Software and hardware deep fusion integrated system for infrastructure

Abstract

The invention discloses an infrastructure software and hardware deep fusion integrated system, and relates to the field of software and hardware fusion. The infrastructure software and hardware deep fusion integrated system comprises control service, computing power sharing, storage service and system security. According to the infrastructure software and hardware deep fusion integrated system, a synchronous heterogeneous data processing platform is set up in the integrated system, a specific local equipment task is decomposed, data distribution is carried out from two parts, namely a computing node and a network node, by adopting a CPU + GPU heterogeneous fusion mode and utilizing an edge server, processing task data decomposition is realized, and the data processing efficiency is improved. Specific tasks of the local equipment are processed in a parallel computing mode, so that an external cloud server provides additional computing power for the local equipment, task processing of the local equipment is accelerated, the local task processing speed is increased on one hand, and on the other hand, some equipment with old hardware conditions can still meet operation and use requirements; the replacement cost is reduced.

Description

Technical Field

[0001] The present invention relates to the technical field of software and hardware integration, and specifically to an infrastructure software and hardware deep integration integrated system. Background Art

[0002] Nowadays, with the development of Internet technology, centralized management of various infrastructures with Internet communication functions is achieved through centralized Internet management systems. On the one hand, the efficiency of equipment management is greatly improved, and on the other hand, the safety of equipment operation and maintenance is ensured by establishing monitoring methods.

[0003] In the existing technology, centralized management technology is mostly used. By establishing a unified management software port and installing corresponding networking software in local devices, data interaction management is performed through software. However, in actual applications, there is essentially technical isolation between software and hardware. The software part only provides data management functions for local hardware and the local operating system of the hardware. The system integration degree of the two is low. Under the current situation of rapid replacement of software information, some old basic equipment and operating systems are facing elimination. The main reason is that the hardware capabilities of old equipment cannot meet the demand for computing power growth. At the same time, the local storage space of the equipment is consumed, resulting in slow response speed when performing local task processing. For this reason, a deep integration system of infrastructure software and hardware is provided to solve the above problems. Summary of the invention

[0004] In view of the shortcomings of the prior art, the present invention provides an integrated system with deep integration of infrastructure software and hardware, which solves the problem that the existing infrastructure management system only provides data management functions for local hardware and the local operating system of the hardware, the degree of integration between the two systems is low, the hardware capabilities of old equipment cannot meet the demand for increased computing power, and the local storage space of the equipment is consumed, resulting in slow response speed when performing local task processing, facing the problem of equipment obsolescence.

[0005] To achieve the above objectives, the present invention is implemented through the following technical solutions: an infrastructure software and hardware deep integration integrated system, including;

[0006] Control services, by building an integrated system, provide more control services for infrastructure through a centralized network system, including communication connections, network services and data hyper-convergence;

[0007] Computing power sharing: cloud servers of components in the integrated system provide additional cloud computing power sharing for local infrastructure. When local devices perform task processing, the system estimates and judges the task equivalent, and intelligently provides corresponding additional cloud computing power for different local devices to accelerate the task processing speed of local devices. This process builds a synchronous heterogeneous data processing platform in the integrated system, decomposes specific local device tasks, adopts a heterogeneous CPU+GPU approach, uses edge servers, and heterogeneously integrates CPU+GPU to distribute data from computing nodes and network nodes, thereby decomposing processing task data.

[0008] Storage service: cloud servers provide cloud storage services for local devices to alleviate storage pressure on local devices;

[0009] System security is achieved by establishing network isolation in the system function framework and establishing a one-way channel between the local device, the integrated system and the Internet to ensure the communication security of local devices. At the same time, by mirroring the local device system and designing a cloud reset solution, the operating system of the local device is encapsulated and protected to achieve control system security. In addition, cloud backup is provided for local device data through storage services to ensure data security.

[0010] Preferably, the integrated system interface architecture includes a device list, a storage system and developer options. The corresponding connected basic devices can be queried through the device list, and the local devices can be operated and managed accordingly in the corresponding basic device interface. A corresponding device list is also established in the storage system. The operation logs and historical storage data of the corresponding devices in the cloud server can be queried through the corresponding device list. The developer options are used by developers to adjust and set the interface framework of this system.

[0011] Preferably, the equipment list includes:

[0012] The device control page, by designing a corresponding virtual operation interface for the operation function of each device, realizes remote control of the device through remote communication technology;

[0013] The device monitoring page uses the communication function to establish connections between existing devices with communication functions, and conducts real-time monitoring of the working status of the device and the internal data of the device, including temperature, voltage and power;

[0014] The data service page allows you to remotely access the internal system data of the local device. At the same time, the local system uses the data service to upload local data to the cloud server.

[0015] Preferably, the communication connection realizes data transmission between the local device and the integrated device by establishing a communication service between the local device and the integrated system. The network service uses the integrated system as a network bridge to provide a network access channel for localized devices without network services, so that the localized devices can perform data updates and system upgrades online, and at the same time provide network online diagnostic services when troubleshooting local device failures.

[0016] Preferably, the data hyper-convergence integrates the technical architecture of computing, storage, network and application services. The architecture centrally manages the resources of local devices of the same type by integrating the independent systems of multiple independent local devices into the system center of the integrated system. The technology includes automated resource management, high-performance network communication and intelligent data processing and storage.

[0017] Preferably, the automated resource management is built-in using existing resource management tools to achieve automatic distribution of local task data and collection of data results. The high-performance network communication utilizes RDMA technology, based on a kernel bypass mechanism, to allow applications to directly access the memory space of the network card and directly write data to the memory space of the network card, allowing the receiving end to read data directly from the memory of the sending end. The intelligent data processing and storage, by optimizing the underlying logic and instruction-level tuning, improves the IO performance response speed between local devices and the integrated system through technologies such as cold and hot data stratification and read-write cache, and optimizes the efficiency of data processing and storage.

[0018] Preferably, in the computing power sharing service, through the communication connection established between the local device and the integrated system, when the local device performs data-related task processing, task data capture is implemented through the device monitoring page to obtain the device task content, and then the data endpoint content interaction is performed through the microservice status between the hardware system and the integrated system, and then the computing power equivalent of the current task is estimated, and the local processing time of the task is calculated based on the current device computing power. When the time exceeds the set threshold, cloud computing power access is performed through intelligent allocation, and additional cloud computing services are provided for the local tasks of the local device, thereby improving the task processing response speed.

[0019] Preferably, when estimating the computing power equivalent of the task, common processing tasks of different types of local devices are exhaustively listed, and the equivalent of each type of task is annotated. For some uncommon tasks, their equivalents are estimated through online analysis, thereby completing the calculation and estimation of the computing power requirements of local tasks.

[0020] Preferably, during the cloud computing power access process, by optimizing data transmission and task allocation and adopting data decomposition, parallel computing data processing tasks are extracted from the main memory and handed over to the GPU for completion, and complex computing tasks are handed over to the cloud CPU for processing. The processed data is then converted into a unified format through the data analysis library and stored in the target file, and the file information is then fed back to the local device.

[0021] Preferably, the network isolation is achieved by setting a network adapter on the data transmission port to allow only a single channel to be opened, so that when the local device interacts with the integrated system for data, it is isolated from the Internet communication port, and when the integrated system interacts with the Internet, it is isolated from the local device communication port, thereby ensuring the data communication security of the local device.

[0022] The present invention discloses an infrastructure software and hardware deep integration integrated system, which has the following beneficial effects:

[0023] 1. The infrastructure is an integrated system with deep integration of software and hardware. By building a synchronous heterogeneous data processing platform in the integrated system, the specific local device tasks are decomposed, and the CPU+GPU heterogeneous mode is adopted. The edge server and the heterogeneous fusion CPU+GPU are used to distribute data from the computing nodes and network nodes, and the processing task data is decomposed. The specific tasks of the local devices are processed in a parallel computing manner, so that the external cloud server can provide additional computing power for the local devices and accelerate the task processing of the local devices. On the one hand, it improves the local task processing speed, and on the other hand, it can still meet the operation requirements of some equipment with outdated hardware conditions and reduce the replacement cost.

[0024] 2. The infrastructure is an integrated system with deep integration of software and hardware, which adopts data hyper-convergence. By integrating the independent systems of multiple independent local devices into the system center of the integrated system, the centralized management of local device resources of the same type is implemented with existing resource management tools built in, so as to realize the automatic distribution of local task data and the collection of data results. The high-performance network communication utilizes RDMA technology and is based on the kernel bypass mechanism, which allows the application to directly access the memory space of the network card, write data directly to the memory space of the network card, and allow the receiving end to read data directly from the memory of the sending end. Intelligent data processing and storage, by optimizing the underlying logic and instruction-level tuning, through the technology of hot and cold data stratification, read-write cache, etc., improves the IO performance response speed between local devices and the integrated system, optimizes the efficiency of data processing and storage, and improves the response speed of local old devices.

[0025] 3. The infrastructure is an integrated system with deep integration of software and hardware. A corresponding device list is also established in the storage system. The operation logs and historical storage data of the corresponding devices in the cloud server are queried through the corresponding device list. The cloud server provides cloud storage services for local devices to alleviate the storage pressure of local devices. The cloud server establishes corresponding storage folders for all local devices, and stores the historical operation logs and local device storage data of local devices respectively, alleviating the hardware storage pressure of local devices. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0027] Figure 1 This is a functional architecture diagram of the integrated system of the present invention;

[0028] Figure 2 It is a system interface architecture diagram of the integrated system of the present invention;

[0029] Figure 3 This is a flowchart of computing power sharing of the present invention;

[0030] Figure 4 Synchronous heterogeneous guide map for computing power allocation of the present invention;

[0031] Figure 5 This is a guide view for allocating computing power between software and hardware according to the present invention. DETAILED DESCRIPTION

[0032] In order to make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention are clearly and completely described. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0033] The embodiments of the present application provide an integrated system with deep integration of infrastructure software and hardware, which solves the problem that the existing infrastructure management system only provides data management functions for local hardware and the local operating system of the hardware, and the system integration between the two is low. The hardware capabilities of old equipment cannot meet the demand for increased computing power. At the same time, the local storage space of the equipment is consumed, resulting in slow response speed when performing local task processing, and facing the problem of equipment obsolescence.

[0034] In order to better understand the above technical solution, the above technical solution will be described in detail below in conjunction with the accompanying drawings and specific implementation methods.

[0035] The embodiment of the present invention discloses an infrastructure software and hardware deep integration integrated system.

[0036] According to the attached Figure 1-5 As shown, including;

[0037] Control services, by building an integrated system, provide more control services for infrastructure through a centralized network system, including communication connections, network services and data hyper-convergence;

[0038] Computing power sharing, through the cloud servers of the components in the integrated system, provides additional cloud computing power sharing for the local infrastructure. When the local device is processing a task, the system estimates the task equivalent and then makes a judgment, and intelligently provides corresponding additional cloud computing power for different local devices, accelerating the task processing speed of local devices. This process builds a synchronous heterogeneous data processing platform in the integrated system, decomposes the specific local device tasks, adopts the heterogeneous method of CPU+GPU, uses edge servers, heterogeneously integrates CPU+GPU, distributes data from computing nodes and network nodes, and realizes the decomposition of processing task data; according to the complexity of instructions, typical processor platforms are roughly divided into CPU, coprocessor, GPU, FPGA, DSA, and ASIC. From left to right, unit calculations become more and more complex, performance is getting better and better, but flexibility is getting lower and lower. Various types of processors such as CPU, GPU, DSA, etc. are essentially software and hardware collaboration based on software and hardware decoupling at different levels. When performing heterogeneous synchronization of computing architectures, they move from homogeneous computing to heterogeneous computing; and from heterogeneous computing to more heterogeneous fusion computing with heterogeneous collaboration. This architecture allows computing tasks to be assigned to the hardware unit that is most suitable for processing the task, thereby improving overall computing efficiency. The CPU and GPU communicate and exchange data through specific interfaces (such as the PCIe bus), offloading computing-intensive tasks to the GPU while keeping other computing on the CPU running, thereby achieving high-performance computing.

[0039] Storage service: cloud servers are used to provide cloud storage services for local devices, alleviating storage pressure on local devices. Cloud servers are used to create corresponding storage folders for all local devices, storing historical operation logs of local devices and local device storage data respectively, alleviating hardware storage pressure on local devices.

[0040] System security is achieved by establishing network isolation in the system function framework and establishing a one-way channel between local devices, integrated systems and the Internet to ensure the communication security of local devices. At the same time, by mirroring the local device system and designing a cloud reset solution, the operating system of the local device is encapsulated and protected to achieve control system security. In addition, cloud backup is provided for local device data through storage services to ensure data security.

[0041] The integrated system interface architecture includes a device list, storage system and developer options. The corresponding connected basic devices can be queried through the device list. In the corresponding basic device interface, the local devices can be operated and managed accordingly. A corresponding device list is also established in the storage system. The operation logs and historical storage data of the corresponding devices in the cloud server can be queried through the corresponding device list. The developer options are used by developers to adjust and set the interface framework of this system.

[0042] The equipment list includes;

[0043] The device control page, by designing a corresponding virtual operation interface for the operation function of each device, realizes remote control of the device through remote communication technology;

[0044] The device monitoring page uses the communication function to establish connections between existing devices with communication functions, and conducts real-time monitoring of the working status of the device and the internal data of the device, including temperature, voltage and power;

[0045] The data service page allows you to remotely access the internal system data of the local device. At the same time, the local system uses the data service to upload local data to the cloud server.

[0046] The communication connection realizes data transmission between local devices and integrated devices by establishing communication services between them. The network service uses the integrated system as a network bridge to provide network access channels for localized devices without network services, so that localized devices can update data and upgrade systems online. At the same time, it provides network online diagnostic services when troubleshooting local device failures.

[0047] Data hyper-convergence is a technical architecture that integrates computing, storage, network and application services. This architecture centrally manages the resources of local devices of the same type by integrating the independent systems of multiple independent local devices into the system center of the integrated system. The technology includes automated resource management, high-performance network communication and intelligent data processing and storage.

[0048] Automated resource management is built-in using existing resource management tools to achieve automatic distribution of local task data and collection of data results. High-performance network communication uses RDMA technology and a kernel bypass mechanism to allow applications to directly access the memory space of the network card and write data directly to the memory space of the network card, allowing the receiving end to read data directly from the memory of the sending end. Intelligent data processing and storage improves the IO performance response speed between local devices and integrated systems and optimizes data processing and storage efficiency by optimizing the underlying logic and instruction-level tuning, and through technologies such as hot and cold data tiering and read-write caching.

[0049] In the computing power sharing service, through the communication connection established between the local device and the integrated system, when the local device performs data-related task processing, the task data is captured through the device monitoring page to obtain the device task content, and then the data endpoint content is interacted through the microservice status between the hardware system and the integrated system. Then, the computing power equivalent of the current task is estimated, and the local processing time of the task is calculated based on the current device computing power. When the time exceeds the set threshold, cloud computing power is accessed through intelligent allocation to provide additional cloud computing services for local tasks of local devices and improve the task processing response speed.

[0050] When estimating the computing power equivalent of a task, we exhaustively list the common processing tasks of different types of local devices, annotate the equivalent of each type of task, and estimate the equivalent of some uncommon tasks through online analysis, thereby completing the calculation and estimation of the computing power requirements of local tasks.

[0051] During the access process of cloud computing power, by optimizing data transmission and task allocation and adopting data decomposition, parallel computing data processing tasks are extracted from the main memory and handed over to the GPU for completion, and complex computing tasks are handed over to the cloud CPU for processing. The processed data is then converted into a unified format through the data analysis library, stored in the target file, and the file information is then fed back to the local device.

[0052] Network isolation is achieved by setting a network adapter on the data transmission port and allowing only a single channel to be opened. This ensures that when the local device interacts with the integrated system, it is isolated from the Internet communication port. When the integrated system interacts with the Internet, it is isolated from the local device communication port, thereby ensuring the data communication security of the local device.

[0053] This infrastructure is an integrated system with deep integration of software and hardware. It provides more control services, computing power sharing services, storage services and system security for the infrastructure through a centralized network system. It realizes remote control and monitoring protection of the equipment by establishing relevant device pages in the integrated system interface. At the same time, by building a synchronous heterogeneous data processing platform in the integrated system, the specific local device tasks are decomposed, and the CPU+GPU heterogeneous mode is adopted. The edge server and the heterogeneous fusion CPU+GPU are used to distribute data from the computing nodes and network nodes to realize the decomposition of processing task data. The specific tasks of the local equipment are processed in a parallel computing mode, so that the external cloud server can provide additional computing power for the local equipment and accelerate the task processing of the local equipment.

[0054] At the same time, additional cloud storage services are provided through cloud servers, and network adapters are set at the data transmission port through network isolation to protect the network services of local devices, prevent the local device system from being intruded, and prevent the data information of local devices from being stolen.

[0055] The above shows and describes the basic principles and main features of the present invention and the advantages of the present invention. It should be understood by those skilled in the art that the present invention is not limited to the above embodiments. The above embodiments and descriptions are only for explaining the principles of the present invention. Without departing from the spirit and scope of the present invention, the present invention may have various changes and improvements, which fall within the scope of the present invention to be protected. The scope of protection of the present invention is defined by the attached claims and their equivalents.

Claims

1. An infrastructure software and hardware deep integration system, characterized by: include; Control services, by building an integrated system, provide more control services for infrastructure through a centralized network system, including communication connections, network services and data hyper-convergence; Computing power sharing: cloud servers of components in the integrated system provide additional cloud computing power sharing for local infrastructure. When local devices perform task processing, the system estimates and judges the task equivalent, and intelligently provides corresponding additional cloud computing power for different local devices to accelerate the task processing speed of local devices. This process builds a synchronous heterogeneous data processing platform in the integrated system, decomposes specific local device tasks, adopts a heterogeneous CPU+GPU approach, uses edge servers, and heterogeneously integrates CPU+GPU to distribute data from computing nodes and network nodes, thereby decomposing processing task data. Storage service: cloud servers provide cloud storage services for local devices to alleviate storage pressure on local devices; System security is achieved by establishing network isolation in the system function framework and establishing a one-way channel between local devices, integrated systems and the Internet to ensure the communication security of local devices. At the same time, by mirroring the local device system and designing a cloud reset solution, the operating system of the local device is encapsulated and protected to achieve control system security. In addition, cloud backup is provided for local device data through storage services to ensure data security.

2. The infrastructure software and hardware deep integration system according to claim 1 is characterized by: The integrated system interface architecture includes a device list, a storage system and developer options. The corresponding connected basic devices can be queried through the device list. In the corresponding basic device interface, the local devices can be operated and managed accordingly. A corresponding device list is also established in the storage system. The operation logs and historical storage data of the corresponding devices in the cloud server can be queried through the corresponding device list. The developer options are used by developers to adjust and set the interface framework of this system.

3. The infrastructure software and hardware deep integration system according to claim 2 is characterized by: The equipment list includes: The device control page, by designing a corresponding virtual operation interface for the operation function of each device, realizes remote control of the device through remote communication technology; The device monitoring page uses the communication function to establish connections between existing devices with communication functions, and conducts real-time monitoring of the working status of the device and the internal data of the device, including temperature, voltage and power; The data service page allows you to remotely access the internal system data of the local device. At the same time, the local system uses the data service to upload local data to the cloud server.

4. The infrastructure software and hardware deep integration system according to claim 1 is characterized by: The communication connection realizes data transmission between the local device and the integrated device by establishing a communication service between the local device and the integrated system. The network service uses the integrated system as a network bridge to provide a network access channel for localized devices without network services, so that the localized devices can perform data updates and system upgrades online, and at the same time provide network online diagnostic services when troubleshooting local device failures.

5. The infrastructure software and hardware deep integration system according to claim 1 is characterized by: The data hyper-convergence integrates the technical architecture of computing, storage, network and application services. The architecture centrally manages the resources of local devices of the same type by integrating the independent systems of multiple independent local devices into the system center of the integrated system. The technology includes automated resource management, high-performance network communication and intelligent data processing and storage.

6. The infrastructure software and hardware deep integration system according to claim 5 is characterized by: The automated resource management is built-in with existing resource management tools to realize automatic distribution of local task data and collection of data results. The high-performance network communication utilizes RDMA technology, based on the kernel bypass mechanism, to allow applications to directly access the memory space of the network card, directly write data to the memory space of the network card, and allow the receiving end to directly read data from the memory of the sending end; The intelligent data processing and storage improves the IO performance response speed between local devices and the integrated system and optimizes the efficiency of data processing and storage by optimizing the underlying logic and instruction-level tuning, and by using technologies such as hot and cold data tiering and read-write cache.

7. The infrastructure software and hardware deep integration system according to claim 1 is characterized by: In the computing power sharing service, through the communication connection established between the local device and the integrated system, when the local device performs data-related task processing, the task data is captured through the device monitoring page to obtain the device task content, and then the data endpoint content is interacted through the microservice status between the hardware system and the integrated system. Then, the computing power equivalent of the current task is estimated, and the local processing time of the task is calculated based on the current device computing power. When the time exceeds the set threshold, cloud computing power is accessed through intelligent allocation to provide additional cloud computing services for local tasks of local devices, thereby improving the task processing response speed.

8. The infrastructure software and hardware deep integration system according to claim 7 is characterized by: When estimating the task computing power equivalent, common processing tasks of different types of local devices are exhaustively listed, and the equivalent of each type of task is annotated. For some uncommon tasks, their equivalents are estimated through online analysis, thereby completing the calculation and estimation of the computing power requirements of local tasks.

9. The infrastructure software and hardware deep integration system according to claim 7 is characterized by: During the cloud computing power access process, by optimizing data transmission and task allocation and adopting data decomposition, parallel computing data processing tasks are extracted from the main memory and handed over to the GPU for completion, and complex computing tasks are handed over to the cloud CPU for processing. The processed data is then converted into a unified format through the data analysis library and stored in the target file, and the file information is then fed back to the local device.

10. The infrastructure software and hardware deep integration system according to claim 1 is characterized by: The network isolation is achieved by setting a network adapter on the data transmission port and allowing only a single channel to be opened, so that when the local device interacts with the integrated system for data, it is isolated from the Internet communication port, and when the integrated system interacts with the Internet, it is isolated from the local device communication port, thereby ensuring the data communication security of the local device.

Images