A storage system upgrade method, device, storage medium, and program product

Abstract

The application discloses a storage system upgrading method and device, a storage medium and a program product, relates to the technical field of system upgrading, and comprises the following steps: collecting the physical characteristics of a target device which stores an image file, and obtaining the digital signature of the target device for verification when the target characteristics in a preset characteristic library and the physical characteristics are successfully matched; if the verification is successful, a preset upgrading engine based on a target directed acyclic graph is used to determine a system upgrading strategy based on the hardware state of the storage system; wherein the target directed acyclic graph records each upgrading step and the relationship between different upgrading steps; and the image file is used to upgrade the target node in the storage system based on the system upgrading strategy. After the target device is successfully identified, the image file is automatically verified, the safety of the system upgrading process is further improved, and the upgrading path is dynamically constructed through the directed acyclic graph according to the current hardware state of the target device, so that the operation efficiency of the system upgrading is effectively improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of system upgrading, in particular to a storage system upgrading method, device, storage medium and program product. BACKGROUND

[0002] A modern storage system is usually composed of a storage medium layer, a controller hardware layer and a system software layer. Among them, the system software layer directly determines the performance and reliability of the storage system due to its stability and functionality. With the continuous change of business needs and the development of technology, the storage system software needs to be updated regularly and upgraded in function. However, when upgrading the storage system software at present, the operation process is complex, and the unified upgrade operation leads to poor software upgrade effect for different manufacturers' devices. SUMMARY

[0003] The present application provides a storage system upgrading method, device, storage medium and program product, which can eliminate the manual operation links such as image making and startup configuration in the traditional scheme, greatly reduce the operation and maintenance complexity, and effectively improve the operation efficiency of system upgrading and the stability of the upgrading process.

[0004] The present application provides a storage system upgrading method, comprising:

[0005] Collecting the physical features of the target device, and matching the target features in the preset feature library with the physical features; the target device stores an image file for upgrading the storage system;

[0006] When the target features and the physical features are successfully matched, obtaining a first digital signature of the target device, and verifying the first digital signature by using a local preset public key; the first digital signature is a signature generated by the target device based on the physical features by using a preset private key;

[0007] If the target device is verified successfully, obtaining the hardware state of the storage system, and determining the system upgrading strategy of the storage system based on the hardware state by using a preset upgrading engine; the preset upgrading engine is an engine constructed based on a target directed acyclic graph, and the target directed acyclic graph is used to record each upgrading step and the relationship between different upgrading steps;

[0008] Reading the image file in the target device, and upgrading the system of the target node in the storage system by using the image file based on the system upgrading strategy.

[0009] The present application also provides an electronic device, comprising: a memory for storing a computer program; and a processor for executing the computer program to implement the steps of any one of the above storage system upgrading methods.

[0010] The application further provides a computer readable storage medium, and the computer readable storage medium stores a computer program.

[0011] The application further provides a computer program product, which comprises a computer program.

[0012] The application can collect the physical features of the target device that stores the image file for upgrading the storage system, match the target features in the preset feature library with the physical features, acquire the first digital signature of the target device generated by using the preset private key when the target features and the physical features are successfully matched, and verify the first digital signature by using the preset public key locally; if the target device is successfully verified, acquire the hardware state of the storage system, and determine the system upgrade strategy of the storage system based on the hardware state by using the preset upgrade engine constructed based on the target directed acyclic graph; the target directed acyclic graph is used to record each upgrade step and the relationship between different upgrade steps; then read the image file in the target device, and perform system upgrade of the target node in the storage system by using the image file based on the system upgrade strategy.

[0013] Through the above technical solution, the application can identify the device that stores the image file, automatically verify the image file after the target device is successfully identified, further improve the security of the system upgrade process, then collect the current hardware state of the target device in real time, dynamically construct the upgrade path of the storage system by using the directed acyclic graph constructed in advance according to each upgrade step and the relationship between different upgrade steps, so that in the storage system upgrade scenario, the relationship between different upgrade steps is determined, the upgrade path is dynamically constructed according to the real-time monitored hardware state, the corresponding upgrade strategy can be set for different devices, the stability and reliability of the upgrade process are effectively guaranteed, the manual operation links such as image making and startup configuration in the traditional scheme are eliminated, the operation and maintenance complexity is greatly reduced, and the operation efficiency of the system upgrade is effectively improved. BRIEF DESCRIPTION OF DRAWINGS

[0014] In order to more clearly illustrate the embodiments of the application, the drawings needed in the embodiments will be briefly introduced. Obviously, the drawings in the following description are only some embodiments of the application, and other drawings can be obtained by those skilled in the art without creative labor.

[0015] Figure 1 A storage system upgrade method flowchart is provided for the embodiments of the application.

[0016] Figure 2A storage system upgrade system architecture diagram provided for an embodiment of the present application;

[0017] Figure 3 A quantum federation device identification system architecture diagram provided for an embodiment of the present application;

[0018] Figure 4 A storage system upgrade flowchart based on a directed acyclic graph upgrade engine provided for an embodiment of the present application;

[0019] Figure 5 A storage system single-node upgrade flowchart provided for an embodiment of the present application;

[0020] Figure 6 A storage system cluster upgrade method flowchart provided for an embodiment of the present application;

[0021] Figure 7 A zero-trust cluster collaboration framework schematic diagram provided for an embodiment of the present application;

[0022] Figure 8 A storage system cluster upgrade flowchart provided for an embodiment of the present application;

[0023] Figure 9 A storage system upgrade device structure schematic diagram provided for an embodiment of the present application. DETAILED DESCRIPTION

[0024] The technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only some of the embodiments of the present application, but not all the embodiments of the present application. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative work fall within the protection scope of the present application.

[0025] It should be noted that, in the description of the present application, the terms “include”, “contain” or any other variant thereof are intended to cover non-exclusive inclusion, so that the process, method, article or device including a series of elements not only includes those elements, but also includes other elements not explicitly listed, or includes elements inherent to such process, method, article or device. The terms “first”, “second” and the like in the present application are used to distinguish similar objects, and are not used to describe a specific order or sequence.

[0026] With the continuous change of business needs and the development of technology, the storage system software needs to be updated and upgraded regularly. At present, when the storage system software is upgraded, the operation process is complex, and the unified upgrade operation leads to poor software upgrade effect for different manufacturers' equipment. The application can identify the equipment of the storage image file, and after the target equipment is successfully identified, the current hardware state of the target equipment is collected in real time. The directed acyclic graph is constructed in advance according to each upgrade step and the relationship between different upgrade steps, the upgrade path of the storage system is dynamically constructed, the stability and reliability of the upgrade process are effectively guaranteed, and the operation efficiency of the system upgrade is improved.

[0027] In order to enable the person skilled in the art to better understand the application scheme, the application will be further described in detail below in combination with the drawings and specific embodiments.

[0028] Next, the embodiment will be described in detail in combination with the execution flow of the storage system upgrade method, as shown in Figure 1 The embodiment of the application provides a storage system upgrade method, which comprises the following steps:

[0029] In step S11, the physical characteristics of the target equipment are collected, and the target characteristics in the preset characteristic library are matched with the physical characteristics. The image file used for upgrading the storage system is saved in the target equipment.

[0030] It can be understood that, as shown in Figure 2 The storage system upgrade architecture in the embodiment mainly comprises: a device access layer: a quantum encryption U disk is accessed through a USB3.0 (Universal Serial Bus) interface, and a device characteristic sensor collects a physical fingerprint; an edge processing layer: a federal learning engine realizes dynamic identification of the equipment, NTRU (Number Theory Research Unit, a kind of public key cryptography system based on lattice) encryption + quantum characteristic extraction guarantees data security, and a DAG (Directed Acyclic Graph) upgrade engine controls the execution flow; a distributed coordination layer: a zero trust blockchain establishes node consensus, and a Merkle tree verifies the integrity of the upgrade instruction; a physical storage layer: a multi-type storage node (SSD (Solid State Drive) + SCM (Storage Class Memory) / HDD (Hard Disk Drive) + Optane / NVMe (Nonvolatile memory express) cluster).

[0031] Specifically, as shown in Figure 3After the target USB device is accessed, the device access layer can first collect the physical characteristics of the target device and match the target characteristics in the preset characteristic library with the physical characteristics. The target device stores an image file for upgrading the storage system, including but not limited to a quantum encryption U disk, i.e., a USB flash disk using quantum encryption technology, which can generate a quantum key and encrypt stored data. During the matching process of the target characteristics and the physical characteristics, when the matching fails, the feature vector of the physical characteristics can be extracted by using the target device identification model, the configuration parameters of the target device are determined, and the target characteristics corresponding to the target device are generated based on the feature vector and the configuration parameters, so as to save the target characteristics corresponding to the target device to the preset characteristic library. Through this process, the edge processing layer can realize dynamic device recognition by using the federated learning engine, solve the USB device compatibility problem in storage device upgrading, and realize automatic recognition and adaptation of multi-brand devices. In the data transmission from the device to the edge layer, the feature data can be transmitted by quantum encryption, and the cross-layer encrypted communication can be realized by the QKD (Quantum Key Distribution) channel of the quantum key management center.

[0032] In a specific embodiment, when collecting physical characteristics, 12-dimensional sensor data can be collected in real time through a USB3.0 interface, and quantum normalization processing can be performed: . Wherein x represents the original collected physical characteristic data, which is the device physical characteristic parameter (such as voltage fluctuation value, timing signal frequency, temperature change gradient, etc.) collected by a quantum sensor array, and is an original signal without processing; represents the mean value of the characteristic data, which is the overall average level of a certain type of characteristic data (such as the average value of the voltage fluctuation collected multiple times), and is used to eliminate the offset in the data and avoid affecting subsequent model training due to the difference in the basic value; represents the standard deviation of the characteristic data, which makes different characteristics (such as voltage and frequency) have the same order of magnitude, and avoids affecting model learning due to the large range of values of a certain type of characteristic; and v represents the normalized characteristic value, i.e., the standardized data finally output, which helps to enhance the stability of the data in quantum encryption transmission. In this way, the original physical characteristic data is converted into standardized data of uniform scale and low noise, providing data input for subsequent device recognition and federated learning, and ensuring that different devices and different types of characteristic data can be uniformly processed by the system. The federated learning processing process is as follows:

[0033] def process_unknown_device(usb_device):

[0034] # 1. Extract physical characteristics

[0035] features = extract_features(usb_device.sensors);

[0036] # 2. Local model inference

[0037] model_output = mobilenet_v3(features);

[0038] # 3. Gradient encryption (Paillier algorithm)

[0039] encrypted_grad = paillier.encrypt(model_output.grad, pub_key);

[0040] # 4. Secure transmission

[0041] send_to_server(encrypted_grad, qkd_channel)。

[0042] Correspondingly, before the feature vector of the physical feature is extracted by using the target device identification model, a preset device identification model can also be deployed at the edge computing layer of the storage system, and the preset device identification model is trained by using the physical feature to obtain an initial device identification model, then the optimization parameters in the initial device identification model are extracted, and the optimization parameters are encrypted by using the preset quantum key, then the encrypted optimization parameters are uploaded to the central server, and an updated device identification model is obtained from the central server, the initial device identification model is covered by using the updated device identification model to obtain the target device identification model. Wherein, the updated device identification model is a model obtained by updating the preset global model based on the target parameters, the target parameters are obtained by the central server based on the federated aggregation algorithm and the encrypted optimization parameters.

[0043] Specifically, when a USB device is connected to the storage system, the system initiates a multi-stage identification process. First, a preliminary feature comparison is performed. The system matches the collected physical characteristics of the device (voltage fluctuations, timing signals, etc.) with a pre-set feature library. If a complete match exists in the feature library, the corresponding device configuration template is directly invoked. If no complete match exists, the device is treated as an unknown device. Specifically, if no matching record exists in the feature library, the system activates a federated learning process: first, a lightweight identification model is deployed on edge computing nodes, and the model is trained using local device feature data. Then, the model parameters are encrypted and uploaded to the central server, allowing the central server to aggregate parameters from multiple nodes, update the global model, and distribute the optimized model to each node. After a new device is identified, the system automatically stores its feature vector and configuration information in the feature library, updating the feature library and achieving a learning-application closed loop. This enables fully automated upgrade operations. The quantum federated device identification system automatically adapts to hardware configurations, and the directed acyclic graph engine intelligently plans upgrade paths, eliminating manual operations such as image creation and startup configuration in traditional solutions, significantly improving operational efficiency. Furthermore, the federated learning mechanism dynamically constructs device feature models, enabling the system to adaptively identify unknown devices, solving the problem of differences in startup methods between devices from different manufacturers, supporting seamless upgrades of heterogeneous device clusters, and effectively enhancing system compatibility.

[0044] Step S12: When the target feature and physical feature are successfully matched, obtain the first digital signature of the target device and verify the first digital signature using the local preset public key; the first digital signature is a signature generated by the target device based on the physical feature using the preset private key.

[0045] In this embodiment, as Figure 3 As shown, when the target feature and physical feature successfully match, the first digital signature of the target device can be obtained, and the first digital signature can be verified using a local preset public key. The first digital signature is a signature generated by the target device based on the physical feature using a preset private key. Specifically, when obtaining the first digital signature of the target device, the hardware feature code of the target device can be obtained, and the target feature code and hardware feature code in the preset device license list can be matched. Therefore, when the target feature code and hardware feature code successfully match, the first digital signature of the target device can be obtained for verification.

[0046] That is, the device legitimacy verification in the embodiment adopts a hierarchical authentication mechanism: in the primary verification, the device provides a hardware feature code, and the system checks whether the feature code is in the permitted list. After the primary verification is completed, deep verification is performed, and the key device can be enabled for NTRU quantum signature verification. The process includes: the device uses a private key to generate a digital signature for the feature data, and the system receives the digital signature, decrypts and verifies the signature authenticity through the pre-stored public key, and generates a quantum random session key after the verification is passed, so as to ensure the security of the system upgrade based on the quantum random session key in the subsequent upgrade process. For example: the communication content between the encryption device and the system, when the quantum encryption U disk passes the legitimacy verification, the quantum random session key generated by the system is used for subsequent interaction data between the encryption device and the storage system (single node or cluster node), including the transmission of the upgrade image, the device configuration instruction, the state feedback information, etc. In this way, since the quantum random key is generated based on the quantum mechanics principle, it has randomness and is resistant to quantum computing cracking, which can effectively prevent the data from being eavesdropped, tampered or forged in the transmission process, and ensure the security of the communication link. And a temporary secure session channel can also be established through the quantum random session key. Since the key is only valid in this upgrade session, it is automatically disabled after the session ends, avoiding the risk of leakage caused by long-term use of fixed keys, and further strengthening the dynamic security of the upgrade process. Through the above technical solution, whether it is the interaction between the device and the local system in the single node upgrade, or the collaborative instruction transmission between the nodes in the cluster upgrade, the quantum random session key can provide independent encryption protection for each temporary session, ensuring the end-to-end security of the entire upgrade process. And the quantum encryption U disk provides a physically unclonable function to realize the verifiable security of the upgrade process.

[0047] Correspondingly, in the system upgrade process in the embodiment, the first digital signature of the target device can also be reacquired based on a preset time period, and the first digital signature is verified by using the local preset public key; if the target device verification is successful, the system upgrade of the target node in the storage system using the image file is continued; if the target device verification fails, the system upgrade of the target node in the storage system using the image file is stopped. In this way, by continuously verifying the digital signature during the system upgrade, the device certificate can be periodically rechecked to prevent man-in-the-middle attacks and further ensure the system security.

[0048] Step S13, if the target device verification is successful, the hardware state of the storage system is acquired, and the system upgrade strategy of the storage system is determined based on the hardware state by using a preset upgrade engine. The preset upgrade engine is an engine based on a target directed acyclic graph, and the target directed acyclic graph is used to record each upgrade step and the relationship between different upgrade steps.

[0049] In the embodiment, as Figure 4As shown, if the target device verification is successful, the hardware state of the storage system is acquired, and a preset upgrade engine is used to determine the system upgrade strategy of the storage system based on the hardware state. The preset upgrade engine is an engine based on a target directed acyclic graph, which is used to record each upgrade step and the relationship between different upgrade steps. Accordingly, before determining the system upgrade strategy of the storage system based on the hardware state by using the preset upgrade engine, the mirror file writing step and the mirror file writing condition of the storage system can be determined, and the writing operation corresponding to different hardware states of the storage system can be determined, and then the target directed acyclic graph is constructed according to the mirror file writing step, the mirror file writing condition and the writing operation corresponding to different hardware states, so as to construct the preset upgrade engine based on the target directed acyclic graph, that is, the directed acyclic graph upgrade engine (DAG-UE, Directed Acyclic Graph Upgrade Engine). When constructing the node dependency relationship in the directed acyclic graph upgrade engine, the upgrade flow can be modeled as an ordered execution unit module, which specifically includes: a basic node containing necessary operations such as image verification, data writing, boot configuration; a conditional node, a special operation dynamically inserted according to the environment state, for example, automatically adding a cooling node when the temperature exceeds 75℃, activating a sharding transmission node when the bandwidth is lower than 10Mbps. And the dependency relationship of the directed acyclic graph includes: sequential dependency, such as data writing must be executed after the image verification is passed; parallel dependency, such as multiple disks can simultaneously perform data writing; conditional dependency, such as forced insertion of cooling operation in high temperature environment. Through the above directed acyclic graph, the environment perception and system upgrade path adjustment of the system can be realized. The engine first collects parameters, continuously acquires 8 types of environment data such as temperature, voltage and bandwidth, and performs state evaluation according to the collected data, such as determining the environment state level according to the preset threshold, so as to realize dynamic optimization of the upgrade path through real-time monitoring. In some specific embodiments, path reconstruction can be performed, including: the system is in a normal state, and a standard upgrade sequence is executed; the system is in a high temperature state, and a cooling stage is inserted before the heat sensitive operation; the system is in a low bandwidth state, and large data transmission is replaced by a sharding transmission mode. It can be understood that when the environment parameter exceeds the safe range, the upgrade is automatically suspended to realize abnormal interruption of system upgrade.

[0050] In step S14, the mirror file in the target device is read, and the system upgrade of the target node in the storage system is performed by using the mirror file based on the system upgrade strategy.

[0051] In the embodiment, the image file in the target device can be read, and the system upgrade of the target node in the storage system is performed based on the system upgrade strategy and the image file. In the system upgrade process, in one specific embodiment, the system state parameters of the storage system can be collected based on a preset parameter collection period, and it is determined whether the system state parameters meet the preset system upgrade condition. If the system state parameters do not meet the preset system upgrade condition, the system upgrade strategy is adjusted based on the system state parameters by using a preset upgrade engine, and the system upgrade of the target node in the storage system is performed based on the adjusted system upgrade strategy and the image file. In another specific embodiment, the file data amount of the image file written into the storage system can be determined. If the file data amount is equal to a preset data write threshold, the image file written into the storage system is subjected to cyclic redundancy check (CRC). If the image file written into the storage system passes the CRC, and the image file has been completely written into the storage system, the hash value of the image file written into the storage system is determined by using a preset encryption hash function, and the hash value is subjected to check, so that after the hash value passes the check, it is determined that the image file writing is completed. That is, in the image file writing process, data check can be performed in real time, for example, CRC check is performed after 512 KB data is written, and SHA-256 verification of the whole image is performed after the writing is completed, so as to further ensure the reliability and security of the image file writing process.

[0052] And as Figure 4 shown, after the system upgrade of the target node in the storage system is performed based on the system upgrade strategy and the image file, the hot restart operation of the storage system can be performed, and after the hot restart operation of the storage system is completed, the upgraded storage system is detected based on a preset system detection process. If the upgraded storage system passes the detection, the corresponding system upgrade log is generated and recorded. If the upgraded storage system fails the detection, the storage system is rolled back.

[0053] Through the above technical solution, this embodiment can collect the physical characteristics of the target device containing the image file for upgrading the storage system, and match the target features in the preset feature library with the physical features. When the target features and physical features match successfully, the first digital signature generated by the target device using a preset private key is obtained, and the first digital signature is verified using a local preset public key. If the target device is successfully verified, the hardware status of the storage system is obtained, and a preset upgrade engine based on the target directed acyclic graph is used to determine the system upgrade strategy of the storage system based on the hardware status. Then, the image file in the target device is read, and the system upgrade of the target node in the storage system is performed using the image file based on the system upgrade strategy. In this way, quantum encryption and federated learning technologies can be integrated for intelligent device identification, achieving adaptive identification and configuration of device types. After the target device is successfully identified, the image file is automatically verified, further improving the security of the system upgrade process. Then, the current hardware status of the target device is collected in real time, and a dynamic upgrade control engine is built using a directed acyclic graph. This engine supports real-time adjustment of the path based on the environment. In storage system upgrade scenarios, by determining the relationship between different upgrade steps and dynamically constructing upgrade paths based on the real-time monitored hardware status, corresponding upgrade strategies can be set for different devices. This effectively ensures the stability and reliability of the upgrade process and eliminates manual operations such as image creation and startup configuration in traditional solutions. This significantly reduces the complexity of operation and maintenance and effectively improves the efficiency of system upgrade operations.

[0054] Based on the previous embodiment, such as Figure 5 As shown, this embodiment provides a specific implementation of upgrading a storage system with a single node, including:

[0055] When the quantum-encrypted USB flash drive is inserted into the storage device's USB port, the system automatically activates the device identification process: the quantum sensor array starts up and collects the device's physical characteristic parameters, including voltage fluctuation curves, timing signal frequencies, and temperature change gradients; the identification engine matches the collected feature data with a pre-set feature library. If a complete match is found, the corresponding device configuration template is directly loaded; if no match is found, the federated learning process is started. After the federated learning process is completed, the device's legitimacy is verified using the NTRU quantum signature algorithm, and a device authentication certificate is generated.

[0056] After that, the system enters the upgrade preparation stage, automatically loads the upgrade image file into the memory buffer, and performs multi-level security verification: the first level, calculates the quantum hash value of the image file; the second level, compares the pre-stored digital signature; the third level, verifies the validity of the certificate chain. Then, the hardware environment state is comprehensively detected: the storage medium health degree (bad block rate / residual life) is monitored; the power stability (voltage fluctuation range) is evaluated; and the cooling condition (temperature / fan speed) is checked. In the image writing stage, the writing strategy can be dynamically selected according to the environmental detection results: normal environment, full-speed writing of the complete image; high-temperature environment (> 75℃), starting the auxiliary cooling device before writing; low-bandwidth environment, enabling the block transmission mechanism (1MB per block).

[0057] After the image file is written, the boot configuration stage is entered, the device firmware type (UEFI (Unified Extensible Firmware Interface) or BIOS (Basic Input / Output System)) is automatically identified, and the startup parameters are intelligently configured: the firmware adaptation template is loaded; the startup order priority is set; the boot record is generated and the original configuration is backed up. Then, three verifications are performed: boot record integrity check; startup parameter compliance verification; and rollback mechanism feasibility test. In order to complete the system recovery after the upgrade, first, a hot restart operation is performed to preserve the memory key state, then the automatic diagnosis program is started, the kernel version consistency check, the driver compatibility test and the storage service functionality verification are performed, and the health status report is generated according to the test results: the green indicator light is activated for successful diagnosis, and the success log is recorded; the automatic rollback mechanism is triggered for failed diagnosis, and the state before the upgrade is restored.

[0058] Based on the previous embodiment, the present application can identify the device storing the image file and collect the current hardware state of the target device in real time, dynamically construct the upgrade path of the storage system through the directed acyclic graph, and perform system upgrade. Next, the process of simultaneously upgrading multiple nodes in the storage system in the present embodiment will be described in detail. Referring to Figure 6 As shown in the figure, the embodiment of the present application provides a storage system cluster upgrade method, which includes:

[0059] Step S21, if there are multiple target nodes in the storage system, the master node and the slave node in the target node are determined; the master node is used to determine the node upgrade strategy of the target node in the storage system according to the system upgrade strategy.

[0060] In the present embodiment, if there are multiple target nodes in the storage system, the master node and the slave node in the target node are first determined, and the master node is used to determine the node upgrade strategy of the target node in the storage system according to the system upgrade strategy.

[0061] Step S22: Determine the node upgrade operations in the node upgrade strategy, and construct a node upgrade instruction set based on each node upgrade operation; construct a Merkle tree based on the node upgrade instruction set, and determine the first root hash value of the Merkle tree.

[0062] In this embodiment, as Figure 7 As shown, the node upgrade operations in the node upgrade strategy can be determined, and a node upgrade instruction set can be constructed based on each node upgrade operation. Then, a Merkle tree is constructed based on the node upgrade instruction set, and the first root hash value of the Merkle tree is determined. In this embodiment, the master node is in the instruction construction stage. First, a complete upgrade instruction set is created, and the Merkle tree structure of the instruction set is generated, and the root hash value is calculated as the instruction digest. The specific Merkle tree construction process is as follows:

[0063] def build_merkle_tree(upgrade_steps):

[0064] """Building a cluster to upgrade the Merkle tree"

[0065] Args:

[0066] upgrade_steps: A list of upgrade steps, where each element is an step name, such as ["verify_signature", "download_image"].

[0067] Returns:

[0068] merkle_root: Root hash value (hexadecimal string)

[0069] tree: A dictionary of complete tree structures, containing the hash values ​​of each level.

[0070] """

[0071] # Step 1: Generate Leaf Node Hash - Generate a unique leaf node hash for each upgrade operation.

[0072] leaf_hashes = [] # Stores the hash values ​​of all leaf nodes

[0073] For step in upgrade_steps:

[0074] # 1.1 Convert the operation name to binary data (e.g., "download_image" -> b'download_image')

[0075] data_block = step.encode('utf-8')

[0076] # 1.2 Applying dual SHA-256 hashing enhances security

[0077] # - First hash: generates a binary hash value (tamper-proof)

[0078] first_hash = hashlib.sha256(data_block).digest()

[0079] # - Second hash: generates a hexadecimal string (resistant to length extension attacks)

[0080] leaf_hash = hashlib.sha256(first_hash).hexdigest()

[0081] leaf_hashes.append(leaf_hash)# Add to the list of leaf nodes

[0082] # Step 2: Build parent nodes layer by layer - Construct the complete Merkle tree from bottom to top

[0083] tree = {"level_0": leaf_hashes}# Store each level of the tree, starting with the leaf level

[0084] current_level = leaf_hashes # Current processing level (starts as the leaf level)

[0085] level = 0 # Current level depth identifier

[0086] # Loop until only the root node remains (continue when the number of nodes in the level > 1)

[0087] while len(current_level) > 1:

[0088] level += 1 # Move to the next level

[0089] parent_hashes = [] # Store the parent node hashes of the current level

[0090] # 2.1 Iterate through the nodes of the current level (take 2 at a time)

[0091] for i in range(0, len(current_level), 2):

[0092] left = current_level[i] # Left node in the current pair

[0093] # 2.2 Handling Odd Nodes: Copying the Last Node When the Number of Nodes is Odd

[0094] # - If it's the last and only node, pair it with itself

[0095] # - Ensure that each level generates a complete pair of parent nodes

[0096] if i+1 < len(current_level):

[0097] right = current_level[i+1] # Normally take the right node

[0098] else:

[0099] right = current_level[i] # Copy the last node as the right node

[0100] # 2.3 Generating Parent Node Hashes

[0101] # - Concatenate the hash values of the left and right child nodes into a string

[0102] combined = left + right

[0103] # - Perform SHA256 hashing on the concatenated string

[0104] parent_hash = hashlib.sha256(combined.encode()).hexdigest()

[0105] parent_hashes.append(parent_hash)

[0106] # 2.4 Storing Current Level Results and Updating Level State

[0107] tree[f"level_{level}"] = parent_hashes # Record all parent nodes of the current level

[0108] current_level = parent_hashes # Set the parent nodes as the current level for the next round of processing

[0109] # Return Result: Root Node (Top of the Tree Unique Hash) and Complete Tree Structure

[0110] # - When the loop ends, current_level only has one element, which is the root hash

[0111] return current_level[0], tree.

[0112] At step S23, the corresponding node upgrade instruction package is generated based on the first root hash value, and the node upgrade instruction package is broadcast to each slave node in the storage system by the master node; based on the node upgrade instruction package, the system upgrade of the master node and each slave node in the storage system is performed by using the mirror file.

[0113] In this embodiment, the corresponding node upgrade instruction package can be generated based on the first root hash value, and the node upgrade instruction package is broadcast to each slave node in the storage system by the master node, and then based on the node upgrade instruction package, the system upgrade of the master node and each slave node in the storage system is performed by using the mirror file. Specifically, in this embodiment, the Merkle tree can be reconstructed based on the node upgrade instruction package by using the preset operation list of each slave node locally, and the corresponding second root hash value of the reconstructed Merkle tree is determined; if the first root hash value and the second root hash value are equal, the corresponding second digital signature is generated based on the second root hash value, and the second digital signature of each slave node is sent to the master node, so that after the master node verifies the second digital signature, the aggregated signature is generated based on each second digital signature, and the corresponding upgrade execution instruction is generated based on the aggregated signature, and then the aggregated signature and the upgrade execution instruction are broadcast to each slave node by using the master node, so that each target node performs system upgrade according to the network time protocol by using the mirror file based on the aggregated signature and the upgrade execution instruction.

[0114] That is, in this embodiment, the consensus verification of each node can be implemented by using the master node. First, the master node broadcasts the instruction digest to all slave nodes, and the slave nodes verify the matching of the local instruction and the digest, and if the verification is passed, the slave nodes generate a digital signature, and mark the node that refuses to sign as an exception. Then, when performing system upgrade, the master node collects all valid signatures and generates an aggregated signature as an execution credential, and then broadcasts the execution instruction and the aggregated signature, so that the slave nodes verify the signature and synchronously execute the upgrade. And as shown in FIG. 8, in the process of distributing the Byzantine consensus instruction from the coordination layer to the storage layer, the cross-layer encrypted communication can also be implemented by using the quantum key management center through the QKD channel. Figure 2

[0115] ​And after the system upgrade of the master node and each slave node in the storage system, the master node needs to obtain the zero-knowledge proof sent by each slave node; wherein the zero-knowledge proof is generated by each slave node based on the system state parameter of the system after the system upgrade according to the preset constraint system; then the master node verifies each zero-knowledge proof by using the preset verification algorithm; if each zero-knowledge proof is verified successfully, it is determined whether the current system of each target node is consistent based on each zero-knowledge proof, and when the current systems of each target node are consistent, a corresponding system upgrade report is generated. Specifically, each node can generate a state proof after completing the upgrade, and submit a zero-knowledge proof, and the master node verifies the state consistency, which can specifically be that the master node verifies all proof files, and confirms that the upgrade is completed when the states are consistent. At the same time, it should be pointed out that if a node is lost, the cluster upgrade is suspended, and after the node is restored, the nodes are re-synchronized, and when the signature of the slave node is invalid, the problem node is isolated, and the remaining nodes continue to upgrade. In this way, through the zero-trust collaborative protocol based on the instruction verification system of the Merkle tree, the dual verification combining the aggregated signature and the zero-knowledge proof is realized, and the system security is further improved.

[0116] In the embodiment, the state difference nodes in each node can also be recorded. All nodes are rolled back to the snapshot before the upgrade, and the consistency verification process is re-initiated. At the same time, if malicious behavior of a node is detected, the repeatedly violated node is permanently isolated, and the corresponding trust list is updated. In this way, not only does the quantum encryption U disk provide a physically unclonable function, but also the quantum-resistant property of the NTRU algorithm combined with the zero-knowledge proof mechanism realizes verifiable security in the entire upgrade process, which can defend against hardware-level attacks and quantum computing threats and enhances the system security. At the same time, a three-level gradient fault-tolerant mechanism (retry → isolation → rollback) is constructed to automatically handle exceptions, and the zero-trust collaborative framework ensures atomic synchronization of multiple node states through Merkle tree verification and aggregated signature, realizes fault self-healing and cluster consistency guarantee, and improves the system reliability.

[0117] In this way, in combination with the technical solutions in the above embodiments, the quantum federation device identification system confirms the legality of the device, triggers the upgrade process, and then uses the directed acyclic graph upgrade engine to generate an optimal upgrade path according to the current environment state. During the upgrade process, the zero-trust collaborative framework is used to coordinate the synchronous execution of the upgrade instructions by multiple nodes to ensure cluster synchronization, and then the upgrade state is fed back to the identification system in real time to update the device trust state. In this way, the access device is first ensured to be safe and trustworthy, then a customized upgrade path is generated according to the device type, an optimal execution path is planned, and finally the zero-trust collaborative framework is used to ensure the consistency of multiple nodes, effectively solving the compatibility, reliability, and consistency problems of the storage cluster upgrade.

[0118] Based on the above embodiment, as shown in Figure 8 the embodiment provides a specific embodiment of the storage system with multiple nodes for upgrading, which comprises:

[0119] The master node first detects the USB device and parses the upgrade content, and analyzes the node topology, formulates an upgrade sequence strategy, determines a fault-tolerant processing scheme, to create a cluster upgrade plan. Then a global Merkle instruction tree is constructed, the upgrade steps are decomposed into atomic operations, a hash value is generated for each operation, and the parent node hash is calculated level by level to form a tree structure, and finally the root node hash value is obtained as the instruction digest.

[0120] When the cluster is upgraded, the nodes enter collaborative verification. The master node first broadcasts the upgrade instruction package (including the root hash value), and the slave node reconstructs the Merkle tree locally after receiving the instruction to verify consistency. After passing, an ED25519 digital signature is generated and returned to the master node. Then the master node performs Byzantine consensus, collects all node signatures, detects abnormal signatures (missing or invalid) and aggregates valid signatures to generate cluster execution credentials, marks offline nodes and invalid signature nodes, and updates the cluster topology graph.

[0121] During the upgrade process, the cluster nodes synchronize the upgrade. The master node first broadcasts the execution instruction and aggregated signature, and then each node uses the NTP (Network Time Protocol) protocol to achieve millisecond-level time synchronization upgrade, and executes the upgrade path planned by the local DAG engine. Before the key node executes, it performs cross-node state verification. The slave node reports the execution progress regularly to allow the master node to build a cluster state matrix. If the execution progress of the slave node deviates from the threshold (> 5% difference), the collaborative correction mechanism is triggered.

[0122] After the upgrade is completed, consistency needs to be performed. Each node first generates a zero-knowledge proof. As can be understood, the above zero-knowledge proof is constructed based on the R1CS (Rank-1 Constraint System) constraint system and contains the final system state parameters, without revealing specific configuration details. Then the master node runs a verification algorithm to check the validity of the proof, and calculates the cluster consistency index to output a global verification report.

[0123] As shown in Figure 9 The embodiment of the application also provides a storage system upgrade device, which comprises:

[0124] The feature matching module 11 is configured to collect the physical features of the target device, and match the target features in the preset feature library with the physical features; the target device stores an image file used for upgrading the storage system;

[0125] The device verification module 12 is configured to acquire a first digital signature of the target device when the target feature and the physical feature are successfully matched, and verify the first digital signature by using a preset public key; the first digital signature is a signature generated by the target device based on the physical feature by using a preset private key.

[0126] The policy determination module 13 is configured to acquire a hardware state of the storage system if the target device is successfully verified, and determine a system upgrade policy of the storage system based on the hardware state by using a preset upgrade engine; the preset upgrade engine is an engine constructed based on a target directed acyclic graph, and the target directed acyclic graph is used to record each upgrade step and the relationship between different upgrade steps.

[0127] The system upgrade module 14 is configured to read an image file in the target device, and perform system upgrade of a target node in the storage system by using the image file based on the system upgrade policy.

[0128] The features of the embodiments of the storage system upgrade apparatus can be referred to the related descriptions of the embodiments of the storage system upgrade method, which will not be repeated here.

[0129] Through the description of the above implementation, those skilled in the art can clearly understand that the method according to the above embodiments can be realized by means of software and necessary general hardware platform, of course, it can also be realized by hardware, but in many cases, the former is a better implementation.

[0130] The embodiments of the present application also provide an electronic device, comprising a memory and a processor, the memory stores a computer program, and the processor is configured to run the computer program to execute the steps in any of the above-mentioned embodiments of the storage system upgrade method.

[0131] The embodiments of the present application also provide a computer readable storage medium, which stores a computer program, wherein the computer program is configured to execute the steps in any of the above-mentioned embodiments of the storage system upgrade method when running.

[0132] In an exemplary embodiment, the above-mentioned computer readable storage medium can include but is not limited to: a U disk, a read-only memory (ROM), a random access memory (RAM), a mobile hard disk, a magnetic disk or an optical disk, and various media that can store computer programs.

[0133] The embodiments of the present application also provide a computer program product, which comprises a computer program, and the computer program is executed by a processor to implement the steps in any of the above-mentioned embodiments of the storage system upgrade method.

[0134] Embodiments of the present application further provide another computer program product comprising a non-transitory computer readable storage medium storing a computer program which, when executed by a processor, implements the steps of any of the above-mentioned storage system upgrade method embodiments.

[0135] Those skilled in the art will further appreciate that the units and algorithm steps of the examples described in connection with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or both, and that the implementation decisions are within the skill of an informed technician. The exemplary configurations and steps have been described above generally, but separately, from their hardware and software aspects in order to illustrate the inter-changeability of hardware and software features. The particular application implementation disclosed will depend on the specific application and design constraints imposed on the overall system. Skilled artisans will recognize the interchangeability of various hardware and software components and the many specific arrangements that software and computer-aided components will take. For example, those skilled in the art will appreciate that computing devices of the exemplary configurations can employ computer-aided components, such as a CPU, GPU, TPU, ASIC, FPGA, etc., to perform the steps described herein, and that one or more of these components can be shared or distributed as desired. Those skilled in the art will also recognize or be able to ascertain, using no more than routine skill in the art, other equivalent steps and components not expressly discussed herein that are useful in implementing the examples. It is therefore intended that the scope of the application be determined by the following claims rather than by the explicit description above.

[0136] The above provides a storage system upgrade method, device, storage medium and program product. The principles and implementation modes of the present application are described by applying specific examples. The above description of the examples is only applicable to help understand the method and core idea of the present application. It should be pointed out that, for those skilled in the art, without departing from the principles of the present application, some improvements and modifications can be made to the present application. These improvements and modifications also fall within the scope of protection of the claims of the present application.

Claims

1. A storage system upgrade method characterized by comprising: The method comprises the following steps: Collecting physical features of a target device and matching target features in a preset feature library with the physical features; An image file for upgrading a storage system is stored in the target device; When the target features and the physical features are successfully matched, a first digital signature of the target device is obtained, and the first digital signature is verified by using a local preset public key; the first digital signature is a signature generated by the target device based on the physical features by using a preset private key; If the target device is successfully verified, a hardware state of the storage system is obtained, and a system upgrade strategy of the storage system is determined based on the hardware state by using a preset upgrade engine; The preset upgrade engine is an engine constructed based on a target directed acyclic graph, and the target directed acyclic graph is used to record each upgrade step and the relationship between different upgrade steps; The image file in the target device is read, and system upgrade of a target node in the storage system is performed by using the image file based on the system upgrade strategy; The matching of the target features in the preset feature library with the physical features comprises the following steps: When the target features and the physical features are not successfully matched, a feature vector of the physical features is extracted by using a target device recognition model; Configuration parameters of the target device are determined, and target features corresponding to the target device are generated based on the feature vector and the configuration parameters; The target features corresponding to the target device are saved in the preset feature library; Correspondingly, before the feature vector of the physical features is extracted by using the target device recognition model, the following steps are further included: A preset device recognition model is deployed at an edge computing layer of the storage system, and the preset device recognition model is trained by using the physical features to obtain an initial device recognition model; Optimization parameters in the initial device recognition model are extracted, and the optimization parameters are encrypted by using a preset quantum key; The encrypted optimization parameters are uploaded to a central server, and an updated device recognition model issued by the central server is obtained; the updated device recognition model is a model obtained by updating a local preset global model based on target parameters obtained by the central server by fusing the encrypted optimization parameters based on a federated aggregation algorithm; The initial device recognition model is covered by using the updated device recognition model to obtain the target device recognition model.

2. The storage system upgrade method of claim 1, wherein, The first digital signature of the target device is obtained by including the following steps: A hardware feature code of the target device is obtained, and a target feature code in a preset device permission list is matched with the hardware feature code; When the target feature code and the hardware feature code are successfully matched, the first digital signature of the target device is obtained.

3. The storage system upgrade method of claim 2, wherein, In the process of performing system upgrade of a target node in the storage system by using the image file based on the system upgrade strategy, the following steps are further included: The first digital signature of the target device is re-obtained based on a preset time period, and the first digital signature is verified by using the local preset public key; If the target device is verified successfully, continue to upgrade the target node in the storage system by using the image file; If the target device is verified unsuccessfully, stop upgrading the target node in the storage system by using the image file.

4. The storage system upgrade method of claim 1, wherein, Before the step of determining the system upgrade strategy of the storage system based on the hardware state by using the preset upgrade engine, the method further comprises: determining an image file writing step and an image file writing condition of the storage system, and determining a writing operation corresponding to different hardware states of the storage system; constructing the target directed acyclic graph according to the image file writing step, the image file writing condition and the writing operation corresponding to different hardware states, so as to construct the preset upgrade engine based on the target directed acyclic graph.

5. The storage system upgrade method of claim 4, wherein, In the process of upgrading the target node in the storage system by using the image file based on the system upgrade strategy, the method further comprises: collecting system state parameters of the storage system based on a preset parameter collection period; determining whether the system state parameters meet a preset system upgrade condition; if the system state parameters do not meet the preset system upgrade condition, adjusting the system upgrade strategy based on the system state parameters by using the preset upgrade engine, so as to upgrade the target node in the storage system by using the image file based on the adjusted system upgrade strategy.

6. The storage system upgrade method of claim 5, wherein, In the process of upgrading the target node in the storage system by using the image file based on the system upgrade strategy, the method further comprises: determining a file data amount of the image file written into the storage system; if the file data amount is equal to a preset data writing threshold, performing a cyclic redundancy check on the image file written into the storage system; if the image file written into the storage system passes the check and the image file has been completely written into the storage system, determining a hash value of the image file written into the storage system by using a preset encryption hash function, and performing a check on the hash value, so as to determine that the image file writing is completed after the hash value passes the check.

7. The storage system upgrade method of claim 1, wherein, After the step of upgrading the target node in the storage system by using the image file based on the system upgrade strategy, the method further comprises: performing a hot restart operation of the storage system; after the hot restart operation of the storage system is completed, detecting the upgraded storage system based on a preset system detection process; if the upgraded storage system passes the detection, generating and recording a corresponding system upgrade log; if the upgraded storage system fails the detection, rolling back the storage system.

8. The storage system upgrade method according to any one of claims 1 to 7, wherein, In the step of upgrading the target node in the storage system by using the image file based on the system upgrade strategy, the method further comprises: if the storage system has a plurality of target nodes, determining a master node and a slave node in the target nodes; determining a node upgrade strategy of the target node in the storage system based on the system upgrade strategy by using the master node; determining a node upgrade operation in the node upgrade strategy, and constructing a node upgrade instruction set based on each node upgrade operation; constructing a Merkle tree based on the node upgrade instruction set, and determining a first root hash value of the Merkle tree; generating a corresponding node upgrade instruction package based on the first root hash value, and broadcasting the node upgrade instruction package to each slave node in the storage system through the master node; performing system upgrade of the master node and each slave node in the storage system based on the node upgrade instruction package and the mirror file.

9. The storage system upgrade method of claim 8, wherein, The performing system upgrade of the master node and each slave node in the storage system based on the node upgrade instruction package and the mirror file comprises: reconstructing a Merkle tree based on a preset operation list of each slave node, and determining a second root hash value of the reconstructed Merkle tree; if the first root hash value and the second root hash value are equal, generating a corresponding second digital signature based on the second root hash value; sending the second digital signature of each slave node to the master node, so that the master node generates an aggregated signature based on each second digital signature after verifying the second digital signature, and generates a corresponding upgrade execution instruction based on the aggregated signature; broadcasting the aggregated signature and the upgrade execution instruction to each slave node by the master node, so that each target node performs system upgrade based on the aggregated signature and the upgrade execution instruction according to the network time protocol using the mirror file.

10. The storage system upgrade method of claim 9, wherein, After the performing system upgrade of the master node and each slave node in the storage system based on the node upgrade instruction package and the mirror file, the method further comprises: obtaining zero-knowledge proofs sent by each slave node by the master node; the zero-knowledge proofs are generated by the slave nodes based on preset constraint systems according to system state parameters after system upgrade; verifying each zero-knowledge proof by the master node through a preset verification algorithm; if each zero-knowledge proof is verified successfully, determining whether the current systems of each target node are consistent based on each zero-knowledge proof; generating a corresponding system upgrade report when the current systems of each target node are consistent.

11. An electronic device, comprising: comprise: a memory for storing a computer program; a processor for executing the computer program to implement the steps of the storage system upgrade method according to any one of claims 1 to 10.

12. A computer-readable storage medium, characterized in that, The computer readable storage medium stores a computer program, wherein the computer program is executed by the processor to implement the steps of the storage system upgrade method according to any one of claims 1 to 10.

13. A computer program product comprising a computer program, characterized in that, The computer program is executed by the processor to implement the steps of the storage system upgrade method according to any one of claims 1 to 10.

Images