Cloud electronic sports diskless storage method and system

Abstract

The invention discloses a cloud E-sports diskless storage method and system, relates to the technical field of diskless storage, and solves the technical problem that the diskless computer performance is reduced due to the fact that a server side of a diskless computer system is easy to have a single-point fault or a single-point performance bottleneck. The method comprises the following steps: installing a local hard disk on the diskless computer, and setting the local hard disk as a non-system disk; starting the diskless computer, providing read-write capability for the diskless computer through IO service compiled by the diskless server, and performing initial guidance of the diskless computer; a local IO service is automatically started in the operating system, and a network hard disk drive is installed in a local hard disk and configured; and the network hard disk drive redirects the IO service request to a local IO service, and the local IO service accesses a storage cluster of the diskless server. The diskless computer performs data write-back through the local hard disk, the network communication frequency is reduced, meanwhile, the network hard disk drive redirects the IO request to the local IO service, delay is not prone to occurring, and the performance of the diskless computer is ensured.

Description

Technical Field

[0001] The present invention relates to the field of diskless storage technology, and in particular to a cloud e-sports diskless storage method and system. Background Art

[0002] Generally speaking, the operating system is installed on the local hard disk for "disk-based booting". However, in many fields such as Internet cafes, high-performance computing, embedded systems, hotels, and education, in order to reduce the cost of hard disk procurement, facilitate hard disk content management, and prevent users from writing unsafe content to the hard disk, computers usually do not have local hard disks, and need to load and run the operating system from a remote server for "diskless booting". Compared with ordinary computers, diskless computers can start and run the operating system without any external memory (floppy drive, hard disk, CD, etc.) support. The way to achieve diskless booting is generally for the diskless host to download some necessary files in the operating system from the server, such as boot program, configuration file, kernel and image file, etc., to the memory of the diskless workstation, and then start the diskless computer. All information in the diskless boot system during operation is stored in the memory.

[0003] Traditional diskless booting is to deploy an ISCSI (Internet Small Computer System Interface) server on the server, connect storage devices (such as hard disk drives or tape drives) to diskless computers through the network, and provide network hard disk capabilities for diskless computers. Cloud e-sports uses the edge cloud base as a basis to provide a high-configuration computing environment, and uses streaming services to interact with users through the network to render the game screen and operation instructions, which places extremely high demands on the real-time and reliability of diskless system storage. However, the ISCSI server is deployed on a single server, which has a single point of failure or a single point of performance bottleneck. When the server fails, it is easy for diskless computers to freeze, and when the number of machines reaches hundreds, all diskless computers will freeze, affecting the user experience of cloud e-sports. At the same time, if other storage systems (such as the distributed file system ceph) are connected to the ISCSI backend, after multiple network communications, it is easy to have high latency, resulting in low performance of diskless computers, which is difficult to adapt to the needs of cloud e-sports usage scenarios.

[0004] In the process of implementing the present invention, the inventors found that there are at least the following problems in the prior art:

[0005] The server side of a diskless computer system is prone to single point failures or single point performance bottlenecks, which results in high latency, reduces the performance of the diskless computer, and makes it difficult to adapt to the needs of cloud e-sports scenarios, affecting the user experience. Summary of the invention

[0006] The purpose of the present invention is to provide a cloud e-sports diskless storage method and system to solve the technical problem that the diskless computer system server has a single point of failure or a single point of performance bottleneck in the prior art, resulting in high latency, reducing the performance of the diskless computer, making it difficult to adapt to the use needs of cloud e-sports scenarios, and affecting the user experience. The many technical effects that can be produced by the preferred technical solution among the many technical solutions provided by the present invention are described in detail below.

[0007] To achieve the above object, the present invention provides the following technical solutions:

[0008] The present invention provides a cloud e-sports diskless storage method, in which a diskless computer writes back e-sports game content through a local hard disk, comprising the following steps: S100: installing a local hard disk on the diskless computer and setting it as a non-system disk; S200: the diskless computer is started, and the IO service written by the diskless server provides the diskless computer with read and write capabilities, and performs initial booting of the diskless computer; S300: the diskless computer is successfully started and enters the operating system, a local IO service is automatically started in the operating system, and a network hard disk driver is installed and configured on the local hard disk of the diskless computer; S400: the network hard disk driver redirects the IO service request to the local IO service, and the local IO service directly accesses the storage cluster of the diskless server through the network.

[0009] Preferably, in step S300, the network hard disk driver is installed manually or automatically with administrator privileges, and when the device name and driver date in the device manager are consistent with those in the inf driver file, it indicates that the network hard disk driver is successfully installed.

[0010] Preferably, in the step S300, when the network hard disk driver is successfully installed, the IP address of the Log Server IP is modified to the log transfer IP of the current area, and the sc query vfdisk command and log upload query are used to confirm whether the local hard disk is correctly configured.

[0011] Preferably, in step S400, the IO request is redirected to the local IO service through an IO transmission protocol, and the format of the IO transmission protocol includes a common packet header, a type definition, a response return code definition, an unknown type response packet, a heartbeat packet and a read packet.

[0012] Preferably, in the IO transmission protocol, all structures are 1-byte aligned, all fields are in little-endian byte order, all requests and responses are defined from the diskless computer side, and the diskless server side should return an unsupported response if the request packet format is correct but the type is not supported.

[0013] Preferably, the storage cluster includes at least two ISCSI network hard disks or RBD network virtual hard disks, one of which is a system disk.

[0014] A cloud e-sports diskless storage system is used to run a cloud e-sports diskless storage method as described in any of the above items, including several diskless computers, a diskless server, a communication module and a management module; the diskless computer is connected to an interactive device for users to operate cloud e-sports games, and the diskless computer is provided with a local hard disk for writing back e-sports game content and installing a network hard disk driver; the diskless server is provided with a storage cluster to provide users with e-sports game content, and to exchange data with the diskless computer through the communication module; the management module provides volume management services and instance management services, and is scheduled by the diskless server through a shared gateway interface, and the volume management service and instance management service respectively provide diskless storage services and diskless boot services.

[0015] Preferably, the instance management service provides a device discovery function for the diskless computer through a startup service. The diskless computer actively queries the startup service when starting up, and returns a diskless storage IO service to the diskless computer according to the MAC address of the diskless computer.

[0016] Preferably, the diskless storage service is used to manage the states of multiple IO processes on the diskless server, balance the IO processes of the diskless server, and collect the states of each IO service on the diskless server and persist them in a database.

[0017] Preferably, when an IO error occurs, the diskless server handles it by disconnecting, returning an error packet, generating a monitoring log, or manually intervening to investigate.

[0018] Implementing one of the above technical solutions of the present invention has the following advantages or beneficial effects:

[0019] In the present application, the diskless computer writes back data through the local hard disk, which can reduce the number of network communications between the diskless computer and the diskless server. At the same time, the network hard disk driver redirects the IO request to the local IO service, thereby achieving the minimum number of network communications between the diskless computer and the diskless server, thereby effectively reducing IO latency, effectively avoiding single point failures or single point performance bottlenecks of the diskless server, and facilitating the flexible addition of hot data cache to various links of cloud e-sports. Therefore, when the diskless computer uses the hard disk of the diskless server, it basically achieves almost the same performance as the local hard disk, is less prone to delays, effectively ensures the performance of the diskless computer, adapts to the usage needs of cloud e-sports scenarios, and improves the user experience. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. It is obvious that the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative work. In the drawings:

[0021] Figure 1 This is a flow chart of a cloud e-sports diskless storage method according to Embodiment 1 of the present invention;

[0022] Figure 2 This is a schematic diagram of the architecture of a cloud e-sports diskless storage system according to the second embodiment of the present invention. Figure 1 ;

[0023] Figure 3 This is a schematic diagram of the architecture of a cloud e-sports diskless storage system according to the second embodiment of the present invention. Figure 2 . DETAILED DESCRIPTION

[0024] In order to make the purpose, technical solutions and advantages of the present invention clearer, the various exemplary embodiments to be described below will refer to the corresponding drawings, which constitute a part of the exemplary embodiments, wherein various exemplary embodiments that may be used to implement the present invention are described. Unless otherwise indicated, the same numbers in different drawings represent the same or similar elements. The implementation methods described in the following exemplary embodiments do not represent all implementation methods consistent with the present disclosure. It should be understood that they are only examples of processes, methods, devices, etc. that are consistent with some aspects of the present disclosure as detailed in the attached claims, and other embodiments may also be used, or the embodiments listed herein may be modified in structure and function without departing from the scope and essence of the present invention.

[0025] In the description of the present invention, it should be understood that the terms "center", "longitudinal", "lateral", etc. indicate the orientation or positional relationship based on the drawings, which is only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the elements referred to must have a specific orientation, be constructed and operated in a specific orientation. The terms "first", "second", etc. are only used for descriptive purposes, and cannot be understood as indicating or implying relative importance or implicitly indicating the number of technical features indicated. The term "multiple" means two or more. The terms "connected" and "connected" should be understood in a broad sense, for example, it can be a fixed connection, a detachable connection, an integral connection, a mechanical connection, an electrical connection, a communication connection, a direct connection, an indirect connection through an intermediate medium, and can be the internal connection of two elements or the interaction relationship between two elements. The term "and / or" includes any and all combinations of one or more related listed items. For those of ordinary skill in the art, the specific meanings of the above terms in the present invention can be understood according to the specific circumstances.

[0026] In order to illustrate the technical solution of the present invention, a specific embodiment is used below for description, and only the parts related to the embodiment of the present invention are shown.

[0027] Embodiment 1:

[0028] like Figure 1As shown, the present invention provides a cloud e-sports diskless storage method, in which the diskless computer writes back the e-sports game content through the local hard disk. In the existing diskless system, diskless write-back is a cache strategy used to improve data writing efficiency. In the diskless system, the diskless computer saves all modifications in the cache instead of directly writing them into the system image. These modifications are then recorded and written into the system image of the server. In this embodiment, the diskless computer saves all modifications directly to the local hard disk, and no longer needs to frequently modify the data transmission interaction with the diskless server, which reduces the data writing requirements for the diskless server, thereby further improving the data write-back efficiency, and also reduces the read and write requirements for the diskless server, so that the diskless server is not prone to IO bottlenecks. The method comprises the following steps: S100: installing a local hard disk on a diskless computer and setting it as a non-system disk. The system disk of the diskless computer is on the diskless server. In order to ensure the successful rewriting of the e-sports game content, the capacity of the local hard disk is more than 300G, and an NVMe solid-state hard disk is preferably selected. The PCIe 4.0 version can write up to 6000MB / s, thereby effectively ensuring the extremely high data modification and rewriting speed of the diskless system, and achieving a better cloud e-sports game experience. S200: The diskless computer is started, that is, the diskless computer is turned on and has a network connection function. The IO service written by the diskless server provides the diskless computer with read and write capabilities, and performs the initial boot of the diskless computer. The initial boot method of the diskless computer can directly adopt the existing technology, thereby realizing the normal booting of the diskless computer. S300: The diskless computer is successfully started and enters the operating system. A local IO service is automatically started in the operating system, and the network hard disk driver is installed and configured on the local hard disk of the diskless computer. S400: The network hard disk driver redirects the IO request to the local IO service. The local IO service directly accesses the storage cluster of the diskless server through the network. The storage cluster is a storage array formed by multiple hard disks, so that the diskless server is not easily affected by single point failures. In this embodiment, the diskless computer writes back data through the local hard disk, which can reduce the number of network communications between the diskless computer and the diskless server. At the same time, the network hard disk driver redirects the IO request to the local IO service, so that the minimum number of network communications between the diskless computer and the diskless server can be achieved, thereby effectively reducing IO delays, effectively avoiding single point failures or single point performance bottlenecks of the diskless server, and facilitating the flexible addition of hot data caches in various links of cloud e-sports. Therefore, when the diskless computer uses the hard disk of the diskless server, it basically achieves almost the same performance as the local hard disk, and is not prone to delays, effectively ensuring the performance of the diskless computer, adapting to the use needs of cloud e-sports scenarios, and improving the use experience.

[0029] As an optional implementation, in step S300, the network hard disk driver is installed manually or automatically with administrator privileges. The administrator privileges ensure the success of the installation. Manual installation facilitates the handling of various unexpected situations encountered during the installation, and automatic installation facilitates the improvement of installation efficiency. The specific installation process can be selected according to needs. When the device name, driver date and inf (Device INFormation File) driver file in the device manager are consistent, it indicates that the network hard disk driver has been successfully installed. By verifying that the network hard disk driver has been successfully installed, it is convenient to ensure that the IO service request can be redirected to the local IO service in this embodiment. In step S300, after the network hard disk driver is successfully installed, the IP address of the Log Server IP is modified to the log transfer IP of the current area, and the sc query vfdisk command (you can enter the sc query vfdisk command by starting the cmd window with administrator privileges. If the state is running, it indicates that it is working normally) and the log upload query (such as restarting the computer, waiting for a few minutes to check the log upload of the diskless computer on the open source data analysis and visualization platform kibana, indicating that it is working normally) are used to confirm whether the local hard disk is configured correctly. By configuring the network hard disk driver, its normal operation is ensured, so that the IO service request can be redirected to the local IO service.

[0030] As an optional implementation, in step S400, the IO request is redirected to the local IO service through the IO transmission protocol. The format of the IO transmission protocol includes a common header, type definition, response return code definition, unknown type response packet, heartbeat packet and read packet. The above content ensures the integrity and normal operation of the IO transmission protocol. In the IO transmission protocol, all structures are 1-byte aligned, all fields are in little-endian byte order, and all requests and responses are defined from the diskless computer end (in this embodiment, the diskless server end response packet type is based on the client end request packet type, with the highest position 1. For example, the request packet type is 0x00100001, and the response packet type is 0x80100001). The diskless server end should return an unsupported response if the request packet format is correct but the type is not supported (for example, if the request packet type is 0x00700001 and the server end does not support it, then the response packet type is 0x80700001, and the response code is 0x2, indicating that it is not supported).

[0031] As an optional implementation, the storage cluster includes at least two ISCSI network hard disks or RBD network virtual hard disks, the specific number of which can be set according to the requirements, one of which is a system disk. ISCSI network hard disk is a technology that uses TCP / IP protocol to transmit SCSI commands on the network, which can realize the function of remote access to storage devices, and has the advantages of low cost, simple deployment and high flexibility; RBD (RADOS Block Devices) is a block storage device. RBD can provide a high-performance and infinitely scalable storage backend for virtualization technology and cloud services. RBD block storage is one of the commonly used storage types. RBD block devices can be mounted like hard disks, and have features such as snapshots, multiple copies, cloning and consistency. Data is stored in multiple OSDs (Object storage Daemons) of Ceph (distributed storage system) clusters in a striped manner. ISCSI network hard disks or RBD network virtual hard disks can be selected according to usage needs to improve the flexibility and adaptability of diskless systems. Therefore, the system disk can be distinguished from the hard disk that stores data such as game programs, so as to further enhance the gaming experience of users of diskless systems and improve the response speed of game operations.

[0032] The embodiment is only a specific example and does not represent only one implementation mode of the present invention.

[0033] Embodiment 2:

[0034] This embodiment provides a cloud gaming diskless storage system, which is used to run a cloud gaming diskless storage method in Embodiment 1, such as Figure 2 As shown, it includes several diskless computers (the specific scale and quantity are selected according to needs), diskless servers (multiple can be set according to needs to reduce the adverse effects of single point failures on cloud e-sports), communication modules and management modules. The diskless computer is connected to the interactive device for users to operate cloud e-sports games. The diskless computer is provided with a local hard disk for writing back e-sports game content and installing a network hard disk driver. Therefore, the diskless computer saves all modifications directly on the local hard disk. There is no need to frequently modify data transmission interaction with the diskless server, which reduces the demand for data writing to the diskless server, thereby further improving the efficiency of data writing, and also reduces the read and write requirements for the diskless server, so that the diskless server is not prone to IO bottlenecks. The diskless server is provided with a storage cluster (the storage cluster includes multiple hard disks, so that the diskless server is not easily affected by single point failures), provides users with e-sports game content, and exchanges data with the diskless computer through the communication module. The communication module preferably communicates through high-speed optical fiber of more than 10Gbps to ensure the cloud e-sports game experience. As shown Figure 3As shown, the management module provides volume management services and instance management services, and provides scheduling for the diskless server through a shared gateway interface, so that the diskless server can operate the volume management service and instance management service of the management module. The volume management service and the instance management service respectively provide diskless storage services (providing image mounting, image synchronization and IO gateway functions) and diskless boot services, thereby separating the diskless storage service and the diskless boot service in architecture, isolating the data required for diskless computer startup from the data for write-back operations, reducing the data interaction requirements for the diskless server, thereby reducing the hard disk working pressure of the diskless server, and making it less likely to have performance bottlenecks. In this embodiment, the diskless computer writes back data through the local hard disk, which can reduce the number of network communications between the diskless computer and the diskless server. At the same time, the network hard disk driver redirects the IO request to the local IO service, thereby achieving the minimum number of network communications between the diskless computer and the diskless server, thereby effectively reducing IO latency. The diskless storage service and the diskless boot service are separated in architecture, and the data required for the diskless computer to start up and the data for the write-back operation are isolated, thereby reducing the data interaction demand for the diskless server, thereby reducing the hard disk working pressure of the diskless server, effectively avoiding single point failures or single point performance bottlenecks of the diskless server, and facilitating the flexible addition of hot data caches in various links of cloud e-sports. Therefore, when the diskless computer uses the hard disk of the diskless server, it basically achieves almost the same performance as the local hard disk, is less prone to delays, effectively ensures the performance of the diskless computer, adapts to the usage needs of cloud e-sports scenarios, and improves the usage experience.

[0035] As an optional implementation method, the instance management service provides device discovery function for diskless computers by starting the service, thereby ensuring that new devices on the diskless computer can run smoothly, making it convenient for the diskless computer to add new interactive devices according to the operational needs of cloud e-sports, thereby improving the cloud e-sports experience. The diskless computer actively queries the startup service when it starts up, and returns the diskless storage IO service to the diskless computer based on the mac address of the diskless computer. Operations based on the mac address ensure that the diskless computer can correctly receive the diskless storage IO service.

[0036] As an optional implementation, the diskless storage service is used to manage the status of multiple IO processes on the diskless server, balance the IO processes on the diskless server, and collect the status of each IO service on the diskless server and persist it to the database. Since the IO service goal of diskless storage is to provide high-performance block IO to the maximum extent, a service is bound to a fixed hard disk or a fixed server, so it is necessary to manage multiple processes on a single server through the diskless storage service to manage the running status and load balancing of multiple IO processes. The diskless storage service is also used to collect the status of each IO service and persist it to the database, and provide a business interface for fast operation.

[0037] As an optional implementation, when an IO error occurs, the diskless server handles it by disconnecting (such as protocol parsing error, image ID does not exist, read data exceeds the above address, read back to the place where the write file has not been written), returning an error packet (such as hard disk read error, incomplete data read, etc.), generating a monitoring log (to facilitate the analysis of the specific cause of the IO error according to the monitoring log) or human intervention investigation (a diskless computer should not retry when receiving n IO error replies, and it is easy for the server system to be paralyzed due to a small number of problems. Human intervention investigation facilitates rapid analysis and troubleshooting of specific problems). The diskless server can effectively deal with various types of IO errors by handling in the above manner, thereby ensuring that the system can quickly resume normal operation and ensure that the user experience is not affected.

[0038] The above description is only the preferred embodiment of the present invention. It is known to those skilled in the art that various changes or equivalent substitutions may be made to these features and embodiments without departing from the spirit and scope of the present invention. In addition, under the teachings of the present invention, these features and embodiments may be modified to adapt to specific circumstances and materials without departing from the spirit and scope of the present invention. Therefore, the present invention is not limited to the specific embodiments disclosed herein, and all embodiments falling within the scope of the claims of this application belong to the protection scope of the present invention.

Claims

1. A cloud e-sports diskless storage method, characterized in that: The diskless computer writes back the e-sports game content through the local hard disk, including the following steps: S100: Install a local hard disk on the diskless computer and set it as a non-system disk; S200: The diskless computer is started, and the IO service written by the diskless server provides the diskless computer with read and write capabilities, and performs initial booting of the diskless computer; S300: The diskless computer starts successfully and enters the operating system. A local IO service is automatically started in the operating system, and the network hard disk driver is installed and configured on the local hard disk of the diskless computer. S400: The network hard disk driver redirects the IO service request to the local IO service, and the local IO service directly accesses the storage cluster of the diskless server through the network.

2. A cloud e-sports diskless storage method according to claim 1, characterized in that: In the step S300, the network hard disk driver is installed manually or automatically with administrator privileges. When the device name and driver date in the device manager are consistent with those in the inf driver file, it indicates that the network hard disk driver is successfully installed.

3. A cloud e-sports diskless storage method according to claim 2, characterized in that: In the step S300, when the network hard disk driver is successfully installed, the IP address of the Log Server IP is modified to the log transfer IP of the current area, and the sc query vfdisk command and log upload query are used to confirm whether the local hard disk is correctly configured.

4. The cloud e-sports diskless storage method according to claim 1, characterized in that: In the step S400, the IO request is redirected to the local IO service through the IO transmission protocol, and the format of the IO transmission protocol includes a common packet header, a type definition, a response return code definition, an unknown type response packet, a heartbeat packet and a read packet.

5. A cloud e-sports diskless storage method according to claim 4, characterized in that: In the IO transmission protocol, all structures are 1-byte aligned, all fields are in little-endian byte order, all requests and responses are defined from the diskless computer side, and the diskless server side should return an unsupported response if the request packet format is correct but the type is not supported.

6. A cloud e-sports diskless storage method according to any one of claims 1 to 5, characterized in that: The storage cluster includes at least two ISCSI network hard disks or RBD network virtual hard disks, one of which is a system disk.

7. A cloud e-sports diskless storage system, characterized in that: A cloud e-sports diskless storage method for running any one of claims 1-6, comprising a plurality of diskless computers, a diskless server, a communication module and a management module; the diskless computer is connected to an interactive device for users to perform cloud e-sports game operations, and the diskless computer is provided with a local hard disk for writing back e-sports game content and installing a network hard disk driver; the diskless server is provided with a storage cluster to provide users with e-sports game content, and to perform data exchange with the diskless computer through the communication module; the management module provides volume management services and instance management services, and is scheduled by the diskless server through a common gateway interface, and the volume management service and instance management service respectively provide diskless storage services and diskless boot services.

8. The cloud e-sports diskless storage system according to claim 7, characterized in that: The instance management service provides a device discovery function for the diskless computer through a startup service. The diskless computer actively queries the startup service when it is started, and returns a diskless storage IO service to the diskless computer according to the MAC address of the diskless computer.

9. The cloud e-sports diskless storage system according to claim 7, characterized in that: The diskless storage service is used to manage the states of multiple IO processes on the diskless server, balance the IO processes of the diskless server, and collect the states of each IO service on the diskless server and persist them in a database.

10. The cloud e-sports diskless storage system according to claim 7, characterized in that: When an IO error occurs, the diskless server handles it by disconnecting, returning an error packet, generating a monitoring log, or manually intervening to troubleshoot.

Images