consumption of the storage server device drops to a preset threshold, the dirty data identification unit reads and locates dirty data in memory, and redirects the dirty data to a preset storage unit via the PCIe channel. This ensures that dirty data is saved promptly and accurately, preventing data loss. Subsequently, the cyclic redundancy check unit verifies the dirty data redirected to the preset storage unit. Only after successful verification is the storage server device allowed to execute the shutdown procedure, effectively preventing data corruption caused by data transmission errors or storage unit failures, and ensuring data integrity and reliability. This targeted dirty data redirection and recovery mechanism allows operations to be performed only on necessary data in the event of an abnormal power supply failure, avoiding indiscriminate processing of all memory data. This resource optimization mechanism improves system operating efficiency and reduces resource consumption.

[0019] When a power-on signal is received again, dirty data is read from a preset storage unit and restored to the memory cache, enabling the storage server device to quickly return to normal operation. This rapid recovery mechanism reduces system downtime caused by abnormal power supply unit failures, improving system availability and stability.

[0020] Furthermore, additional aspects and advantages of the invention will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of the invention. Attached Figure Description

[0021] Figure 1 This is a general flowchart of the power failure protection method for storage server devices provided in the embodiments of the present invention; Figure 2 This is a detailed flowchart of step S200 in the power failure protection method for storage server devices provided in this embodiment of the invention; Figure 3 This is a detailed flowchart of step S300 in the power failure protection method for storage server devices provided in this embodiment of the invention; Figure 4 This is a detailed flowchart of step S400 in the power failure protection method for storage server devices provided in this embodiment of the invention; Figure 5 This is a detailed flowchart of the first embodiment of the power-loss protection method for storage server devices provided in this invention. Figure 6 This is a detailed flowchart of the second method for power-loss protection of storage server devices provided in this embodiment of the invention; Figure 7 This is a schematic diagram of the power-loss protection system for the storage server device provided in an embodiment of the present invention. Detailed Implementation

[0022] The following will provide a clear and complete description of the concept, specific structure, and technical effects of the present invention in conjunction with the embodiments and accompanying drawings, so as to fully understand the purpose, solution, and effects of the present invention.

[0023] It should be noted that, unless otherwise specified, when a feature is referred to as "fixed" or "connected" to another feature, it can be directly fixed or connected to the other feature, or indirectly fixed or connected to the other feature. The singular forms "a," "described," and "the" used herein are also intended to include the plural forms, unless the context clearly indicates otherwise. Furthermore, unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art. The terminology used in this specification is for the purpose of describing particular embodiments only and not for limiting the invention. The term "and / or" as used herein includes any combination of one or more of the associated listed items.

[0024] It should be understood that although the terms first, second, third, etc., may be used to describe various elements in this disclosure, these elements should not be limited to these terms. These terms are used only to distinguish elements of the same type from one another. For example, a first element may also be referred to as a second element without departing from the scope of this disclosure, and similarly, a second element may also be referred to as a first element. Any and all instances or exemplary language (“e.g.,” “such as,” etc.) provided herein are intended only to better illustrate embodiments of the invention and, unless otherwise required, do not impose a limitation on the scope of the invention.

[0025] This invention provides a power-loss protection method, system, and computer device for storage server devices, which effectively solves the problems mentioned in the background art and can effectively improve the data protection capability and system stability of storage server devices under abnormal power conditions. See details below. Figures 1 to 7 The power failure protection method, system, and computer device for storage server equipment provided in the embodiments of the present invention will be further described.

[0026] Reference Figure 1 As shown, Figure 1 This is a general flowchart of the power-loss protection method for a storage server device provided in this embodiment of the invention. The power-loss protection method for the storage server device includes, but is not limited to, steps S100 to S500. Specifically, S100: Monitors the output signal of the power supply unit in real time through CPLD. When an abnormality is detected in the output signal of the power supply unit, it determines that the power supply unit is in an abnormal power-off state and generates a non-maskable interrupt signal. S200: In response to a non-maskable interrupt signal, the power supply switching control unit controls the power supply mode to switch from the power supply unit to the battery backup unit. Through ACPI and the hardware coordination module, it controls the relevant necessary hardware to enter the core shutdown mode and interrupts the power supply to the non-necessary hardware. S300: When receiving an indication signal that the power consumption of the storage server device has dropped to a preset threshold, the dirty data identification unit is controlled to read the dirty data bitmap marked by the storage server device and locate the dirty data to be protected in the memory of the storage server device; S400: Dirty data is transferred to a preset storage unit via the PCIe channel, and the dirty data transferred to the preset storage unit is verified by the cyclic redundancy check unit to verify the integrity of the transferred dirty data. S500: When the dirty data verification passes, it controls the storage server device to execute the shutdown process and controls the battery backup unit to stop supplying power. When the power-on signal is received again, it reads the dirty data from the preset storage unit and restores it to the memory cache so that the storage server device can be restored to normal operation.

[0027] In some embodiments of the present invention, the power-loss protection method for storage server devices includes: real-time monitoring of the output signal of the power supply unit using a CPLD (Complex Programmable Logic Device), which enables immediate identification of abnormal output signals from the power supply unit, allowing for rapid response to abnormal situations and preventing data loss or system crashes caused by abnormal power loss of the power supply unit. Simultaneously, upon detecting an abnormal power loss of the power supply unit, a non-maskable interrupt signal (NMI) is generated. The use of the non-maskable interrupt signal prevents response failures caused by the accidental masking of ordinary interrupts, further improving system reliability.

[0028] In response to a non-maskable interrupt signal, the power supply switching control unit can quickly switch the power supply mode to the battery backup unit. Through ACPI (Advanced Configuration and Power Interface) and the hardware coordination module, it controls necessary hardware to enter a core shutdown mode, interrupting power supply to non-essential hardware. This coordinated control mechanism not only ensures stable system operation when the power supply unit malfunctions, but also minimizes energy waste, reduces power consumption, extends the battery backup unit's power supply time, and improves energy efficiency.

[0029] When the power consumption of the storage server device drops to a preset threshold, the dirty data identification unit reads and locates dirty data in memory, and then transfers the dirty data to a preset storage unit via the PCIe channel. This process ensures that dirty data can be saved promptly and accurately in the event of a power supply unit failure, preventing data loss. Afterwards, the dirty data transferred to the preset storage unit is verified by a cyclic redundancy check unit to confirm data integrity. Only after successful verification is the storage server device allowed to execute the shutdown procedure. This verification mechanism effectively avoids data corruption caused by data transmission errors or storage unit failures, ensuring data integrity and reliability.

[0030] The targeted transfer and recovery mechanism for dirty data allows operations to be performed on only the necessary data in the event of an abnormal power supply failure, avoiding indiscriminate processing of the entire memory data. This resource optimization mechanism improves system efficiency and reduces resource consumption.

[0031] When a power-on signal is received again, dirty data is read from a preset storage unit and restored to the memory cache, enabling the storage server device to quickly return to normal operation. This rapid recovery mechanism reduces system downtime caused by abnormal power supply unit failures, improving system availability and stability.

[0032] Through various technical means such as real-time monitoring, rapid response, data protection, power switching, hardware collaborative control, power consumption management, and modular design, it achieves high reliability, data protection and integrity, system stability and recovery capability, energy saving and resource optimization, flexibility and scalability, reduced maintenance costs, and improved user experience.

[0033] It is understood that the power failure protection method for storage server devices of the present invention has at least the following advantages over the prior art: More precise and efficient power consumption control, adapting to BBU load capacity: Through CPLD microsecond-level power failure detection, ACPI is linked to control the necessary hardware to enter the core shutdown mode and interrupt the power supply to the non-essential hardware. It can reduce the power consumption from more than 1000W to less than 500W within 100 microseconds, which is fully matched to the load capacity of the battery backup unit (50A peak discharge, 500W+ maximum power consumption).

[0034] Deep hardware-software collaboration forms a complete protection loop: Existing solutions only achieve basic linkage between the CPLD and the battery backup unit, without involving the collaboration between the dedicated ACPI and the hardware collaboration module. This invention innovatively designs the ACPI and hardware collaboration module, including a hardware unloading plugin, a data transfer plugin, and distributed storage software, constructing an end-to-end closed loop of "power failure detection - interrupt triggering - power consumption control - data transfer - status synchronization - power-on recovery". Compared with the existing solution's "single power supply switching + data backup" mode, the protection process is more complete and has stronger adaptability.

[0035] Enhanced data consistency assurance and superior transfer efficiency and verification capabilities: Compared to existing solutions' simple data backup (lacking targeted transfer or verification mechanisms), this invention employs a "dirty data bitmap identification + targeted transfer of storage units + dual verification of redundancy values" mechanism, achieving a transfer rate ≥2GB / s and ensuring no data loss or corruption. Automatic data recovery and cluster state synchronization upon power-on significantly improve data security; cluster data reconstruction is triggered upon verification failure, further guaranteeing storage pool consistency.

[0036] Highly compatible, with low cost increment and easy implementation: The ACPI and hardware collaboration module adopts the PCIe 4.0 / 5.0 standard interface and is adapted to mainstream distributed storage software stacks (Ceph / Gluster) through a data transfer plugin. It can be directly integrated into existing distributed storage nodes without refactoring the hardware architecture. Only the battery backup unit and the ACPI and hardware collaboration module need to be added, with a cost increase of less than 20%, which is significantly more cost-effective than similar redundant design solutions.

[0037] In some embodiments of the present invention, step S100 includes, but is not limited to, steps S110 to S140; specifically, S110: Real-time monitoring of the POK signal, output voltage, and AC-OK signal output by the power supply unit via CPLD; S120: When the POK signal output by the power supply unit is low, it is determined that the output signal of the power supply unit is abnormal and the power supply unit is in an abnormal power-off state. S130: When the output voltage of the power supply unit is lower than the preset value, it is determined that the output signal of the power supply unit is abnormal and the power supply unit is in an abnormal power-off state. S140: When the AC-OK signal output by the power supply unit is lost or is low, it is determined that the output signal of the power supply unit is abnormal and the power supply unit is in an abnormal power-off state.

[0038] In some embodiments of the present invention, when an abnormal output signal of the power supply unit is detected, determining that the power supply unit is in an abnormal power-off state includes: real-time monitoring of the POK signal, output voltage and AC-OK signal output by the power supply unit. By monitoring these three signals simultaneously, the state of the power supply unit can be determined from multiple dimensions, avoiding the risk of misjudgment caused by a single signal.

[0039] When the POK signal output by the power supply unit is low, it indicates that the power supply is not stable or has failed. At this time, it is determined that the output signal of the power supply unit is abnormal and the power supply unit is in an abnormal power-off state.

[0040] The output voltage of the power supply unit is directly monitored to ensure it is within the preset normal range. If the output voltage is lower than the preset value (e.g., 12V), it indicates insufficient power supply capacity or a fault. In this case, the output signal of the power supply unit is determined to be abnormal, and the power supply unit is determined to be in an abnormal power-off state.

[0041] When the AC-OK signal output by the power supply unit is lost or goes low, it indicates that the AC power supply has been interrupted or is in an unstable state. At this time, it is determined that the output signal of the power supply unit is abnormal and the power supply unit is in an abnormal power failure state.

[0042] By setting multiple anomaly detection conditions (POK signal low level, output voltage below a preset value, AC-OK signal loss or low level), the status of the power supply unit can be determined from multiple dimensions, enabling more accurate identification of abnormal states. This multi-condition judgment mechanism reduces the possibility of false positives and improves the reliability of power supply anomaly state identification.

[0043] Reference Figure 2 As shown, Figure 2 This is a detailed flowchart of step S200 in the power-loss protection method for a storage server device provided in this embodiment of the invention. Step S200 includes, but is not limited to, steps S210 to S230. Specifically, S210: Accurately identifies necessary and unnecessary hardware through a preset hardware priority list. Necessary hardware includes the first CPU and the second CPU, while unnecessary hardware includes the memory associated with the first CPU, PCIe peripherals, and non-critical hard drives. S220: By configuring ACPI parameters through the BIOS of the storage server device, the ACPI and hardware coordination module triggers the first CPU to enter the core shutdown mode based on the ACPI parameters, and controls the second CPU to reduce its frequency to the lowest operating frequency; S230: Cuts off power to the first CPU-associated memory, PCIe peripherals, and non-critical hard drives via ACPI and hardware coordination module.

[0044] In some embodiments of the present invention, controlling necessary hardware to enter a core-shutdown mode and interrupting power supply to non-essential hardware via ACPI and a hardware coordination module includes: using a preset hardware priority list to clearly distinguish between necessary hardware (such as the first CPU and the second CPU) and non-essential hardware (such as the first CPU-associated memory, PCIe peripherals, and non-critical hard drives). The PCIe peripherals can be network cards or expansion cards. This classification method provides a foundation for subsequent power management and resource optimization. The hardware priority list can be adjusted and expanded according to different application scenarios and system requirements, enabling the system to adapt to diverse hardware configurations and operating modes.

[0045] By configuring ACPI (Advanced Configuration and Power Interface) parameters in the storage server device's BIOS, the operating status of the hardware can be flexibly controlled according to a preset priority list and power management strategy. Based on the ACPI parameters, the ACPI and hardware coordination module triggers the first CPU to enter a core-shutdown mode and controls the second CPU to reduce its frequency to the lowest possible operating frequency, thereby reducing CPU power consumption. Simultaneously, power to unnecessary hardware is cut off, enabling the system to operate efficiently in a low-power mode. This layered power management strategy minimizes energy consumption while ensuring basic system functionality.

[0046] By cutting off power to non-essential hardware such as the primary CPU-associated memory, PCIe peripherals, and non-critical hard drives, the system's power consumption was further reduced. This refined power management mechanism not only extended the battery backup unit's operating time but also optimized the use of system resources.

[0047] Through the coordinated control of hardware priority lists and ACPI parameters, hardware resources can be accurately identified and managed. This precise control mechanism reduces system instability caused by hardware malfunctions or resource conflicts, thus improving system reliability. The system can operate stably even when the power supply unit malfunctions or the system enters a low-power mode, reducing system crashes or data loss due to improper power management.

[0048] In some embodiments of the present invention, the necessary hardware also includes the core memory of the storage server device, a storage unit, a controller with built-in redundant hard disks and fan groups, and the power-loss protection method of the storage server device further includes, but is not limited to, steps S240 to S250. Specifically, S240: Power is supplied to the second CPU, the core memory of the storage server device, the storage unit, and the redundant hard drive built into the controller via a battery backup unit; S240: By using the BMC of the storage server device to perform parallel execution of fan speed reduction or only retaining minimal heat dissipation for the second CPU, the power consumption of the storage server device is reduced to a preset threshold within a preset time.

[0049] In embodiments of the present invention, when the power supply unit experiences an abnormal power failure, the battery backup unit supplies power to the second CPU, core memory, preset storage unit, and controller-built-in redundant hard drive. This power supply strategy ensures that the system can maintain the operation of critical functions even during a power outage. The choice to power the second CPU instead of the first CPU may be based on the fact that the first CPU does not require additional power in core-deactivated mode, while the second CPU still needs to maintain basic functions after frequency reduction. The core memory and preset storage unit are critical components for data storage, while the controller-built-in redundant hard drive is used for data backup and redundancy protection. Powering these components ensures data integrity and system stability. By setting preset time and power consumption thresholds, the system can execute power-down protection procedures in a planned manner. This preset management mechanism allows the system to orderly enter a low-power mode when the power supply unit malfunctions, reducing the burden on the battery backup unit.

[0050] The storage server's BMC (Browser Control Center) can reduce fan speeds or provide only minimal cooling to the secondary CPU. This thermal management strategy aims to reduce system power consumption, ensure the cooling needs of critical hardware, and simultaneously log power-down events and operations to further reduce unnecessary power consumption. During power-down protection, the BMC executes thermal management operations in parallel to ensure that power consumption is reduced to a preset threshold (e.g., to 500W) within a preset time, achieving a balance between power optimization and thermal management. This balanced strategy considers both power reduction and avoids hardware failures caused by insufficient cooling.

[0051] By powering critical hardware with battery backup units and performing thermal management operations in parallel through the BMC, storage server devices can achieve fine-grained power consumption management and thermal balance during power failure protection. This technology significantly improves system reliability, power management capabilities, recovery capabilities, and user experience, while reducing system maintenance costs. This comprehensive power failure protection method provides strong support for the stable operation of storage server devices under abnormal power conditions.

[0052] Reference Figure 3 As shown, Figure 3 This is a detailed flowchart of step S300 in the power failure protection method for a storage server device provided in this embodiment of the invention. Step S300 includes, but is not limited to, steps S310 to S330. Specifically, S310: When an indication signal is received that the power consumption of the storage server device drops to a preset threshold within a preset time, the dirty data identification unit is controlled to read the dirty data bitmap marked by the distributed storage software in the storage server device. S320: Accurately locates non-persistent dirty data in the memory of the storage server device and treats the non-persistent dirty data as dirty data to be protected; S330: Calculate and store the original redundancy value of the dirty data to be protected.

[0053] In some embodiments of the present invention, when an indication signal indicating that the power consumption of the storage server device has decreased to a preset threshold is received, the dirty data identification unit is controlled to read the dirty data bitmap marked by the storage server device to locate the dirty data to be protected in the memory of the storage server device. This includes: pre-setting a preset time and power consumption threshold to ensure that the system power consumption can be reduced to a safe level within a specified time when the power supply unit malfunctions. This preset mechanism provides time and power consumption guarantees for subsequent data protection operations. When the power consumption of the storage server device decreases to the preset threshold, an indication signal is triggered to start the dirty data identification and protection process. This triggering mechanism ensures the timeliness of data protection operations.

[0054] By reading the dirty data bitmap marked by the distributed storage software in the storage server device, unpersisted dirty data in memory can be accurately located. This marking mechanism makes dirty data identification more efficient and accurate. The dirty data bitmap provides specific location information of dirty data in memory, locating the dirty data to be protected in memory (such as the address range 0x3800000000-0x3c00000000), and thus accurately identifying unpersisted dirty data as the dirty data to be protected. This precise location mechanism reduces redundant operations in the data protection process and improves efficiency.

[0055] Calculate the original redundancy value (e.g., 0xcf8dc8be) of the dirty data to be protected, transfer it, and then re-verify it to ensure the data is undamaged. This redundancy value is used for subsequent data verification and recovery operations. The redundancy value can be calculated using Cyclic Redundancy Check (CRC), checksum, or other data integrity verification algorithms. Store the calculated original redundancy value to provide a basis for subsequent data integrity verification. This storage mechanism ensures that the integrity and accuracy of the data can be verified during data recovery.

[0056] In some embodiments of the present invention, the preset threshold is 500W and the preset time is 100 to 150 microseconds. Those skilled in the art can set it according to the actual situation. The embodiments of the present invention do not limit the preset threshold and preset time.

[0057] By using preset power consumption thresholds and time management, combined with a dirty data bitmap marked by distributed storage software, dirty data in memory can be accurately identified and located, and its original redundancy value can be calculated and stored. This technique significantly improves the reliability of data protection, system efficiency, and recovery capabilities, while reducing maintenance costs and enhancing user experience. This comprehensive data protection method provides strong protection for the data security of storage server devices in the event of power failure.

[0058] Reference Figure 4As shown, Figure 4 This is a detailed flowchart of step S400 in the power failure protection method for a storage server device provided in this embodiment of the invention. Step S400 includes, but is not limited to, steps S410 to S440. Specifically, S410: The dirty data after transfer is checked by the cyclic redundancy check unit to obtain the redundancy check value of the dirty data; S420: Perform a consistency comparison between the original redundant value and the redundant check value of the dirty data to be protected; S430: When the original redundancy value of the dirty data and the redundancy check value of the dirty data are consistent, it is determined that the integrity check of the transferred dirty data has passed; S440: When the original redundancy value of dirty data and the redundancy check value of dirty data are inconsistent, it is determined that the integrity check of the transferred dirty data has failed.

[0059] In some embodiments of the present invention, verifying the integrity of dirty data transferred to a preset storage unit by using a cyclic redundancy check unit includes: after transferring the dirty data, verifying the transferred dirty data using the cyclic redundancy check unit, calculating the redundancy check value of the dirty data, and comparing the redundancy check value of the transferred dirty data with the previously calculated and stored original redundancy value. This comparison mechanism is used to verify whether the data remains intact during the transfer process.

[0060] If the redundancy check value of the transferred dirty data matches the original redundancy value, the integrity check of the transferred dirty data is considered successful. This indicates that no errors or corruption occurred during the data transfer process. If the redundancy check value of the transferred dirty data does not match the original redundancy value, the integrity check of the transferred dirty data is considered unsuccessful. This indicates that errors or corruption may have occurred during the data transfer process, requiring further processing. This comprehensive data integrity check method provides strong assurance for the stability of storage server equipment during data transfer and recovery.

[0061] Reference Figure 5 As shown, Figure 5 This is a detailed flowchart of the first embodiment of the power-loss protection method for a storage server device provided by the present invention. The power-loss protection method for the storage server device also includes, but is not limited to, steps S600 to S630. Specifically, S600: When the power supply unit is in an abnormal power failure state, the distributed storage software of the storage server device sends the protection status of the cluster synchronization node to inform other nodes to suspend data writing to the node in order to suspend data interaction. S610: When the power supply unit is detected to be powered on, read the dirty data and its redundancy check value from the storage unit; S620: If the dirty data transferred to the preset storage unit passes the verification, the dirty data is restored to the memory cache, asynchronously written to the storage pool, and the state is restored to the cluster synchronization node to restore data interaction. At the same time, the BMC restores power to the fan, non-critical hard drive, and PCIe peripheral, wakes up the first CPU and restores the frequency so that the storage server device returns to normal operation. S630: If the verification of dirty data transferred to the preset storage unit fails, the cluster data reconstruction process is triggered to synchronize the dirty data from the replica node to ensure the consistency of the storage pool.

[0062] In some embodiments of the present invention, the power failure protection method for the storage server device further includes: when the power supply unit experiences an abnormal power failure, the distributed storage software of the storage server device synchronizes the node protection status with the cluster, informing other nodes to suspend data write operations to that node. This mechanism ensures that the node will not receive new data writes during the power failure, avoiding data conflicts and inconsistencies. By suspending data interaction, the focus can be placed on protecting dirty data in the current node, ensuring that data can be safely recovered after power is restored.

[0063] Upon power-up of the power supply unit, dirty data and its redundancy check value are read from the storage unit to verify data integrity. The integrity of the dirty data is determined by comparing the transferred redundancy check value with the original redundancy value. This verification mechanism ensures data accuracy during the transfer and recovery processes.

[0064] If the dirty data verification passes, the dirty data is restored to the memory cache and asynchronously written to the storage pool. This asynchronous write mechanism improves the efficiency of data recovery and reduces system waiting time. After data interaction is restored, the state is restored to the cluster synchronization node to ensure that other nodes in the cluster can continue to interact with that node. The BMC restores power to the fans, non-critical hard drives, and PCIe peripherals, and wakes up the first CPU to restore its operating frequency. This gradual restoration mechanism of hardware resources ensures a smooth transition to normal operation after power restoration.

[0065] If dirty data verification fails, the cluster data reconstruction process is triggered, synchronizing the dirty data from the replica nodes to ensure the consistency of the storage pool. This mechanism leverages the redundancy of the cluster to ensure data integrity and consistency under abnormal conditions.

[0066] By synchronizing node protection states during power outages, new data writes can be effectively prevented during abnormal power outages, reducing the risk of data conflicts and inconsistencies. After power is restored, integrity checks and recovery of corrupted data are performed, enabling the storage server to effectively protect data and quickly return to normal operation. This technique significantly improves data integrity and reliability, and enhances the system's recovery capabilities.

[0067] Reference Figure 6 As shown, Figure 6 This is a detailed flowchart of the second embodiment of the power-loss protection method for storage server devices provided by the present invention. The power-loss protection method for storage server devices also includes, but is not limited to, steps S700 to S730. Specifically, S700: After the power supply switching control unit switches the power supply mode from the power supply unit to the battery backup unit, it monitors the battery backup unit's power level in real time. S710: Issues an alarm message when the battery backup unit's power level falls below a safety threshold; S720: When the battery backup unit's power level is lower than the first preset power level, charge the battery backup unit. S730: When the battery backup unit reaches the second preset power value, stop charging the battery backup unit.

[0068] In some embodiments of the present invention, the power failure protection method for the storage server device further includes: when the power supply unit experiences an abnormal power failure, the power supply switching control unit switches the power supply mode from the power supply unit to the battery backup unit. This switching mechanism ensures that the system can continue to operate when the power is interrupted. After switching to the battery backup unit, the battery backup unit's power level is monitored in real time. This monitoring mechanism provides data support for subsequent power management and alarms.

[0069] A pre-set safety threshold, a first preset battery level, and a second preset battery level are used to determine the battery backup unit's power status. When the battery backup unit's power level falls below the safety threshold, an alarm message is issued to alert the administrator or the system to take action. This alarm mechanism ensures timely action is taken when the power is insufficient, preventing system shutdown due to depletion of power. When the battery backup unit's power level falls below the first preset battery level, charging is initiated. This charging mechanism ensures that the battery backup unit can be replenished promptly when its power is low. When the battery backup unit's power level reaches the second preset battery level, charging stops. This mechanism prevents overcharging and extends battery life.

[0070] In some embodiments of the present invention, the security threshold can be set according to the operating load on the storage server device. The embodiments of this application do not limit the size of the security threshold.

[0071] In some embodiments of the present invention, the first preset power value is 80% and the second preset power value is 95%. Those skilled in the art can set the size of the first preset power value and the second power value according to the actual situation. The embodiments of this application do not limit the size of the first preset power value and the second power value.

[0072] The system switches the power supply to the battery backup unit via a power switching control unit and monitors the battery backup unit's power level in real time. This allows the system to continue operating even if the main power supply unit experiences a power failure, and a power management mechanism ensures the battery backup unit's power level remains within a reasonable range. This technology significantly improves system reliability, availability, and maintenance efficiency, while also enhancing the user experience.

[0073] In some embodiments of the present invention, the power-loss protection method for the storage server device further includes: presetting the following parameters in the BIOS of the storage server device to ensure that the power-loss protection function is adaptable to different scenarios: Enable battery backup unit function: Controls whether the battery backup unit participates in power failure protection; Set the dump data space size of the preset storage unit: supports options such as 8G / 16G / 32G / 64G, and the 64G option is compatible with two controller built-in redundant hard drives; Enable the "Back up data on shutdown" function: Ensure that the data transfer is complete before shutting down; Power backup enhancement function: Optimize the discharge curve of the battery backup unit to adapt to changes in the power consumption of storage devices; Forced data recovery option: Forcefully restores the data in the preset storage unit to memory upon power-up to ensure data integrity.

[0074] Add battery backup unit configuration options to the BIOS, including: enabling / disabling battery backup unit function, selecting the size of the data dump space, enabling / disabling data backup during power-off, and switching on / off the power backup enhancement function. Configuration parameters are tied to the number of redundant hard drives built into the controller (e.g., a 64GB NU only supports 2 redundant hard drives), ensuring sufficient data dump space during power failure and adapting to different node configuration requirements.

[0075] This invention also provides a power failure protection system for a storage server device, as described in the embodiments of the present invention. Figure 7As shown, the power failure protection system of the storage server device includes a CPLD, a power supply switching control unit, an ACPI and hardware coordination module, a dirty data identification unit, a cyclic redundancy check unit, and a storage coordination module. The CPLD monitors the output signal of the power supply unit in real time. When it detects an abnormal output signal, it determines that the power supply unit is in an abnormal power failure state and generates a non-maskable interrupt signal. The power supply switching control unit responds to the non-maskable interrupt signal and controls the power supply mode to switch from the power supply unit to the battery backup unit. The ACPI and hardware coordination module controls necessary hardware to enter a core-shutdown mode and interrupts power supply to non-essential hardware. The dirty data identification unit is used for... When an indication signal indicating that the power consumption of the storage server device has decreased to a preset threshold is received, the dirty data bitmap marked by the storage server device is read to locate the dirty data to be protected in the memory of the storage server device; the cyclic redundancy check unit is used to transfer the dirty data to be protected to a preset storage unit through the PCIe channel, and to verify the integrity of the transferred dirty data in the preset storage unit; the storage coordination module is used to control the storage server device to perform a shutdown process and control the battery backup unit to stop power supply when the dirty data verification passes. When a power-on signal is received again, the dirty data is read from the preset storage unit and restored to the memory cache so that the storage server device returns to normal operation.

[0076] The power-loss protection system of the storage server equipment achieves a comprehensive improvement in high reliability, data protection, power consumption management, rapid recovery, and user experience through the coordinated operation of the CPLD, power switching control unit, ACPI and hardware coordination module, dirty data identification unit, cyclic redundancy check unit, and storage coordination module. This modular design not only improves the overall performance of the system but also facilitates system expansion and maintenance.

[0077] It should be understood that the method steps in the embodiments of the present invention can be implemented or carried out by computer hardware, a combination of hardware and software, or by computer instructions stored in a non-transitory computer-readable storage medium. The method can use standard programming techniques. Each program can be implemented in a high-level procedural or object-oriented programming language to communicate with the computer system. However, if necessary, the program can be implemented in assembly or machine language. In any case, the language can be a compiled or interpreted language. Furthermore, for this purpose, the program can run on a programmed application-specific integrated circuit (ASIC).

[0078] Furthermore, the procedures described herein may be performed in any suitable order unless otherwise indicated herein or otherwise clearly contradicted by the context. The procedures described herein (or variations and / or combinations thereof) may be executed under the control of one or more computer systems configured with executable instructions, and may be implemented by hardware or a combination thereof as code (e.g., executable instructions, one or more computer programs, or one or more applications) that commonly executes on one or more processors. The computer program comprises a plurality of instructions executable by one or more processors.

[0079] Furthermore, the method can be implemented in any suitable type of computing platform, including but not limited to personal computers, minicomputers, mainframes, workstations, networked or distributed computing environments, standalone or integrated computer platforms, or in communication with charged particle tools or other imaging devices, etc. Aspects of the invention can be implemented as machine-readable code stored on a non-transitory storage medium or device, whether removable or integrated into a computing platform, such as a hard disk, optical read and / or write storage medium, RAM, ROM, etc., such that it is readable by a programmable computer, and when the storage medium or device is read by the computer, it can be used to configure and operate the computer to perform the processes described herein. Furthermore, the machine-readable code, or portions thereof, can be transmitted via wired or wireless networks. The invention described herein includes these and other different types of non-transitory computer-readable storage media when such media comprises instructions or programs that implement the steps described above in conjunction with a microprocessor or other data processor. When programmed according to the methods and techniques described in the invention, the invention may also include the computer itself.

[0080] A computer program can be applied to input data to perform the functions described herein, thereby transforming the input data to generate output data stored in non-volatile memory. The output information can also be applied to one or more output devices, such as a display. In a preferred embodiment of the invention, the transformed data represents physical and tangible objects, including specific visual depictions of physical and tangible objects generated on the display.

[0081] The above description is merely a preferred embodiment of the present invention. The present invention is not limited to the above-described embodiments. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention, as long as they achieve the technical effects of the present invention by the same means, should be included within the scope of protection of the present invention. Within the scope of protection of the present invention, the technical solutions and / or implementation methods can have various modifications and variations.

Claims

1. A power-loss protection method for storage server equipment, characterized in that, include: The output signal of the power supply unit is monitored in real time by CPLD. When the abnormal output signal of the power supply unit is detected, it is determined that the power supply unit is in an abnormal power failure state and an unmaskable interrupt signal is generated. In response to the non-maskable interrupt signal, the power supply switching control unit controls the power supply mode to switch from the power supply unit to the battery backup unit, and controls the relevant necessary hardware to enter the core shutdown mode through ACPI and hardware coordination module, and interrupts the power supply to the non-necessary hardware. When an indication signal is received that the power consumption of the storage server device has decreased to a preset threshold, the dirty data identification unit is controlled to read the dirty data bitmap marked by the storage server device and locate the dirty data to be protected in the memory of the storage server device. The dirty data is transferred to a preset storage unit via the PCIe channel, and the dirty data transferred to the preset storage unit is verified by the cyclic redundancy check unit to verify the integrity of the transferred dirty data. When the dirty data verification passes, the storage server device is controlled to execute the shutdown process, and the battery backup unit is controlled to stop supplying power. When the power-on signal is received again, the dirty data is read from the preset storage unit and restored to the memory cache, so that the storage server device is restored to normal operation.

2. The power failure protection method for storage server equipment according to claim 1, characterized in that, The step of monitoring the output signal of the power supply unit in real time via CPLD, and determining that the power supply unit is in an abnormal power-off state when an abnormal output signal of the power supply unit is detected, includes: The POK signal, output voltage, and AC-OK signal output by the power supply unit are monitored in real time using a CPLD. When the POK signal output by the power supply unit is low, it is determined that the output signal of the power supply unit is abnormal, and the power supply unit is in an abnormal power-off state. When the output voltage of the power supply unit is lower than the preset value, it is determined that the output signal of the power supply unit is abnormal, and the power supply unit is in an abnormal power failure state. When the AC-OK signal output by the power supply unit is lost or is low, it is determined that the output signal of the power supply unit is abnormal, and the power supply unit is in an abnormal power-off state.

3. The power failure protection method for storage server equipment according to claim 1, characterized in that, The step of controlling necessary hardware to enter a core-shutdown mode and interrupting power supply to non-essential hardware via ACPI and hardware coordination module includes: The relevant necessary hardware and non-necessary hardware are accurately identified by a preset hardware priority list. The relevant necessary hardware includes a first CPU and a second CPU, and the non-necessary hardware includes memory associated with the first CPU, PCIe peripherals, and non-critical hard drives. By configuring ACPI parameters in the BIOS of the storage server device, the ACPI and hardware coordination module triggers the first CPU to enter the core shutdown mode based on the ACPI parameters, and controls the second CPU to reduce its frequency to the lowest operating frequency. The ACPI and hardware coordination module disconnects the power supply to the first CPU-associated memory, the PCIe peripherals, and the non-critical hard drive.

4. The power failure protection method for the storage server device according to claim 3, characterized in that, The necessary hardware also includes the core memory of the storage server device, storage units, a controller with built-in redundant hard drives and fan groups, and the method further includes: The battery backup unit provides power to the second CPU, the core memory of the storage server device, the storage unit, and the redundant hard drive built into the controller. The storage server device's BMC performs parallel actions to either reduce the fan speed of the fan group or maintain only minimal cooling for the second CPU, so that the power consumption of the storage server device drops to a preset threshold within a preset time.

5. The power-loss protection method for a storage server device according to claim 1, characterized in that, When an indication signal indicating that the power consumption of the storage server device has decreased to a preset threshold is received, the dirty data identification unit is controlled to read the dirty data bitmap marked by the storage server device and locate the dirty data to be protected in the memory of the storage server device, including: When an indication signal is received that the power consumption of the storage server device drops to a preset threshold within a preset time, the dirty data identification unit is controlled to read the dirty data bitmap marked by the distributed storage software in the storage server device. Accurately locate the non-persistent dirty data in the memory of the storage server device, and treat the non-persistent dirty data as dirty data to be protected; Calculate and store the original redundancy value of the dirty data to be protected.

6. The power-loss protection method for a storage server device according to claim 5, characterized in that, The dirty data transferred to the preset storage unit is verified by a cyclic redundancy check unit to verify the integrity of the transferred dirty data, including: The dirty data after being transferred is checked by the cyclic redundancy check unit to obtain the redundancy check value of the dirty data. The original redundancy value of the dirty data to be protected and the redundancy check value of the dirty data are compared for consistency. When the original redundancy value of the dirty data and the redundancy check value of the dirty data are consistent, it is determined that the integrity check of the transferred dirty data has passed. If the original redundancy value of the dirty data and the redundancy check value of the dirty data are inconsistent, it is determined that the integrity check of the transferred dirty data has failed.

7. The power-loss protection method for a storage server device according to claim 6, characterized in that, Also includes: When the power supply unit is in an abnormal power failure state, the distributed storage software of the storage server device informs the protection status of the cluster synchronization node, and tells other nodes to suspend data writing to the node in order to suspend data interaction. When the power supply unit is detected to be powered on, the dirty data and its redundancy check value in the storage unit are read. If the dirty data transferred to the preset storage unit passes the verification, the dirty data is restored to the memory cache, asynchronously written to the storage pool, and the state is restored to the cluster synchronization node to restore data interaction. At the same time, the BMC restores power to the fan, non-critical hard drive, and PCIe peripheral, wakes up the first CPU and restores the frequency so that the storage server device returns to normal operation. If the verification of dirty data transferred to the preset storage unit fails, the cluster data reconstruction process is triggered to synchronize the dirty data from the replica node to ensure the consistency of the storage pool.

8. The power-loss protection method for a storage server device according to claim 1, characterized in that, Also includes: After the power supply switching control unit switches the power supply mode from the power supply unit to the battery backup unit, it monitors the power level of the battery backup unit in real time. When the battery backup unit's power level falls below a safety threshold, an alarm message is issued; When the battery backup unit's power level is lower than a first preset power level, the battery backup unit is charged. When the battery backup unit reaches the second preset power value, charging of the battery backup unit is stopped.

9. A power-loss protection system for a storage server device, used to execute the power-loss protection method for a storage server device as described in any one of claims 1 to 8, characterized in that, include: The CPLD is used to monitor the output signal of the power supply unit in real time. When the abnormal output signal of the power supply unit is detected, it determines that the power supply unit is in an abnormal power failure state and generates a non-maskable interrupt signal. A power supply switching control unit is used to switch the power supply mode from the power supply unit to the battery backup unit in response to the non-maskable interruption signal. ACPI and the hardware coordination module are used to control the necessary hardware to enter the core shutdown mode and interrupt the power supply to the non-essential hardware. The dirty data identification unit is used to read the dirty data bitmap marked by the storage server device and locate the dirty data to be protected in the memory of the storage server device when it receives an indication signal that the power consumption of the storage server device has dropped to a preset threshold. A cyclic redundancy check unit is used to transfer the dirty data to be protected to a preset storage unit through a PCIe channel, and to check the dirty data transferred to the preset storage unit to verify the integrity of the transferred dirty data. The storage coordination module is used to control the storage server device to perform a shutdown process and control the battery backup unit to stop supplying power when the dirty data verification passes. When a power-on signal is received again, the module reads the dirty data from the preset storage unit and restores it to the memory cache so that the storage server device can be restored to normal operation.

10. A computer device comprising a memory and a processor, characterized in that, When the processor executes a computer program stored in the memory, it performs the method as described in any one of claims 1 to 8.

Images