WO2019084853A1 - 裸金属服务器的硬件属性发现方法和系统 - Google Patents

裸金属服务器的硬件属性发现方法和系统 Download PDF

Info

Publication number
WO2019084853A1
WO2019084853A1 PCT/CN2017/108949 CN2017108949W WO2019084853A1 WO 2019084853 A1 WO2019084853 A1 WO 2019084853A1 CN 2017108949 W CN2017108949 W CN 2017108949W WO 2019084853 A1 WO2019084853 A1 WO 2019084853A1
Authority
WO
WIPO (PCT)
Prior art keywords
bare metal
server
metal server
manager
information collection
Prior art date
Application number
PCT/CN2017/108949
Other languages
English (en)
French (fr)
Inventor
沈培
沈钢纲
黄彪
雷舒莹
Original Assignee
华为技术有限公司
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 华为技术有限公司 filed Critical 华为技术有限公司
Priority to CN201780089457.6A priority Critical patent/CN110495136B/zh
Priority to PCT/CN2017/108949 priority patent/WO2019084853A1/zh
Priority to EP17930891.1A priority patent/EP3684009A4/en
Publication of WO2019084853A1 publication Critical patent/WO2019084853A1/zh
Priority to US16/860,938 priority patent/US20200257519A1/en

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F8/00—Arrangements for software engineering
    • G06F8/60—Software deployment
    • G06F8/65—Updates
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F9/00—Arrangements for program control, e.g. control units
    • G06F9/06—Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
    • G06F9/44—Arrangements for executing specific programs
    • G06F9/455—Emulation; Interpretation; Software simulation, e.g. virtualisation or emulation of application or operating system execution engines
    • G06F9/45533—Hypervisors; Virtual machine monitors
    • G06F9/45541—Bare-metal, i.e. hypervisor runs directly on hardware
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F8/00—Arrangements for software engineering
    • G06F8/60—Software deployment
    • G06F8/61—Installation
    • G06F8/63—Image based installation; Cloning; Build to order
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F9/00—Arrangements for program control, e.g. control units
    • G06F9/06—Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
    • G06F9/44—Arrangements for executing specific programs
    • G06F9/4401—Bootstrapping
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F9/00—Arrangements for program control, e.g. control units
    • G06F9/06—Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
    • G06F9/44—Arrangements for executing specific programs
    • G06F9/4401—Bootstrapping
    • G06F9/4406—Loading of operating system
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F9/00—Arrangements for program control, e.g. control units
    • G06F9/06—Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
    • G06F9/44—Arrangements for executing specific programs
    • G06F9/4401—Bootstrapping
    • G06F9/4416—Network booting; Remote initial program loading [RIPL]
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F9/00—Arrangements for program control, e.g. control units
    • G06F9/06—Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
    • G06F9/44—Arrangements for executing specific programs
    • G06F9/455—Emulation; Interpretation; Software simulation, e.g. virtualisation or emulation of application or operating system execution engines
    • G06F9/45533—Hypervisors; Virtual machine monitors
    • G06F9/45558—Hypervisor-specific management and integration aspects
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L41/00—Arrangements for maintenance, administration or management of data switching networks, e.g. of packet switching networks
    • H04L41/08—Configuration management of networks or network elements
    • H04L41/0803—Configuration setting
    • H04L41/0813—Configuration setting characterised by the conditions triggering a change of settings
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F9/00—Arrangements for program control, e.g. control units
    • G06F9/06—Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
    • G06F9/44—Arrangements for executing specific programs
    • G06F9/455—Emulation; Interpretation; Software simulation, e.g. virtualisation or emulation of application or operating system execution engines
    • G06F9/45533—Hypervisors; Virtual machine monitors
    • G06F9/45558—Hypervisor-specific management and integration aspects
    • G06F2009/45575—Starting, stopping, suspending or resuming virtual machine instances

Definitions

  • the present invention relates generally to information technology (IT), and more particularly to a method and system for discovering hardware attributes of a bare metal server.
  • Server virtualization technology refers to the technology that can run multiple virtual servers on a single physical server, and the above virtual servers are almost indistinguishable from physical servers in terms of users, applications, and even operating systems. Users can flexibly install any software on a virtual server.
  • the virtualized server is called a virtual machine (English: virtual machine, VM).
  • the operating system running in the virtual machine is called the guest operating system (English: guest OS).
  • the software responsible for managing virtual machines is called virtual machine monitor (VMM), and can also be called manager (English: hypervisor).
  • a system that provides services using server virtualization technology can be referred to as a cloud computing system.
  • the physical infrastructure in a cloud computing system includes various physical devices for providing computing, storage, or network resources.
  • the bare metal server is one of the physical infrastructure of the cloud computing system.
  • the cloud computing system needs to solve the problem of setting up the basic environment, such as solving the problems of adding, deleting, power management, and operating system deployment of the bare metal server.
  • the life cycle of the bare metal server can be managed by using dedicated software in the cloud computing system.
  • Openic OpenStack is an open source cloud computing management platform project
  • the Ironic component provides an application programming interface (API) for managing bare metal servers.
  • the Ironic component provides APIs to manage bare metal servers.
  • bare metal server management software collects the hardware properties of bare metal servers before the bare metal server is issued (English: provision).
  • the hardware attributes can be hard disk information of the bare metal server (such as the size of the hard disk storage space and / or hard disk type), memory information, central processing unit (CPU) information (such as the number of cores of the CPU and / or the frequency of each core), Network card information or any combination thereof.
  • CPU central processing unit
  • the management software of the bare metal server issues the resources in the bare metal server to the user.
  • the bare metal server can be directly distributed to users. For example, services can be directly provided for high-intensity computing or high-time-demanding services. Virtualized resources of bare metal servers can also be virtualized and then distributed to users' virtualized resources.
  • the hardware management software of the bare metal server needs to start the operating system on the bare metal server through the network startup mode, and then obtain the relevant hardware attributes through the application in the operating system. This method takes a lot of time and cannot meet the requirements of the bare metal manager to quickly issue bare metal servers.
  • This paper proposes various technical solutions to collect the hardware properties of the bare metal server without relying on the operating system, to meet the needs of the bare metal manager to quickly issue bare metal servers.
  • the present application provides a cloud computing system for discovering hardware attributes of a bare metal server, the bare metal manager including a bare metal manager and a bare metal server for powering up a bare metal server, such as Controlling the booting system of the bare metal server by a startup command, the booting system controls the power module to the bare gold
  • the server is powered on; after the bare metal server is powered on, the booting system in the bare metal server, such as the BIOS, is started; the booting system is used to run the pre-boot execution environment PXE client and boot through the PXE boot mode.
  • An operating system is started, and an information collecting module is run before the operating system is started; the information collecting module is configured to collect hardware attributes of the bare metal server, and send the hardware attribute of the bare metal server to the bare metal through a network. Manager.
  • the information collection module in the bare metal server runs in the pre-boot phase of the operating system in the boot system operating environment, so that the discovery of the hardware attribute can be independent of the operating system, and the bare metal server can be completed without starting the operating system.
  • the acquisition of hardware attributes is quickly sent to the bare metal manager, which greatly reduces the time required for the bare metal manager to obtain the hardware properties of the bare metal server, and improves the deployment efficiency of the bare metal server, especially for large data centers.
  • the situation of the bare metal server greatly improves the deployment efficiency.
  • the information collection module is small, the operation and download do not need to consume a large amount of network traffic, which reduces network congestion.
  • the implementation of the above method to the cloud computing system is small, and the development cost is relatively small.
  • a possible implementation manner is that the bare metal manager is further used to enable the PXE boot mode of the bare metal server.
  • This mode is used in the scenario where the operating system is not set up on the bare metal server or the operating system is installed on the bare metal server.
  • the operating system startup mode is non-PXE startup mode. In the above two scenarios, the operating system startup mode can be switched to the PXE startup mode.
  • a further possible implementation manner is that the guiding system includes an information collecting interface, and the information collecting module is configured to invoke the information collecting interface to obtain hardware attributes of the bare metal server.
  • the booting system needs to be extended, so that the booting system adds an information collecting interface, and the information collecting module calls the information collecting interface to collect hardware attributes. Since the booting system has acquired hardware attributes in the initialization phase, the corresponding information is directly added to the booting system. Interface to collect hardware properties for easy implementation.
  • another possible implementation manner is to add an information setting interface in the booting system, and the bare metal manager is further configured to send a hardware attribute modification request to the bare metal server through the network, where the information collecting module further And configured to modify the hardware attribute of the bare metal server by calling the information setting interface according to the hardware attribute modification request.
  • the implementation further provides the possibility that the bare metal manager can modify the hardware attributes of the bare metal server before the operating system of the bare metal server is started, so that the bare metal manager can manage the bare metal server more flexibly.
  • a further possible implementation manner is that the information collection module is pre-stored in a bare metal manager, and the PXE client in the bare metal server is configured to acquire an IP address allocated for the bare metal server, using the The IP address communicates with the bare metal manager and the information collection module is downloaded from the bare metal manager.
  • the bare metal manager provides an information collecting module, that is, in a specific implementation, the bare metal manager is extended in the bare metal manager to enable the bare metal manager to provide information collection. Module.
  • the bare metal server obtains the information collection module without an additional process, but obtains the information collection module in the PXE startup process.
  • an optional specific implementation manner is that the information collection module is stored in the FTP server instead of the PXE booting boot module, and the bare metal manager notifies the address of the FTP server to the bare metal server.
  • a further possible implementation manner is that the information collecting module is pre-stored in the bare metal server, and the guiding system of the bare metal server loads and runs the letter before the operating system starts.
  • Information collection module is not on the bare metal manager side, but is stored in the bare metal server.
  • the information collection module may not be downloaded from the remote end before the operating system is loaded.
  • the information collection module can be loaded and executed, which further saves execution time and improves the efficiency of the bare metal manager to discover the hardware properties of the bare metal server.
  • an optional specific implementation manner is that the bare metal server further includes an information collecting switch, where the information collecting switch is configured to indicate whether the guiding system runs the information collecting module; if the information collecting switch Instructing to run the information collection module, the guidance system in the bare metal server is used to load the information collection module.
  • the boot system runs the information collection module according to the indication of the information collection switch .
  • a further possible implementation manner is that the information collecting module in the bare metal server carries the hardware attributes of the bare metal server by using a dynamic host setting protocol DHCP message or a representative state transmission REST message.
  • the present application provides a hardware attribute discovery method for a bare metal server, which is applied to a bare metal manager, where the bare metal manager manages a bare metal server; the method includes: powering up the bare metal server And receiving a hardware attribute of the bare metal server sent by the bare metal server through a network, wherein the booting system in the bare metal server is used to run a pre-boot execution environment PXE client and boot operation through a PXE boot mode; The system is started, and an information collection module is run before the operating system of the bare metal server is started, and the information collection module is configured to collect and send hardware attributes of the bare metal server through a network.
  • the method of the foregoing second aspect is applied to the bare metal manager.
  • the specific implementation of the foregoing second aspect reference may be made to the specific implementation part of the bare metal manager in the foregoing first aspect, and details are not described herein again.
  • the application provides a management apparatus for managing a bare metal server;
  • the management apparatus includes: a storage device for storing instructions; and at least one processor coupled to the storage device;
  • the processor in the at least one server executes the instruction, the instruction causes the processor to perform the method of the second aspect above.
  • the application provides a bare metal manager for managing a bare metal server, the manager including: a startup management module, configured to power on the bare metal server; and a discovery module, configured to: Receiving hardware attributes of the bare metal server sent by the bare metal server through a network, wherein a booting system in the bare metal server is configured to run a PXE client and boot an operating system boot by a PXE boot mode, and in the Before the operating system of the bare metal server starts, an information collecting module is run, and the information collecting module is configured to collect and send the hardware attributes of the bare metal server through a network.
  • a startup management module configured to power on the bare metal server
  • a discovery module configured to: Receiving hardware attributes of the bare metal server sent by the bare metal server through a network, wherein a booting system in the bare metal server is configured to run a PXE client and boot an operating system boot by a PXE boot mode, and in the Before the operating system of the bare metal server starts,
  • a possible implementation manner is that the startup management module is further configured to enable the PXE startup mode of the bare metal server.
  • a further possible implementation manner is that the discovery module is further configured to send a hardware attribute modification request to the bare metal server through a network.
  • a further possible implementation manner is that the discovery module is further configured to save the information collection module, and provide the information collection module to the bare metal server through a network.
  • a further possible implementation manner is that the discovery module is specifically configured to receive the bare gold a dynamic host setting protocol DHCP protocol message sent by the server, the hardware attribute of the bare metal server carried in the DHCP protocol message is obtained by the DHCP server in the discovery module; or the discovery module is specifically configured to receive The representative state of the bare metal server transmits a REST protocol message, and the hardware module of the bare metal server carried in the REST protocol message is obtained by the parsing module in the discovery module.
  • the present application provides a hardware attribute discovery method for a bare metal server, which is characterized in that it is applied to a bare metal server, and a bare metal manager manages a bare metal server; the method includes: receiving the bare metal management The boot command of the device is powered on according to the boot command; after the bare metal server is powered on, the boot system in the bare metal server is started, and the boot system boots through the pre-boot execution environment PXE boot mode. The operating system is booted; and before the operating system is started, the hardware attributes of the bare metal server are collected, and the hardware attributes of the bare metal server are sent to the bare metal manager through the network.
  • a possible implementation manner is that before the collecting the hardware attributes of the bare metal server, the method further includes: acquiring an IP address allocated for the bare metal server, using the IP address and the bare The metal manager communicates and downloads an information collection module from the bare metal manager and runs the information collection module in the guidance system.
  • a further possible implementation manner is that before the collecting the hardware attributes of the bare metal server, the method further includes: the booting system loading from the bare metal server in a pre-boot phase of the operating system And run the information collection module.
  • a further possible implementation manner is that the information collecting module invokes an information collecting interface in the guiding system to acquire hardware attributes of the bare metal server.
  • a further possible implementation manner is that the method further includes: receiving a hardware attribute modification request sent by the bare metal manager, according to the hardware attribute modification request, the information collection module calls the The information setting interface in the bare metal server modifies the hardware attributes of the bare metal server.
  • the application provides a bare metal server, and the bare metal manager manages the bare metal server;
  • the server includes: a boot controller, a central processing unit, a first internal memory, a second internal memory, and a network card
  • the boot controller is configured to receive a boot command of the bare metal manager, perform a power-on operation according to the boot command, and save a boot system in the second internal memory, after the bare metal server is powered on
  • the boot system is loaded into the first internal memory, the central processor runs the boot system; the boot system is configured to run the PXE client and boot an operating system boot by a PXE boot mode, and Running the information collection module before the operating system is started; the information collection module is configured to collect hardware attributes of the bare metal server, and send the hardware attributes of the bare metal server to the bare metal management through the network card Device.
  • the foregoing sixth aspect is applied to the bare metal server.
  • the application provides a computer program product, the computer program product being stored on a readable storage medium, the computer program product comprising instructions running in a running environment of a booting system in a computer, and Being executed before the operating system of the computer is started, the instructions causing the processor to: invoke an information collection interface in the boot system, collect hardware attribute information of the bare metal server; and The hardware attribute information of the bare metal server is sent to the bare metal manager.
  • the application provides a computer system including a manager and a server; the server includes a booting system on the server After booting, the booting system is started; the booting system is configured to run a pre-boot execution environment PXE client and boot an operating system boot by a PXE boot mode, and run an information collecting module before the operating system starts; An information collecting module, configured to collect hardware attributes of the server, and send the hardware attributes of the server to the manager through a network; the manager is configured to receive hardware attributes of the server sent by the server.
  • the computer system provided by the foregoing eighth aspect may be the cloud computing system provided by the foregoing first aspect, or may be a traditional data center system, and the manager may include a bare metal manager, or may include the function of implementing the seventh aspect of the present application.
  • the server may include a bare metal server, or may include any other form of physical server that implements the functions of the seventh aspect of the present application.
  • the specific implementation manner of the foregoing eighth aspect may refer to the specific implementation manner of the foregoing first aspect.
  • 1A is a simplified block diagram of a cloud computing system
  • 1B is another simplified block diagram of a cloud computing system
  • 2A is a schematic diagram of stages of obtaining hardware attributes of a bare metal server of a conventional solution
  • 2B is a schematic flowchart of obtaining hardware attributes of a bare metal server according to the present invention
  • 3A is a schematic diagram of stages of obtaining hardware attributes of a bare metal server
  • 3B is a schematic diagram showing the relationship between each stage of the conventional solution, each stage of the solution of the present invention, and each stage of the bare metal server startup;
  • FIG. 4 is a block diagram showing the composition of a bare metal server in the solution of the present application.
  • 5A is a simplified block diagram of a computer system composed of a bare metal manager and a bare metal server in an embodiment of the present application;
  • 5B is a schematic flowchart of acquiring hardware attributes of a bare metal server by a bare metal manager according to an embodiment of the present application
  • 6A is a simplified block diagram of a computer system composed of a bare metal manager and a bare metal server in still another embodiment of the present application;
  • 6B is a schematic flowchart of acquiring hardware attributes of a bare metal server by a bare metal manager according to still another embodiment of the present application;
  • a bare metal server one of the physical facilities in a cloud computing system, also referred to as a bare metal node or a physical machine of a "naked” operating system, generally refers to a server that does not have an operating system or virtualization software installed.
  • a physical machine of a "naked” operating system may also refer to a physical machine that has an operating system installed but does not have virtualization software installed.
  • the present application uses the term "bare metal server” to refer to a cloud computing system.
  • the bare metal server referred to in this application may have multiple names, such as “to be issued server” and “to be deployed server”, which are not enumerated here.
  • Bare metal manager for managing bare metal servers also known as the management component of bare metal servers, where the term "management component” or “manager” may refer to software, firmware, middleware. , microcode, hardware, and/or any combination thereof.
  • the management component or manager can be any one or combination of modules, processes, threads running on a general purpose server, or the general purpose server itself.
  • the management component or manager can be one or more, and the management component or manager can be located on one general server or distributed on two or more general purpose servers.
  • the management component or manager can be obtained and executed from a computer readable medium having various data structures stored therein, or can be implemented by various logical combinations of hardware circuits.
  • the bare metal manager is a term selected based on the clear uniformity of the expression in the patent application.
  • the bare metal manager can have multiple implementation modes. Therefore, there may be various expressions, such as a controller of a bare metal server, a control device of a bare metal server, an Ironic component, or a component of another name.
  • the above-described general purpose server is a physical server including a processor and a storage device known to those skilled in the art, the storage device is for storing instructions, and the processor is for executing instructions.
  • PXE boot mode is the current mainstream diskless boot technology, which enables the computer to boot through the network instead of from the local hard disk, optical drive and other devices, using the network mode of Client/Server.
  • the terminal requires the server to assign an IP address, and then uses a file transfer protocol (FTP).
  • the file transfer protocol may include a trivial file transfer protocol (TFTP) or a multicast simple file transfer protocol (MTFTP).
  • TFTP trivial file transfer protocol
  • MTFTP multicast simple file transfer protocol
  • the multicast trivial file transfer protocol downloads a boot package to the local memory for execution.
  • the startup software package completes the terminal basic software settings to boot the terminal operating system pre-installed on the server.
  • the PXE booting mode is the way in which the network is booted.
  • the bare metal server executes the PXE client, and the bare metal manager executes the PXE server.
  • the PXE server you need to support the DHCP function and TFTP function; DHCP (Dynamic Host Configuration Protocol), dynamic host setting protocol, PXE server supports the DHCP function to assign an IP address to the PXE client, the IP address is used for the PXE client.
  • the PXE server communicates with the PXE server.
  • the PXE server supports the TFTP function to transfer files to the PXE client through the TFTP protocol, such as an operating system-related boot module or an image file of the operating system.
  • the server in the present application has a distinction between hardware and software.
  • the software server refers to computer software that manages resources and provides corresponding services.
  • the server corresponding to the client refers to a software server, for example, The PXE server, the DHCP server, and the TFTP server
  • the hardware server refers to a physical computing device different from the personal computer form.
  • the bare metal server, the configured server, and the general server are hardware servers.
  • the booting system is used to manage the startup of the bare metal server, as well as the control and driving of important hardware, and provides the base layer call for the high-level software.
  • the module can be saved with the basic input/output system of the bare metal server.
  • the booting system may be in the form of a firmware written in a read-only memory in the bare metal server.
  • the booting system may be a Basic Input Output System (BIOS), and specifically, may be a BIOS using UEFI or Use a non-UEFI BIOS.
  • BIOS Basic Input Output System
  • UEFI Unified Extensible Firmware Interface
  • a unified and extensible firmware interface is a new firmware standard that is programmable and extensible.
  • the server power-on in this application refers to the state that the server has been powered on from the state that it has never been powered on, instead of the state that the server has not been powered on but has not been powered on.
  • the configured server 11 (1 shown in the figure, the actual number may be 1 or more) in the cloud computing system is a server that has been powered on and installed with an operating system or virtualization software, and includes hardware.
  • the platform 110 and the hardware platform 110 may include a hard disk, a central processing unit, a network interface card, a memory, and the like in a physical structure of the server.
  • the software platform 112 can be run on the hardware platform 110.
  • the software platform 112 can include software such as virtualization software, an operating system of the virtual machine, and application software of the virtual machine.
  • Virtualization software (for example, Hypervisor) is the middle layer between physical hardware and virtual machines, and is responsible for coordinating the access of each virtual machine to the hardware platform.
  • the Hypervisor is also called Virtual Machine Monitor (VMM).
  • the VMM includes a virtual hardware platform for implementing virtual machines.
  • the virtual hardware platform includes virtual memory, central processing unit, hard disk, graphics card and network interface card.
  • a plurality of virtual machines VM1 to VMn are run on top of the VMM, and the virtual machine's software system runs on the virtual hardware platform of the virtual machine monitor.
  • the system of FIG. 1A includes a bare metal server 12 (one shown in the figure, and the actual number may be one or more) in addition to the configured server 11.
  • a bare metal server to join a cloud computing system, the cloud computing system needs to solve the problem of setting up the basic environment, such as solving the problems of adding, deleting, power management, and operating system deployment of the bare metal server.
  • a dedicated bare metal manager 201 can be specifically deployed to manage the life cycle of the bare metal server.
  • the Ironic component is provided in OpenStack, and the Ironic component provides a The series manages the physical machine's API, which can manage the physical machine of the "naked" operating system, including the management of the entire life cycle from the physical machine power-on installation of the operating system to the physical machine power-off maintenance.
  • the cloud computing control platform 20 can be deployed on a single server or on multiple servers.
  • FIG. 1B illustrates another composition of the cloud computing system.
  • the cloud computing control platform 20 is implemented by a virtual machine in the configured server 11, that is, the cloud computing control platform is not separate. It is deployed by a physical machine, but runs in a virtual machine in the configured server 11 (the illustration is deployed on the virtual machine VMn, or can be actually deployed in multiple virtual machines), and the virtual machine fulfills the cloud computing control platform. 20 duties.
  • the cloud computing control platform 20 may include other components, such as a virtual machine management component, a security component, or a billing component or an energy consumption component.
  • the bare metal servers described above in FIG. 1A and FIG. 1B are all implemented in a cloud computing system (cloud data center).
  • the bare metal server in this patent application can also be deployed in a traditional computer system (for example, a traditional data center).
  • the bare metal manager and the bare metal server may be respectively referred to as a manager and a server; in addition, the present patent application is also applicable to an initial scenario of a cloud computing system, for example, a scenario in which all servers in a cloud computing system are bare metal servers. .
  • the technical solution of the present application is applicable to both the scenario of the cloud computing system and the scenario of the conventional computer system.
  • the system to which the technical solution of the present application is applicable is a computer system.
  • the bare metal manager 201 needs to obtain the hardware attributes of the bare metal server before the bare metal server is issued.
  • the hardware attributes may be hard disk attributes (such as size or type) of the bare metal server, memory attributes, CPU attributes (architecture, The number of cards, the number of cores or the frequency, or the attributes of the network card (MAC address), the Baseboard Management Controller (BMC), or any combination thereof.
  • the above hardware attributes are not exhaustive of the hardware attributes of all bare metal servers. Those skilled in the art can extend the hardware attributes of other bare metal servers other than the above examples of hardware attributes. After the bare metal manager 201 obtains the hardware attributes, the bare metal server can be operated accordingly.
  • the cloud computing system issues a bare metal server
  • the hardware resources of the bare metal server can be utilized for computing virtualization or storage virtualization.
  • Bare metal servers can also be used to directly provide services for high-intensity computing or other services with high real-time requirements.
  • the process of acquiring the hardware attributes of the bare metal server by the bare metal manager is referred to as the discovery process of the hardware properties of the bare metal server
  • the discovery process of the hardware properties of the bare metal server may include how the bare metal manager triggers the discovery process
  • the process of finally obtaining hardware attributes can also include the process of how the bare metal server collects hardware attributes and reports hardware attributes to the bare metal manager.
  • the method in which the conventional bare metal manager 201 finds the hardware properties of the bare metal server includes five stages.
  • the first stage as in step 201, the bare metal manager bare metal server is powered on at time T0, the bare metal manager 201 enables the PXE boot mode of the bare metal server 12, and the boot system in the bare metal server starts and runs the network card.
  • the PXE client next, in steps 202-205, the PXE client of the bare metal server 12 performs an IP address allocation process with the bare metal manager 201 (the bare metal manager 201 has the function of the DHCP server), including DHCP discovery, DHCP.
  • the bare metal manager 201 acts as a DHCP server to provide the bare metal server with an IP address and a download address of the file to be downloaded (for example, a TFTP server address), and then, in step 206, the bare metal server 12 Download the PXE boot boot module (such as the PXE boot loader) from the bare metal manager at time T1.
  • the PXE boot boot module such as the PXE boot loader
  • the second stage since the PXE boot boot module is started, the bare metal server 12 needs to perform a second IP address allocation process, such as steps 207-210, the bare metal server obtains a new IP address, and then in steps 211-212, bare metal The server sends a file acquisition request to the bare metal manager again (which may be a TFTP request sent according to the TFTP server address), and then the bare metal server acquires an image of the operating system at time T2.
  • the third stage after obtaining the operating system image, start the operating system on the bare metal server 12 (generally, the operating system started is a streamlined Linux operating system), and then launch a special application in the system. Module, this application module starts running at time T3.
  • the fourth stage after the special application module runs, it starts to collect the hardware attributes of the bare metal service period, and acquires the hardware attributes at time T4, as shown in step 213.
  • the fifth stage as in step 214, the bare metal server 12 transmits the acquired hardware attributes to the bare metal management 201 through the network, and the bare metal manager 201 receives the hardware attributes sent by the bare metal server 12 at time T5.
  • the above method has problems of low efficiency and poor portability.
  • the operation of the operating system itself takes time.
  • the hardware attributes of multiple bare metal servers are collected in batches, downloading the operating system image over the network will result in bandwidth usage and increased latency.
  • the extra operating system used to collect hardware attributes is currently only capable of supporting X86 type bare metal servers but not ARM type bare metal servers, which leads to poor portability.
  • the present patent application is based on the above problems of the prior art, and provides a system and method for discovering hardware attributes of a bare metal server without relying on an operating system.
  • the present patent application utilizes a booting system in a bare metal server to run an information gathering module on the booting system after the booting system in the bare metal server is booted and before the booting system boots the operating system, the information collecting module collecting bare metal The hardware properties of the server and send the collected hardware properties to the bare metal manager over the network.
  • the information collection module is a new functional module provided for the application.
  • the module runs in the pre-loading stage of the operating system after the booting system is started, and is used to implement the discovery of the hardware properties of the bare metal server.
  • the information collection module can be a small application or an applet, such as an information collection application or an information collection program.
  • the boot system running environment includes the following stages: first, the initialization phase of the bare metal server, followed by the pre-boot phase of the operating system (also known as the operating system before loading). Phase), and finally the operating system startup phase (after the operating system startup is completed, the boot system is no longer running), the information collection module in this application runs in the pre-boot phase of the operating system in the boot system runtime environment.
  • the bare metal manager 201 collects the hardware attributes of the bare metal server and includes four stages. The first stage: at time T0, the bare metal server 12 is powered on, the boot system in the bare metal server 12 starts to boot, and the bare metal manager 201 enables the PXE boot mode of the bare metal server 12; then the bare metal server at time T1 12 Running information collection module (The bare metal server of the present application can load and run the information collection module from the outside, and can also load and run the information collection module from the internal, and the subsequent embodiments will be specifically described).
  • the second stage the information collection module startup phase, in the T2 time information collection module runs in the operating environment of the boot system in the bare metal server 12 (this phase is the pre-boot phase of the operating system in the boot system operating environment).
  • the third stage the information collection module invokes the information collection interface in the boot system to obtain the hardware attributes of the bare metal server 12.
  • the information collection module acquires the hardware attributes of the bare metal server 12.
  • the fourth stage the reporting stage of the hardware attribute of the bare metal server 12, the information collecting module transfers the collected hardware attributes of the bare metal server 12 to the bare metal manager 201 through the network, and the bare metal manager 201 acquires the data at the time T4. Hardware properties.
  • the bare metal server 12 Before performing the fourth stage of transmitting hardware attributes, the bare metal server 12 also needs to complete the process of IP address acquisition, in which the bare metal server 12 sends a DHCP request to obtain an IP address, and the DHCP server in the bare metal manager 201 responds to the DHCP.
  • the bare metal server 12 is requested and assigned an IP address.
  • FIG. 3B illustrates the relationship between the process of discovering the hardware attributes of the bare metal server and the stages after the bare metal server is powered on, and also shows the process of discovering the hardware attributes of the bare metal server and the bare metal server in the prior art. The relationship between the various stages after the electricity.
  • the solution information collection module provided by the present application runs in the pre-boot phase of the operating system of the boot system, so that the discovery of the hardware attributes can be independent of the operating system, and the bare metal server can be completed without starting the operating system.
  • the acquisition of hardware attributes is quickly sent to the bare metal manager, which greatly reduces the time required for the bare metal manager to obtain the hardware properties of the bare metal server, and improves the deployment efficiency of the bare metal server, especially for large data centers.
  • the situation of the bare metal server greatly improves the deployment efficiency.
  • the information collection module is small, the operation and download do not need to consume a large amount of network traffic, which reduces network congestion.
  • the implementation of the above method by the booting system has little improvement on the system, and the development cost is relatively small.
  • the prior art method can only support the bare metal server of the X86 type, and cannot support the problem of the bare metal server of the Advanced RISC Machine (ARM) type, because the prior art is used for the installation specific
  • the operating system of the software cannot be applied to bare metal servers of the ARM type.
  • the method of the present application does not rely on the operating system to complete the discovery of hardware attributes, solves the above problems, and improves the versatility of the discovery manner of the bare metal service hardware attributes.
  • the processor 401 may include a CPU or a plurality of CPUs.
  • the first internal memory 402 may be a random access memory RAM for reading commands after the CPU is started. And make the CPU execute the instructions.
  • the second internal memory 403 can be a read-only memory ROM or a flash, wherein firmware can be set on the ROM, and the firmware is a module written in the ROM, generally serving as the basis of the system and the underlying hardware setting and control work. .
  • the booting system may be stored in the second internal memory 403.
  • the booting system may be a BIOS using UEFI or a BIOS not using UEFI.
  • the booting system is loaded into the first internal memory after the bare metal server is powered on.
  • the processor 401 is caused to run the boot system to boot the operating system of the bare metal server.
  • the NIC 404 is configured with a PXE client.
  • the PXE client is actually a client module located in the ROM of the NIC. This application is based on the PXE client in the NIC.
  • the PXE client can also be integrated in the field.
  • the PXE client is also loaded into the first internal memory 402 so that the processor 401 runs the PXE client, and after the PXE client is loaded into the memory, it At the same time, with the function of the DHCP client and the TFTP client, the DHCP client requests the DHCP server to allocate the IP address to the bare metal server, and then the PXE client downloads the remote file to the local operation through the TFTP;
  • the memory 405 is used to save data. Before the bare metal server is released, the external memory 405 generally does not store data.
  • the startup controller 406 is configured to communicate with the bare metal management system through an Intelligent Platform Management Interface (IPMI) to receive a startup command of the bare metal management system, and the startup command is used to power on the bare metal server to make the bare metal
  • IPMI Intelligent Platform Management Interface
  • the startup command can also be used to enable the PXE boot mode of the bare metal server.
  • a specific implementation of the boot controller 406 is a Baseboard Management Controller (BMC).
  • the basic management controller is generally used to measure internal physical variables of a bare metal server, such as temperature, humidity, power supply voltage, and fan speed. Communication parameters and operating system functions, if any of these variables is outside the specified limits, the BMC issues an alert notification.
  • the BMC can also be used to implement the functions of the above-mentioned boot controller.
  • the setting of the PXE boot mode of the bare metal server is generally enabled by the bare metal manager in the cloud computing scenario.
  • the manual mode of the bare metal server can be manually set or PXE boot mode. Activated by default.
  • the bare metal server is set to the PXE boot mode by default.
  • the PXE boot mode of the bare metal server is activated by default after the boot system is started. At this time, the bare metal manager does not need to enable the PXE boot mode.
  • the bare metal manager 201 enables the PXE boot mode of the bare metal server.
  • the bare metal manager can remotely set the boot mode of the bare metal server to the PXE boot mode, or the non-PXE boot that is set by default when the bare metal server is shipped from the factory.
  • the mode is changed to the PXE startup mode.
  • the bare metal manager can enable the PXE boot mode of the bare metal server through the baseboard management controller. If the bare metal server does not have a pre-installed operating system, the bare metal manager can directly set the boot mode of the bare metal server to the PXE boot mode. If the bare metal server has the operating system pre-installed, the bare metal manager enables the bare metal.
  • the PXE startup mode of the server may include: changing the operating system startup mode of the bare metal server from a hard disk/optical disk startup to a PXE startup. Further, the bare metal manager may also restart the bare metal server after changing the startup mode.
  • the bare metal server On the bare metal server, the bare metal server is controlled by the bare metal manager in the cloud computing scenario. However, in some special scenarios, it is also possible to manually power up the bare metal server. The discovery method of the bare metal hardware attribute of the present application can still be performed.
  • the power-on of the bare metal server and the enablement of the PXE startup mode may be performed simultaneously or sequentially, and can be flexibly implemented by those skilled in the art.
  • the application may have two improvements to the booting system in the second internal memory 403.
  • an information collecting interface is added in the booting system for collecting hardware attributes of the bare metal server
  • the guiding system is used for providing information collection.
  • the running environment of the module, the information collecting module runs in the booting system environment, and before the booting system boots the operating system to start, the information collecting interface is called to obtain the hardware attributes of the bare metal server, and the acquired hardware attributes of the bare metal server are sent.
  • the information collection module may be a firmware module preset in a second internal memory ROM in the bare metal server, or may be a software module downloaded from the bare metal manager in the PXE phase. The two implementations are described in detail below.
  • FIG. 5A is a simplified block diagram of a cloud computing system composed of a bare metal manager and a bare metal server in an embodiment of the present application
  • FIG. 5B is a hardware attribute of the bare metal manager acquiring a bare metal server in the system shown in FIG. 5A.
  • Method flow chart when the cloud computing system needs to increase server resources, the bare metal manager needs to issue a bare metal server. The bare metal manager needs to collect the hardware properties of the bare metal server before the bare metal server is released.
  • the information collecting module is pre-replaced on the TFTP server of the PXE server on the bare metal manager side instead of the booting boot module, that is, the information collecting module is downloaded from the bare metal manager in the PXE phase.
  • the bare metal manager 201 enables the PXE boot mode of the bare metal server 12 and powers up the bare metal server 12.
  • the boot management module 2011 in the bare metal manager may send a boot command to the boot system of the bare metal server, and control the power-on and PXE boot mode of the bare metal server by controlling the boot controller 406 in the bare metal server. Enable.
  • the boot system in the second internal memory 403 is loaded into the first internal memory 402, and the processor 401 of the bare metal server 12 executes the boot system, which is booted.
  • the booting system loads the PXE client in the network card 404 and runs the PXE client.
  • the bare metal server 12 After the PXE client runs on the bare metal server, the bare metal server 12 performs the PXE client function.
  • the bare metal manager 201 performs the function of the PXE server, and the PXE client performs a process of obtaining an IP address.
  • the process of obtaining an IP address includes steps 502-505, whether the booting system or the PXE client is run by the processor 401 of the bare metal server, and the following process will describe the IP address acquisition by using the bare metal server as the main body. Process.
  • the bare metal server 12 sends a DHCP discovery message, which is generally a broadcast message, and the discovery module 2012 of the bare metal manager 201 has the function of a PXE server, which receives the DHCP discovery message sent in broadcast form.
  • the bare metal manager 201 sends a DHCP offer to the bare metal server 12, and in step 504, the bare metal server 12 sends a DHCP to the bare metal manager 201 after receiving the DHCP offer sent by the bare metal manager 201.
  • the request is to obtain an IP address, and after receiving the DHCP request, the bare metal manager 201 assigns an IP address to the bare metal server 12.
  • the bare metal manager 201 may also notify the bare metal server 12 of the information of the file to be downloaded when the IP address is sent to the bare metal server 12, and the information of the file to be downloaded is used to indicate The bare metal server 12 downloads the file to be downloaded from the bare metal manager to perform the related functions of the file to be downloaded.
  • the information of the file to be downloaded may include information such as address information of the download server or a save address of the file to be downloaded or a name of the file to be downloaded.
  • the file may be a TFTP server by transmitting the file through the TFTP.
  • the bare metal manager 201 notifies the bare metal server 12 that the information of the file to be downloaded may be the address of the TFTP server (the download server is generally set in the bare metal manager). Side, and in some cases, the bare metal manager itself is the download server).
  • the file to be downloaded saved in the TFTP server is a booting boot module (for example, a PXE boot loader for booting the boot operating system).
  • the TFTP server on the bare metal manager 201 side does not The boot module is saved, and the information collection module is saved. That is to say, the file to be downloaded saved on the TFTP server is the information collection module, and the boot module is no longer started.
  • the information collection module is configured to enable the bare metal server to acquire and send hardware attributes of the bare metal server 12. Therefore, in step 505, the bare metal server 12 receives the IP address sent by the bare metal manager 201 and the information of the file to be downloaded.
  • the information of the file to be downloaded in this embodiment is the storage address of the information collecting module.
  • Step 506 The bare metal server 12 downloads the information collection module from the download server according to the save address of the information collection module, and loads the information collection module into the memory of the bare metal server.
  • the discovery module 2012 in the bare metal manager 201 notifies the bare metal server 12 of the save address of the information collection module, and receives a file acquisition request sent by the bare metal server 12 according to the save address of the information collection module, bare metal.
  • the manager 201 provides the information collection module to the bare metal server 12 over the network. After the bare metal server 12 downloads the information collection module, the information collection module runs in the operating environment of the boot system, and invokes the information collection interface in the boot system to obtain the hardware attributes of the bare metal server 12.
  • the information collecting module may be a UEFI Application, which runs in a UEFI BIOS environment.
  • the bare metal server 12 downloads and runs the information collection module, since the original IP address cannot be transmitted to the information collection module, the IP address is re-acquired, and the steps 507-510 are the same as the process of obtaining the IP address in the above steps 502-505, no longer. Narration.
  • the bare metal server 12 transmits the hardware attribute of the bare metal server acquired by the information collecting phase to the bare metal manager 201 in step 511.
  • the bare metal server 12 is downloaded to the information collection module in step 506, the information collection module is executed and information collection is started.
  • the IP address acquisition phase of steps 507-510 can be synchronized with the information collection phase, and the bare metal server is used.
  • the time when the IP address is obtained and the time when the hardware attribute is collected may be simultaneously or sequentially.
  • the embodiment of the present patent application has no special requirements here.
  • the bare metal manager of the present application obtains the hardware attributes of the bare metal server
  • the bare metal manager provides an information collecting module and runs in the operating environment of the booting system on the bare metal server. After the information collecting module is run
  • the information collection interface newly provided by the boot system can be called to obtain the hardware properties of the bare metal server. Since the information booting module runs in the pre-boot phase of the operating system executed by the booting system, the solution of the present invention can complete the acquisition of the hardware attributes of the bare metal server without relying on the startup of the operating system, so that the bare metal manager can be faster.
  • Obtaining the hardware properties of the bare metal server greatly improves the distribution efficiency of the bare metal server.
  • the information collection module is relatively small, the operation and download do not need to consume a large amount of network traffic, which reduces network congestion.
  • the hardware properties of bare metal servers can be discovered without much improvement on bare metal servers.
  • the information collecting module may be a UEFI Application.
  • the UEFI Application is hardware initialization completed, and the operating system starts.
  • the previously running application, UEFI Application runs in UEFI operation
  • the Transient System Load (TSL) phase in the environment.
  • the entire operating environment after the UEFI BIOS is started includes the following phases: 1. Security Phase (SEC). After the computing device is powered on, it enters this phase. 2.
  • the pre-EFI Initilization (PEI) phase is used to prepare the execution environment for the next phase. 3.
  • the Driver Execution Environment (DXE) phase which performs most of the system initialization work.
  • the memory can be completely used, so the complex work can be performed at this stage.
  • the Boot Device Selection (BDS) phase which mainly executes the startup strategy.
  • the main functions include: initializing the console device; loading the device. Drive; load and execute startup items based on system settings.
  • the above 1-4 stages may be referred to as an initialization stage.
  • system resources are still controlled by the UEFI core.
  • the role of the TSL phase is to prepare the execution environment for the OS Loader, so this phase is called a temporary system.
  • the BS's ExitBootServices service is called, the system goes to the next stage. 6.
  • the information collecting module provided by the application of the present invention is an application running in the stage before the loading of the operating system of the fifth stage, thereby realizing the method of acquiring the hardware attribute and sending it to the bare metal manager before the operating system is started, thereby improving the nakedness. Metal server distribution efficiency.
  • the booting system on the bare metal server is the UEFI BIOS
  • information may be added to the existing Boot Service or Runtime Service of the UEFI.
  • the interface is collected and the UEFI Application calls the information collection interface to obtain the hardware attributes of the bare metal server.
  • the UEFI new information collection interface description can include:
  • the hardware attributes of the bare metal server are passed to the caller (information collection module, such as UEFI Application).
  • the hardware attributes can be encapsulated by a typical structure or by an array.
  • the specific contents include but are not limited to: CPU architecture, Number of logical CPUs, hard disk size, memory size, MAC address information, etc.
  • the UEFI new information collection interface definition can include:
  • an implementation manner of the hardware attribute of the bare metal server 12 may be sent by using a DHCP protocol, for example, using a DHCP option to carry hardware attributes of the bare metal server.
  • the bare metal server 12 obtains the IP address assigned by the DHCP server in the bare metal manager.
  • the information collecting module acquires the hardware attribute of the bare metal server, in step 511, the information collecting module may send the information again.
  • the DHCP request and the collected hardware attributes are constructed in the option code in the DHCP REQUEST message.
  • the discovery module 2012 (including the PXE Server function) in the bare metal manager 201 receives the DHCP REQUEST message and will respond to the option code according to the configuration.
  • any reserved bits in the reserved code93 or code128-code254 may be used to construct the hardware attribute field of the bare metal server, for example, the code 93 is defined as the architecture field of the bare metal server ( Client-architecture), define code128 as the system CPU number field of the bare metal server (client-vcpu), define code129 as the system hard disk size (client-disk) of the bare metal server, and define the code130 as the bare metal server system.
  • the memory size (client-memory), the code131 is defined as the bare metal server's network card information (client-macs) and so on.
  • the bare metal server can determine which DHCP message to use to carry the hardware attribute according to the time when the hardware attribute is actually obtained.
  • the information collecting module sends the DHCP request again. Carrying the hardware attributes of the bare metal server, in fact, the hardware attributes may also be carried in the message in step 507 or step 509.
  • the bare metal server 12 sends the hardware attribute may be a Representational State Transfer (REST) protocol message transmission using EFI.
  • the peer module that is, the discovery module 2012 in the bare metal manager 201, includes the function of parsing the REST protocol, and is used for receiving and parsing the hardware attributes of the bare metal server carried in the REST message.
  • an information setting interface may be added, which is used for hardware attributes of the bare metal server. Modify the IP address information of the Baseboard Management Controller (BMC), for example. Similar to the above information collection interface, the information setting interface can also be added in the existing booting service or running service of the UEFI. Accordingly, the bare metal manager can also be used to send a hardware attribute modification request to the bare metal server; the information collecting module It is also possible to add an information setting function for calling the information setting interface to set the hardware properties of the bare metal server.
  • the specific definition of the information setting interface can be as follows:
  • the above information setting interface may exist separately from the information collecting interface, that is, the modification of the hardware attribute may not depend on the collection of hardware attributes, before the operating system is started.
  • the technical method related to the present application can also be used to directly set the hardware attributes, that is, the bare metal manager modifies the hardware attributes of the bare metal server before collecting the hardware attributes of the bare metal server.
  • FIG. 6A is a simplified block diagram of a computer system composed of a bare metal manager and a bare metal server in still another embodiment of the present application
  • FIG. 6B is a hardware attribute of the bare metal manager acquiring a bare metal server in the system shown in FIG. 6A.
  • the information collection module has been previously saved in the ROM of the bare metal server, and specifically, for example, can be integrated in the boot system, that is, the information collection function is newly added by improving the boot system of the bare metal server. Module.
  • step 601 the bare metal manager 201 enables the PXE boot mode of the bare metal server 12 and powers up the bare metal server 12.
  • step 602 after the bare metal server 12 is powered on, the booting system in the second internal memory 403 is loaded into the first internal memory 402, and the processor 401 of the bare metal server 12 executes the booting system, and the booting system is started. After the booting system finds that the PXE is enabled, the PXE client in the network card 404 is loaded and the PXE client is run.
  • the bare metal server 12 further includes an information collecting switch, the information collecting switch is configured to indicate whether the guiding system runs the information collecting module, and the guiding system determines, according to the indication of the information collecting switch, whether to load the Information collection module.
  • the information collection switch may be controlled by the bare metal manager through the BMC, for example, when the bare metal manager determines the hardware attribute of the bare metal server, or when the bare metal server is to be issued, the information is The collection switch is turned on. Further, before the PXE client runs to download the PXE operating system boot module to the TFTP server in the bare metal manager 201, the booting system runs the information collecting module according to the indication of the information collecting switch to complete the nakedness.
  • the indication of the information collection switch is that the information collection module needs to be loaded. Therefore, when the boot system runs to the pre-loading stage of the operating system, the information collecting module starts running in the environment of the booting system, and performs the collection of the hardware attributes of the bare metal server, and the information collecting module calls the information collecting interface in the booting system to obtain the naked Hardware properties of the metal server 12.
  • the information collecting module may be a UEFI Application, which runs in a UEFI BIOS environment.
  • Steps 603-606 the process of the IP address acquisition by the bare metal server is performed by the bare metal server, which is described in the foregoing embodiment, and is not described here.
  • the bare metal server 12 After obtaining the IP address, the bare metal server 12 transmits the hardware attribute of the bare metal server acquired in the information collecting phase to the bare metal manager 201 in step 607.
  • the storage location of the information collection module is not on the bare metal manager side, but is stored in the bare metal server.
  • the information collection module may not be downloaded from the remote end before the operating system is loaded. Load execution can be performed to discover the hardware properties of bare metal servers.
  • the execution time is further saved, and the efficiency of the bare metal manager to obtain the hardware properties of the bare metal server is improved, thereby improving the bare metal manager to issue the bare metal. Server efficiency.
  • the bare metal server in this application generally does not have an operating system installed.
  • the bare metal manager can directly enable the PXE boot mode and power on the bare metal server to implement subsequent collection hardware attributes.
  • the method of sending to the bare metal manager, for a bare metal server that has an operating system installed, the bare metal manager can modify the boot mode of the bare metal server before restarting the bare metal server.

Landscapes

  • Engineering & Computer Science (AREA)
  • Software Systems (AREA)
  • Theoretical Computer Science (AREA)
  • General Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Computer Security & Cryptography (AREA)
  • Computer Networks & Wireless Communication (AREA)
  • Signal Processing (AREA)
  • Stored Programmes (AREA)

Abstract

本发明提出了一种裸金属服务器的硬件属性的发现方法和系统。针对裸金属管理器需要在裸金属服务器上启动操作系统之后才能获得裸金属服务器硬件属性这种效率较低的问题，本发明提供的裸金属服务器的硬件属性的发现方法不依赖于操作系统，在裸金属服务器中的引导系统启动之后，以及在引导系统引导操作系统启动之前，本发明在该引导系统上运行信息收集模块，该信息收集模块收集并向裸金属管理器发送裸金属服务器的硬件属性。

Description

裸金属服务器的硬件属性发现方法和系统 技术领域
本发明大体上涉及信息技术(IT)，尤其涉及一种裸金属服务器的硬件属性的发现方法和系统。
背景技术
服务器虚拟化技术是指能够在一台物理服务器上运行多台虚拟服务器的技术，并且上述虚拟服务器在用户、应用软件甚至操作系统看来，几乎与物理服务器没有区别。用户可以在虚拟服务器上灵活地安装任何软件。在服务器虚拟化技术中，被虚拟出来的服务器称为虚拟机(英文：virtual machine，VM)。运行在虚拟机里的操作系统称为访客操作系统(英文：guest OS)。负责管理虚拟机的软件称为虚拟机监视器(英文：virtual machine monitor，VMM)，也可以称为管理者(英文：hypervisor)。
采用服务器虚拟化技术提供服务的系统可以称为云计算系统。云计算系统中的物理基础设施包括各种物理设备，用于提供计算、存储或者网络资源。其中，裸金属服务器是云计算系统的物理基础设施中的一种。对于加入云计算系统中的裸金属服务器，云计算系统需要解决基础环境的搭建问题，例如解决裸金属服务器的添加、删除、电源管理、操作系统部署等问题。对于上述问题，可以在云计算系统中采用专用的软件对该裸金属服务器的生命周期进行管理。例如OpenStack(OpenStack是一个开源的云计算管理平台项目)中提供了Ironic组件。Ironic组件提供了管理裸金属服务器的应用编程接口(英文：application programming interface，API)。Ironic组件提供的API可以管理裸金属服务器。
云计算系统中，裸金属服务器的管理软件(例如Ironic)在裸金属服务器发放(英文：provision)之前，收集裸金属服务器的硬件属性。硬件属性可以是裸金属服务器的硬盘信息(如硬盘存储空间的大小和/或硬盘类型)、内存信息、中央处理器(CPU)信息(如CPU的核心的数量和/或各核心的频率)、网卡信息或其任意组合。裸金属服务器的管理软件在收集到硬件属性之后，将该裸金属服务器中的资源发放给用户。裸金属服务器可以直接被发放给用户，例如，为高强度计算或实时性要求高的服务直接提供业务，也可以将裸金属服务器的资源虚拟化后再发放给用户虚拟化的资源。
裸金属服务器的管理软件收集裸金属服务器的硬件属性需要先通过网络启动的方式在裸金属服务器上启动操作系统，再通过操作系统中的应用获取相关的硬件属性。该方法耗时较多，不能满足裸金属管理器快速发放裸金属服务器的需求。
发明内容
本文提出了各种技术方案，以实现不依赖操作系统的情况下，收集裸金属服务器的硬件属性，用来满足裸金属管理器快速发放裸金属服务器的需求。
第一方面，本申请提供一种云计算系统，用于发现裸金属服务器的硬件属性，该云计算系统包括裸金属管理器以及裸金属服务器，该裸金属管理器用来上电裸金属服务器，例如通过启动命令控制裸金属服务器的启动系统，该启动系统控制电源模块以对该裸金 属服务器进行上电操作；在该裸金属服务器上电之后，该裸金属服务器中的引导系统，例如BIOS，被启动；该引导系统用于运行预启动执行环境PXE客户端并通过PXE启动方式引导操作系统启动，以及在所述操作系统启动之前运行信息收集模块；该信息收集模块，用于收集该裸金属服务器的硬件属性，并将该裸金属服务器的硬件属性通过网络发送给所述裸金属管理器。
上述方案中，裸金属服务器中的信息收集模块运行在引导系统运行环境下的操作系统启动前阶段，使得硬件属性的发现可以不依赖于操作系统，不需要启动操作系统即可完成裸金属服务器的硬件属性的获取，并快速发送给裸金属管理器，使得裸金属管理器获得裸金属服务器的硬件属性的时间大大减少，提高了裸金属服务器的部署效率，尤其是对于大型的数据中心需要部署多台裸金属服务器的情况，大大提高了部署效率；同时，由于信息收集模块很小，运行和下载都不需要耗费网络大量的流量，降低了网络拥塞。另外，该引导系统执行上述方法对云计算系统的改变很小，所花费的开发代价也比较小。
对于上述第一方面，一种可能的实现方式是该裸金属管理器还用于使能该裸金属服务器的PXE启动方式。这种方式用于裸金属服务器没有设置操作系统启动方式的场景或者裸金属服务器安装有操作系统的场景(已经预装操作系统的场景下操作系统启动方式为非PXE启动方式)，上述实现使得在上述两种场景下操作系统启动方式都能切换到PXE启动方式。
对于上述第一方面，再一种可能的实现方式是所述引导系统包括信息收集接口，所述信息收集模块用于调用所述信息收集接口以获取所述裸金属服务器的硬件属性。这种方式需要对引导系统进行扩展，使得引导系统增加信息收集接口，信息收集模块调用该信息收集接口以收集硬件属性，由于引导系统在初始化阶段已经获取到硬件属性，直接在引导系统中增加相应的接口来收集硬件属性，便于实现。
对于上述第一方面，再一种可能的实现方式是在引导系统中再增加信息设置接口，裸金属管理器还用于通过网络向该裸金属服务器发送硬件属性修改请求，所述信息收集模块还用于根据所述硬件属性修改请求，调用所述信息设置接口对所述裸金属服务器的硬件属性进行修改。本实现方式进一步提供了裸金属管理器在裸金属服务器的操作系统启动之前，就对裸金属服务器的硬件属性进行修改的可能性，以实现裸金属管理器对裸金属服务器更灵活的管理。
对于上述第一方面，再一种可能的实现方式是所述信息收集模块预先保存在裸金属管理器，裸金属服务器中的PXE客户端用于获取为裸金属服务器分配的IP地址，使用所述IP地址与裸金属管理器进行通信，并从裸金属管理器下载所述信息收集模块。本实现方式中，裸金属管理器提供信息收集模块，也就是说，在具体实现中，利用裸金属管理器易于扩展的特性，在裸金属管理器进行扩展，使得裸金属管理器能够提供信息收集模块。而且，裸金属服务器获得信息收集模块不需要额外的流程，而是在PXE启动流程中获得所述信息收集模块。
在上述实现方式中，一种可选的具体实现方式是所述信息收集模块代替PXE启动引导模块保存在FTP服务器，裸金属管理器将FTP服务器的地址通知到裸金属服务器。
对于上述第一方面，再一种可能的实现方式是所述信息收集模块预先保存在所述裸金属服务器中，裸金属服务器的引导系统在所述操作系统启动前阶段加载并运行所述信 息收集模块。本实现方式中，信息收集模块的存储位置不在裸金属管理器侧，而是保存在裸金属服务器中，在引导系统运行时，在操作系统加载前阶段，该信息收集模块可以不用从远端下载即可加载执行该信息收集模块，进一步节约了执行时间，提高了裸金属管理器发现裸金属服务器硬件属性的效率。
在上述实现方式中，一种可选的具体实现方式是裸金属服务器还包括信息收集开关，所述信息收集开关用于指示所述引导系统是否运行所述信息收集模块；如果所述信息收集开关指示运行所述信息收集模块，所述裸金属服务器中的引导系统用于加载所述信息收集模块。
可选地，在所述PXE客户端运行之后并且在向所述裸金属管理器中的FTP服务器下载PXE启动引导模块之前，所述引导系统根据所述信息收集开关的指示运行所述信息收集模块。
对于上述第一方面，再一种可能的实现方式是裸金属服务器中的信息收集模块用动态主机设定协议DHCP消息或代表性状态传输REST消息携带所述裸金属服务器的硬件属性。
第二方面，本申请提供一种裸金属服务器的硬件属性发现方法，应用于裸金属管理器，所述裸金属管理器对裸金属服务器进行管理；所述方法包括：上电所述裸金属服务器；以及，接收所述裸金属服务器通过网络发送的所述裸金属服务器的硬件属性，其中，所述裸金属服务器中的引导系统用于运行预启动执行环境PXE客户端并通过PXE启动方式引导操作系统启动，以及在所述裸金属服务器的操作系统启动之前运行信息收集模块，所述信息收集模块用于收集并通过网络发送所述裸金属服务器的硬件属性。
上述第二方面的方法应用于裸金属管理器，上述第二方面的可能的具体实现方式可以参考上述第一方面中裸金属管理器的具体实现部分，在此不再赘述。
第三方面，本申请提供一种管理装置，用于对裸金属服务器进行管理；所述管理装置包括：存储设备，用于存储指令；以及至少一台处理器，与所述存储设备耦合；其中，当所述至少一台服务器中的处理器执行所述指令时，所述指令致使所述处理器执行上述第二方面的方法。
第四方面，本申请提供一种裸金属管理器，用于对裸金属服务器进行管理，所述管理器包括：启动管理模块，用于上电所述裸金属服务器；以及，发现模块，用于接收所述裸金属服务器通过网络发送的所述裸金属服务器的硬件属性，其中，所述裸金属服务器中的引导系统用于运行PXE客户端并通过PXE启动方式引导操作系统启动，以及在所述裸金属服务器的操作系统启动之前运行信息收集模块，所述信息收集模块用于收集并通过网络发送所述裸金属服务器的硬件属性。
对于上述第四方面，一种可能的实现方式是启动管理模块还用于使能所述裸金属服务器的PXE启动方式。
对于上述第四方面，再一种可能的实现方式是所述发现模块，还用于通过网络向所述裸金属服务器发送硬件属性修改请求。
对于上述第四方面，再一种可能的实现方式是所述发现模块还用于保存所述信息收集模块，以及通过网络向所述裸金属服务器提供所述信息收集模块。
对于上述第四方面，再一种可能的实现方式是所述发现模块具体用于接收所述裸金 属服务器发送的动态主机设定协议DHCP协议消息，通过所述发现模块中的DHCP服务器获取所述DHCP协议消息中携带的所述裸金属服务器的硬件属性；或者，所述发现模块具体用于接收所述裸金属服务器发送的代表性状态传输REST协议消息，通过所述发现模块中的解析模块获取所述REST协议消息中携带的所述裸金属服务器的硬件属性。
第五方面，本申请提供一种裸金属服务器的硬件属性发现方法，其特征在于，应用于裸金属服务器，裸金属管理器对裸金属服务器进行管理；所述方法包括：接收所述裸金属管理器的启动命令，根据所述启动命令进行上电操作；在所述裸金属服务器上电后，所述裸金属服务器中的引导系统被启动，所述引导系统通过预启动执行环境PXE启动方式引导操作系统启动；以及在操作系统启动之前，收集所述裸金属服务器的硬件属性，并将所述裸金属服务器的硬件属性通过网络发送给所述裸金属管理器。
对于上述第五方面，一种可能的实现方式是在所述收集所述裸金属服务器的硬件属性之前还包括：获取为所述裸金属服务器分配的IP地址，使用所述IP地址与所述裸金属管理器进行通信，并从所述裸金属管理器下载信息收集模块，以及在所述引导系统中运行所述信息收集模块。
对于上述第五方面，再一种可能的实现方式是在所述收集所述裸金属服务器的硬件属性之前还包括：所述引导系统在所述操作系统启动前阶段从所述裸金属服务器中加载并运行信息收集模块。
对于上述第五方面，再一种可能的实现方式是所述信息收集模块调用所述引导系统中的信息收集接口获取所述裸金属服务器的硬件属性。
对于上述第五方面，再一种可能的实现方式是所述方法还包括：接收所述裸金属管理器发送的硬件属性修改请求，根据所述硬件属性修改请求，所述信息收集模块调用所述裸金属服务器中的信息设置接口，对所述裸金属服务器的硬件属性进行修改。
第六方面，本申请提供一种裸金属服务器，裸金属管理器对所述裸金属服务器进行管理；所述服务器包括：启动控制器、中央处理器、第一内部存储器、第二内部存储器以及网卡；所述启动控制器，用于接收所述裸金属管理器的启动命令，根据所述启动命令进行上电操作；所述第二内部存储器中保存有引导系统，所述裸金属服务器上电之后所述引导系统被加载到所述第一内部存储器中，所述中央处理器运行所述引导系统；所述引导系统用于运行所述PXE客户端并通过PXE启动方式引导操作系统启动，以及在所述操作系统启动之前运行信息收集模块；所述信息收集模块，用于收集所述裸金属服务器的硬件属性，并将所述裸金属服务器的硬件属性通过所述网卡发送给所述裸金属管理器。
上述第六方面应用于裸金属服务器，上述第六方面的可能的具体实现方式可以参考上述第一方面中裸金属服务器的具体实现部分，在此不再赘述。
第七方面，本申请提供一种计算机程序产品，所述计算机程序产品保存在可读存储介质上，所述计算机程序产品包含指令，所述指令运行在计算机中的引导系统的运行环境下，并在所述计算机的操作系统启动前被运行，所述指令致使处理器执行以下操作：调用所述引导系统中的信息收集接口，收集所述裸金属服务器的硬件属性信息；以及通过网络将所述裸金属服务器的硬件属性信息发送给裸金属管理器。第八方面，本申请提供一种计算机系统，包括管理器与服务器；所述服务器包括引导系统，在所述服务器上 电之后，所述引导系统被启动；所述引导系统，用于运行预启动执行环境PXE客户端并通过PXE启动方式引导操作系统启动，以及在所述操作系统启动之前运行信息收集模块；所述信息收集模块，用于收集所述服务器的硬件属性，并将所述服务器的硬件属性通过网络发送给所述管理器；所述管理器用于接收所述服务器发送的所述服务器的硬件属性。
上述第八方面提供的计算机系统，可以为上述第一方面提供的云计算系统，也可以为传统的数据中心系统，管理器可以包含裸金属管理器，也可以包含实现本申请第七方面功能的其它任何形态的管理器，服务器可以包含裸金属服务器，也可以包含实现本申请第七方面功能的任何其他形态的物理服务器。
进一步，上述第八方面的具体实现方式可以参考上述第一方面的具体实现方式。
在研究下文附图和详细描述之后其它系统、方法、特征和优点对于本领域普通技术人员来说将是或变得显而易见。希望所有这些其它系统、方法、特征和优点包含在本描述中，在本发明的范围内，并且受所附权利要求书的保护。
附图说明
现将参考附图通过非限制性的实施例描述本发明的优选特征，其中：
图1A为云计算系统的简化方框图；
图1B为云计算系统的又一简化方框图；
图2A为传统方案的裸金属服务器的硬件属性的获取的各阶段示意图；
图2B为本发明的裸金属服务器的硬件属性的获取的流程示意图；
图3A为裸金属服务器的硬件属性的获取的各阶段示意图；
图3B为传统方案的各阶段、本发明方案各阶段分别与裸金属服务器启动各阶段间的关系示意图；
图4为本申请方案中裸金属服务器的组成方框图；
图5A为本申请的一个具体实施例中裸金属管理器与裸金属服务器组成的计算机系统的简化方框图；
图5B为本申请的一个具体实施例中裸金属管理器获取裸金属服务器的硬件属性的流程示意图；
图6A为本申请的又一个具体实施例中裸金属管理器与裸金属服务器组成的计算机系统的简化方框图；
图6B为本申请的又一个具体实施例中裸金属管理器获取裸金属服务器的硬件属性的流程示意图；
在整个附图中，相同的参考符号和描述表示类似的但不一定完全相同的元件。尽管本文所述的示例性实施例很容易进行各种修改或替换成其它形式，所以特定实施例已经通过附图中的示例示出并将在本文中进行详细描述。然而，本文所述的示例性实施例并不旨在限制所公开的特定形式。相反，当前发明涵盖属于所附权利要求书的范围内的所有修改、等效物以及替代物。
具体实施方式
为了使本申请的上述目的、技术方案和优点更易于理解，下文提供了详细的描述。所述详细的描述通过使用方框图、流程图和/或示例提出了设备和/或过程的各种实施 例。由于这些方框图、流程图和/或示例包含一个或多个功能和/或操作，所以本领域内人员将理解可以通过许多硬件、软件、固件或它们的任意组合单独和/或共同实施这些方框图、流程图或示例内的每个功能和/或操作。
以下为本申请文件中包含的术语：
裸金属服务器，云计算系统中的物理设施中的一种，也可称为裸金属节点或“裸”操作系统的物理机，一般指没有安装操作系统或虚拟化软件的服务器。“裸”操作系统的物理机也可指安装有操作系统但是没有安装虚拟化软件的物理机。为了与云计算系统中的其它对象(例如传统的用来进行资源虚拟化的物理服务器或者虚拟化得到的虚拟机)进行区分，本申请使用“裸金属服务器”这种称谓来表示云计算系统中待发放或待部署的物理服务器，实践中，本申请所指的裸金属服务器可以有多种称谓，例如“待发放服务器”与“待部署服务器”，在此不一一列举。
裸金属管理器，用于对裸金属服务器进行管理，也可以称为裸金属服务器的管理组件，其中，在此处的术语“管理组件”或者“管理器”可以指代软件、固件、中间件、微代码、硬件和/或其任意组合。例如，管理组件或管理器可以是在通用服务器上运行的模块、进程、线程任意之一或组合，也可以就是通用服务器自身。管理组件或管理器可以是一个或者多个，管理组件或管理器可以位于一台通用服务器上或者分布在两台或两台以上通用服务器上。管理组件或管理器可以从存储有各种数据结构的计算机可读介质中获取并执行，也可以通过硬件电路的各种逻辑组合来实现。基于上述描述可知，裸金属管理器是本专利申请中基于表述的清楚统一性而选择的一种称谓，在本申请所涉及的技术被具体实施时，裸金属管理器可以有多种实现方式，因此亦可以存在多种表达方式，例如，裸金属服务器的控制器、裸金属服务器的控制装置、Ironic组件或者其他名称的组件。上述通用服务器为本领域技术人员所知的包含处理器和存储设备的物理服务器，存储设备用于存储指令，处理器用于执行指令。
预启动执行环境(preboot execute environment，PXE)，PXE启动方式是目前主流的无盘启动技术，它可以使计算机通过网络而不是从本地硬盘、光驱等设备启动，采用Client/Server的网络模式，在启动过程中，终端要求服务器分配IP地址，再用文件传输协议(file transfer protocol，FTP)，文件传输协议可以包括简单文件传输协议(trivial file transfer protocol，TFTP)或多播简单文件传输协议MTFTP(multicast trivial file transfer protocol)协议，下载一个启动软件包到本机内存中执行，由这个启动软件包完成终端基本软件设置，从而引导预先安装在服务器中的终端操作系统。计算机启动操作系统的方式有网络启动、光盘启动或硬盘启动等几种方式，PXE启动方式是其中网络启动的方式。本申请中，裸金属服务器执行所述PXE客户端，裸金属管理器执行所述PXE服务器。对于PXE服务器，需要支持DHCP功能与TFTP功能；DHCP(Dynamic Host Configuration Protocol)，动态主机设定协议，PXE服务器支持该DHCP功能即可为PXE客户端分配IP地址，该IP地址用于PXE客户端通过网络与PXE服务器通信；PXE服务器支持该TFTP功能即可通过TFTP协议向PXE客户端传输文件，例如操作系统相关的引导模块或者操作系统的镜像文件。
服务器，本申请中的服务器有硬件与软件的区分，软件服务器指管理资源并提供对应服务的计算机软件，在本申请中，与客户端相对应的服务器均指软件服务器，例如 PXE服务器、DHCP服务器以及TFTP服务器，硬件服务器指不同于个人计算机形态的物理计算设备，在本申请中，裸金属服务器、已配置服务器与通用服务器为硬件服务器。
引导系统，不同于操作系统，用于管理裸金属服务器的启动，以及重要硬件的控制和驱动，并为高层软件提供基层调用，具体可以保存有裸金属服务器的基本输入输出系统的模块、开机后自检模块和系统自启动模块。该引导系统可以是以写入裸金属服务器中的只读存储器的固件形态存在，例如该引导系统可以是基本输入输出系统(Basic Input Output System，BIOS)，具体地，可以是采用UEFI的BIOS或者采用非UEFI的BIOS。UEFI(Unified Extensible Firmware Interface)，统一的可扩展固件接口，是一种可编程和可扩展的新型的固件标准。
上电，本申请中的服务器上电指服务器从未开机的状态进入开机状态，而不是服务器从未通电状态进入通电但并未开机的状态。
以下对本申请的具体实施场景和方法进行详细介绍。
图1A和图1B分别为本申请的云计算系统的组成框图。如图1A所示，云计算系统中的已配置服务器11(图示1台，实际数量可为1台或多台)为已经上电并安装了操作系统或者虚拟化软件的服务器，其包含硬件平台110，硬件平台110中可以包括服务器物理结构中的硬盘、中央处理器、网络接口卡和存储器等。在硬件平台110上可运行软件平台112，软件平台112可包含虚拟化软件、虚拟机的操作系统，以及虚拟机的应用软件等软件。虚拟化软件(例如，Hypervisor)是物理硬件和虚拟机之间的中间层，负责协调各个虚拟机对硬件平台的访问。Hypervisor又称为虚拟机监视器VMM(Virtual Machine Monitor)，VMM包括实现虚拟机的虚拟硬件平台，虚拟硬件平台中包含了虚拟出来的存储器，中央处理器，硬盘，显卡和网络接口卡等。在VMM之上运行了多个虚拟机VM1至VMn，虚拟机的软件系统运行在虚拟机监视器的虚拟硬件平台上。
图1A的系统除了包含已配置服务器11之外，还包含裸金属服务器12(图示1台，实际数量可为1台或多台)。对于加入云计算系统中的裸金属服务器，云计算系统需要解决基础环境的搭建问题，例如解决裸金属服务器的添加、删除、电源管理、操作系统部署等问题。对于上述问题，在云计算系统中的云计算控制平台20中可以专门部署专用的裸金属管理器201对该裸金属服务器的生命周期进行管理，例如OpenStack中提供了Ironic组件，Ironic组件提供了一系列管理物理机的API，可以对“裸”操作系统的物理机进行管理，包含从物理机上电安装操作系统到物理机下电维修整个生命周期的管理。云计算控制平台20可以单独部署在一台服务器上也可以部署在多台服务器上。
图1B示出了云计算系统的另外一种组成方式，与图1A不同之处在于，云计算控制平台20由已配置服务器11中的虚拟机来实现，也就是说云计算控制平台并不单独由一台物理机部署，而是运行在已配置服务器11中的虚拟机之中(图示部署在虚拟机VMn上，也可以实际部署在多台虚拟机)，由虚拟机履行云计算控制平台20的职责。
云计算控制平台20中除了设置有裸金属管理器201之外，还可以包含其它的组件，例如虚拟机管理组件、安全组件或者计费组件或者能耗组件等等。
上述图1A与图1B描述的裸金属服务器均在云计算系统(云数据中心)中实现，实际上，本专利申请中裸金属服务器也可以部署在传统的计算机系统中(例如传统的数据中心)，即计算机系统中的已部署服务器上不再虚拟化出虚拟机的场景，此种场景下， 裸金属管理器与裸金属服务器可以分别称为管理器与服务器；另外，本专利申请同时也适用于云计算系统的初始的场景，例如，云计算系统中所有的服务器均为裸金属服务器的场景。总之，无论是云计算系统的场景，还是传统的计算机系统的场景，本申请的技术方案均可适用，本申请技术方案所适用的系统为计算机系统。
裸金属管理器201在裸金属服务器发放(Provision)之前，需要获取裸金属服务器的硬件属性，硬件属性可以是裸金属服务器的硬盘属性(如大小或类型)、内存属性、CPU属性(体系结构、个数、核数或频率等)或网卡属性(MAC地址)、基板管理控制器(Baseboard Management Controller，BMC)的属性或其任意组合，上述硬件属性并未穷举所有的裸金属服务器的硬件属性，本领域技术人员可以扩展出上述硬件属性的举例以外其他裸金属服务器的硬件属性。裸金属管理器201在获取到硬件属性之后，才能对该裸金属服务器进行相应的操作，例如，云计算系统发放裸金属服务器，可以利用裸金属服务器的硬件资源进行计算虚拟化或存储虚拟化，也可以利用裸金属服务器直接为高强度计算或其它实时性要求高的服务提供业务。本申请中，将裸金属管理器获取裸金属服务器的硬件属性的过程称为裸金属服务器硬件属性的发现过程，裸金属服务器硬件属性的发现过程，即可以包含裸金属管理器如何触发发现流程并最终获得硬件属性的过程，也可以包含裸金属服务器如何收集硬件属性并向裸金属管理器报告硬件属性的过程。
如图2A和图2B所示，传统的裸金属管理器201发现裸金属服务器的硬件属性的方法包括五个阶段。第一阶段：如步骤201，在T0时刻裸金属管理器裸金属服务器上电，裸金属管理器201使能裸金属服务器12的PXE启动方式，裸金属服务器中的引导系统启动并运行网卡中的PXE客户端；接下来如步骤202-205，裸金属服务器12的PXE客户端会与裸金属管理器201(裸金属管理器201具备DHCP Server的功能)进行IP地址分配流程，包括DHCP发现、DHCP提供、DHCP请求、DHCP确认，在此过程中裸金属管理器201作为DHCP服务器为裸金属服务器提供IP地址和待下载文件的下载地址(例如TFTP服务器地址)，然后在步骤206，裸金属服务器12在T1时刻从裸金属管理器上下载到PXE启动引导模块(如PXE boot loader)。第二阶段：由于启动了PXE启动引导模块，裸金属服务器12需要进行第二次IP地址分配过程，如步骤207-210，裸金属服务器获得新的IP地址，接着在步骤211-212，裸金属服务器再次向裸金属管理器发送文件获取请求(可以是根据TFTP服务器地址发送的TFTP请求)，然后在T2时刻裸金属服务器获取到操作系统的镜像。第三阶段：在获取到操作系统镜像之后，在裸金属服务器12上启动这个操作系统(一般来说，所启动的操作系统是一个精简的Linux操作系统)，然后在这个系统中启动特别的应用模块，这个应用模块在T3时刻开始启动运行。第四阶段：特别的应用模块运行后开始收集裸金属服务期的硬件属性，在T4时刻获取到硬件属性，如步骤213。第五阶段：如步骤214，裸金属服务器12通过网络向裸金属管理201发送获取到的硬件属性，裸金属管理器201在T5时刻接收到裸金属服务器12发送的硬件属性。
上述方法存在效率低和可移植性差的问题。启动操作系统本身就需要消耗时间，当批量收集多台裸金属服务器的硬件属性时，通过网络下载操作系统镜像，将导致带宽占用，时延增加。另外，额外运行的用于收集硬件属性的操作系统目前仅能够支持X86类型的裸金属服务器却不能支持ARM类型的裸金属服务器，这导致了可移植性差的问题。
本专利申请基于现有技术的上述问题，提供一种不依赖操作系统即可发现裸金属服务器的硬件属性的系统和方法。本专利申请借助裸金属服务器中的引导系统，在裸金属服务器中的引导系统启动之后，以及在引导系统引导操作系统启动之前，在该引导系统上运行信息收集模块，该信息收集模块收集裸金属服务器的硬件属性，并将收集到的硬件属性通过网络发送给裸金属管理器。
信息收集模块为本申请新提供的功能模块，该模块运行在引导系统启动之后的操作系统加载前阶段，用于实现裸金属服务器硬件属性的发现。一般来说，该信息收集模块可以为小应用或小程序，例如信息收集应用或信息收集程序。一般来说，裸金属服务器上电之后，引导系统被启动，引导系统运行环境包括如下几个阶段：首先是裸金属服务器的初始化阶段，接着是操作系统启动前阶段(也称为操作系统加载前阶段)，最后是操作系统启动阶段(在操作系统启动完成之后，引导系统不再运行)，本申请中的信息收集模块就运行在引导系统运行环境中的操作系统启动前阶段。
如图3A所示本申请所提供的方案中，裸金属管理器201收集裸金属服务器的硬件属性包括四个阶段。第一阶段：在T0时刻，裸金属服务器12上电，裸金属服务器12中的引导系统开始启动，以及裸金属管理器201使能裸金属服务器12的PXE启动方式；然后在T1时刻裸金属服务器12运行信息收集模块(本申请裸金属服务器可从外部加载运行该信息收集模块，也可从内部加载运行该信息收集模块，后续实施例将具体展开描述)。第二阶段：信息收集模块启动阶段，在T2时刻信息收集模块在裸金属服务器12中的引导系统的运行环境中运行(该阶段为引导系统运行环境中的操作系统启动前阶段)。第三阶段：信息收集模块调用引导系统中的信息收集接口获取裸金属服务器12的硬件属性，在T3时刻，信息收集模块获取到裸金属服务器12的硬件属性。第四阶段：裸金属服务器12的硬件属性的上报阶段，信息收集模块将收集到的裸金属服务器12的硬件属性通过网络传递给裸金属管理器201，裸金属管理器201在T4时刻获取到该硬件属性。在执行第四阶段发送硬件属性之前，裸金属服务器12还需要完成IP地址获取的过程，该过程中裸金属服务器12发送DHCP请求以获取IP地址，裸金属管理器201中的DHCP服务器响应该DHCP请求并为裸金属服务器12分配IP地址。
图3B示出了本申请方案发现裸金属服务器硬件属性的流程与裸金属服务器上电后各阶段间的关系，同时也示出了现有技术发现裸金属服务器硬件属性的流程与裸金属服务器上电后各阶段间的关系。从图3B可见，本申请提供的方案信息收集模块运行在引导系统运行环境的操作系统启动前阶段，使得硬件属性的发现可以不依赖于操作系统，不需要启动操作系统即可完成裸金属服务器的硬件属性的获取，并快速发送给裸金属管理器，使得裸金属管理器获得裸金属服务器的硬件属性的时间大大减少，提高了裸金属服务器的部署效率，尤其是对于大型的数据中心需要部署多台裸金属服务器的情况，大大提高了部署效率；同时，由于信息收集模块很小，运行和下载都不需要耗费网络大量的流量，降低了网络拥塞。另外，该引导系统执行上述方法对系统的改进很小，所花费的开发代价也比较小。另外，现有技术的方法只能支持X86类型的裸金属服务器，不能支持高级精简指令集机器(Advanced RISC Machine，ARM)类型的裸金属服务器的问题，原因是现有技术中的用于安装特定软件的操作系统不能应用于ARM类型的裸金属服务器。然而 本申请的方法不依赖操作系统完成硬件属性的发现，解决了上述问题，提高了裸金属服务硬件属性的发现方式通用性。
图4为本申请方案中的裸金属服务器12的组成框图，处理器401中可以包含一个CPU或者多个CPU，第一内部存储器402可以为随机存储器RAM，用于在CPU启动之后，读取指令并使得CPU执行指令。第二内部存储器403可以为只读存储器ROM或者flash，其中，在ROM上可以设置有固件(firmware)，固件是写入ROM中的模块，一般担任着系统的基础和底层的硬件设置和控制工作。本申请实施例中，第二内部存储器403中可以保存有引导系统，引导系统可以是采用UEFI的BIOS或者不采用UEFI的BIOS，该引导系统在裸金属服务器上电后被加载到第一内部存储器402中使得处理器401运行该引导系统，以引导裸金属服务器的操作系统启动。网卡404中配置有PXE客户端(PXE客户端实际是位于网卡内部的ROM中的客户端模块，本申请以PXE客户端设置于网卡中为例，本领域技术人员也可以将PXE客户端集成在引导系统中)，用于与裸金属管理器进行通信，该PXE客户端也会被加载到第一内部存储器402中使得处理器401运行该PXE客户端，PXE客户端被加载入内存后，它同时具有DHCP客户端和TFTP客户端的功能，DHCP客户端会向DHCP服务器请求用来分配给裸金属服务器的IP地址，然后由PXE客户端将放置在远端的文件通过TFTP下载到本地运行；外部存储器405用于保存数据，该裸金属服务器未发放之前，外部存储器405中一般没有保存数据。启动控制器406用于通过智能平台控制接口(IPMI，Intelligent Platform Management Interface)与裸金属管理系统通信，接收裸金属管理系统的启动命令，该启动命令用于上电裸金属服务器，以使得裸金属服务器中的引导系统被启动。进一步，该启动命令还可以用于使能裸金属服务器的PXE启动方式。上述启动控制器406的一种具体实现是基板管理控制器(Baseboard Management Controller，BMC)，基本管理控制器一般用来测量裸金属服务器内部物理变量，例如：温度，湿度，电源电压，风扇速度，通信参数和操作系统函数，如果这些变量任何一个超出了指定限制的范围之外，BMC发出告警通知。本申请中，BMC还可以用来实现上述启动控制器的功能。
关于裸金属服务器的PXE启动方式的设置，在云计算场景下一般由裸金属管理器使能，但特殊的场景下，也可以由人工对裸金属服务器的启动方式进行手动设置或者是PXE启动方式被默认激活。例如裸金属服务器出厂时默认设置为PXE启动方式，裸金属服务器的PXE启动方式在引导系统启动之后被默认激活，此时，裸金属管理器不需要额外使能PXE启动方式。裸金属管理器201使能裸金属服务器的PXE启动方式，可以是裸金属管理器远程将裸金属服务器的启动方式设置为PXE启动方式，也可以是将裸金属服务器出厂时默认设置的非PXE启动方式修改为PXE启动方式。具体地，可以是裸金属管理器可以通过基板管理控制器使能所述裸金属服务器的PXE启动方式。如果裸金属服务器没有预装操作系统，裸金属管理器可以直接设置所述裸金属服务器的启动方式为PXE启动方式，如果裸金属服务器已经预装操作系统，所述裸金属管理器使能裸金属服务器的PXE启动方式可以包括：将所述裸金属服务器的操作系统启动方式由硬盘/光盘启动变更为PXE启动。进一步，在变更启动方式之后，所述裸金属管理器还可以重启所述裸金属服务器。
关于裸金属服务器的上电，在云计算场景下由裸金属管理器通过启动命令来控制，但在一些特殊的场景下，也有可能由人工手动来为裸金属服务器进行上电，此种情况下，本申请的裸金属硬件属性的发现方法仍然可以执行。
裸金属服务器的上电与PXE启动方式的使能可以是同时执行的动作，也可以是先后执行的动作，本领域技术人员可以灵活实现。
本申请对第二内部存储器403中的引导系统可以有两方面的改进，一方面在引导系统中增加信息收集接口，用于收集裸金属服务器的硬件属性，另一方面引导系统用于提供信息收集模块的运行环境，该信息收集模块运行在引导系统环境下，在引导系统引导操作系统启动之前，调用信息收集接口以获取裸金属服务器的硬件属性，并将获取到的裸金属服务器的硬件属性发送给裸金属管理器。其中，该信息收集模块可以是预先设置在裸金属服务器中的第二内部存储器ROM中的固件模块，也可以是PXE阶段从裸金属管理器上下载得到的软件模块。下面就这两种实施方式进行详细介绍。
图5A为本申请的一个具体实施例中裸金属管理器与裸金属服务器组成的云计算系统的简化方框图，图5B为在图5A所示系统中，裸金属管理器获取裸金属服务器的硬件属性的方法流程图，当云计算系统中需要增加服务器资源时，裸金属管理器需要发放裸金属服务器。在裸金属服务器被发放之前，裸金属管理器需要收集裸金属服务器的硬件属性。在本实施例中，信息收集模块预先代替启动引导模块保存在裸金属管理器侧的PXE服务器的TFTP服务器上，即信息收集模块是PXE阶段从裸金属管理器上下载得到。
在步骤501，裸金属管理器201使能裸金属服务器12的PXE启动方式，并给裸金属服务器12上电。具体地，裸金属管理器中的启动管理模块2011可以向裸金属服务器的启动系统发送启动命令，通过控制裸金属服务器中的启动控制器406来分别实现裸金属服务器的上电和PXE启动方式的使能。裸金属服务器12上电后，第二内部存储器403中的引导系统被加载到第一内部存储器402，裸金属服务器12的处理器401执行所述引导系统，所述引导系统被启动。所述引导系统发现PXE启动方式使能后，加载网卡404中的PXE客户端并运行该PXE客户端(当PXE客户端在裸金属服务器上运行后，裸金属服务器12执行PXE客户端的功能，相对的，裸金属管理器201执行PXE服务器的功能)，所述PXE客户端执行获取IP地址的流程。具体地，获取IP地址的流程包括步骤502-505，无论是引导系统还是PXE客户端都是通过裸金属服务器的处理器401来运行的，下述流程将以裸金属服务器为主体描述IP地址获取流程。在步骤502，裸金属服务器12发送DHCP发现消息，该DHCP发现消息一般是广播消息，裸金属管理器201的发现模块2012具备PXE服务器的功能，它会接收到以广播形式发送的DHCP发现消息，在步骤503裸金属管理器201向裸金属服务器12发送DHCP提供(offer)，在步骤504，裸金属服务器12在接收到裸金属管理器201发送的DHCP offer后，向裸金属管理器201发送DHCP请求以获取IP地址，裸金属管理器201接收到该DHCP请求后，为裸金属服务器12分配IP地址。此外，在PXE启动过程中，裸金属管理器201在将IP地址发送给裸金属服务器12的时候或者之后，还可以通知裸金属服务器12待下载文件的信息，该待下载文件的信息用以指示裸金属服务器12从裸金属管理器下载待下载文件，以执行待下载文件的相关功能。该待下载的文件的信息可以包括下载服务器的地址信息或待下载文件的保存地址或待下载文件的名称等信息。一般来说，在PXE启动过程中，PXE客户端和PXE服务器之间 通过TFTP传送文件，上述的下载服务器可以是TFTP服务器，因此，裸金属管理器201通知裸金属服务器12的待下载文件的信息可以是TFTP服务器的地址(下载服务器一般来说设置在裸金属管理器侧，并且在一些情况下，裸金属管理器本身即是下载服务器)。现有PXE技术中，TFTP服务器中保存的待下载文件是启动引导模块(例如PXE boot loader，用于引导启动操作系统)，本申请的实施例中，裸金属管理器201侧的TFTP服务器上不再保存启动引导模块，而是保存信息收集模块，也就是说TFTP服务器上保存的待下载文件为信息收集模块，而不再是启动引导模块。该信息收集模块用于使得裸金属服务器获取裸金属服务器12的硬件属性并发送。因此，在步骤505中，裸金属服务器12接收到裸金属管理器201发送的IP地址以及待下载文件的信息，本具体实施例中待下载文件的信息即信息收集模块的保存地址。
步骤506，裸金属服务器12根据信息收集模块的保存地址从下载服务器上下载信息收集模块，并将信息收集模块加载到裸金属服务器的内存中。具体地，裸金属管理器201中的发现模块2012通知裸金属服务器12所述信息收集模块的保存地址，并接收裸金属服务器12根据所述信息收集模块的保存地址发送的文件获取请求，裸金属管理器201通过网络向裸金属服务器12提供所述信息收集模块。当裸金属服务器12下载到该信息收集模块之后，该信息收集模块在引导系统的运行环境中运行，调用引导系统中的信息收集接口以获取裸金属服务器12的硬件属性。具体地，如果引导系统为UEFI BIOS，信息收集模块可以为UEFI Application，该UEFI Application在UEFI BIOS环境中运行。在裸金属服务器12下载并运行信息收集模块之后，由于原IP地址无法传递给信息收集模块，因此要重新获取IP地址，步骤507-510与上述步骤502-505中获取IP地址流程相同，不再赘述。在重新获取到IP地址之后，裸金属服务器12在步骤511将信息收集阶段获取到的裸金属服务器的硬件属性发送给裸金属管理器201。
需要说明的是，上述裸金属服务器12在步骤506下载到信息收集模块之后，即运行该信息收集模块并启动信息收集，步骤507-510的IP地址获取阶段可以与信息收集阶段同步，裸金属服务器获取到IP地址的时间与收集到硬件属性的时间可以同时，也可以有先后之分，本专利申请的实施例在此无特别要求。裸金属服务器12在获取到IP地址(IP地址用于网络通信)以及硬件属性(硬件属性用于发送到裸金属管理器)之后，执行步骤511的通过网络发送硬件属性。
上述本申请的裸金属管理器获取裸金属服务器的硬件属性的实施例中，裸金属管理器提供信息收集模块，并在裸金属服务器上的引导系统的运行环境中运行，该信息收集模块运行后可调用引导系统新提供的信息收集接口以获取裸金属服务器的硬件属性。由于信息引导模块运行在引导系统所执行的操作系统启动前阶段，因此本发明申请的方案不依赖于操作系统的启动即可完成裸金属服务器的硬件属性的获取，使得裸金属管理器能够更快地获得裸金属服务器的硬件属性，大大提高了裸金属服务器的发放效率，同时由于信息收集模块比较小，运行和下载都不需要耗费网络大量的流量，降低了网络拥塞。另外，不需要对裸金属服务器做太多的改进即可实现裸金属服务器的硬件属性的发现。
在上述实施例的一种更为具体的实现中，裸金属服务器上的引导系统为UEFI BIOS的情况下，信息收集模块可以是UEFI Application，根据UEFI规范，UEFI Application是硬件初始化完成，操作系统启动之前运行的应用，UEFI Application运行在UEFI运行 环境中的系统加载瞬间(Transient System Load，TSL)阶段。UEFI BIOS启动后的整个运行环境包括以下几个阶段：1、安全验证阶段(Security Phase，SEC)，计算设备加电后进入此阶段。2、前EFI初始化(Pre-EFI Initilization，PEI)阶段，用于为下一阶段准备执行环境；3、驱动执行环境(Driver Execution Environment，DXE)阶段，该阶段执行大部分系统初始化工作，进入此阶段时，内存已可以完全被使用，因而此阶段可以进行复杂工作；4、启动设备选择(Boot Device Selection，BDS)阶段，该阶段主要执行启动策略，主要功能包括：初始化控制台设备；加载设备驱动；根据系统设置加载和执行启动项。其中上述1-4阶段可以通称为初始化阶段。5、TSL阶段，此时系统资源仍由UEFI内核控制。TSL阶段的作用是为OS Loader准备执行环境，因此该阶段被称为临时系统。在TSL阶段，当BS的ExitBootServices服务被调用后，系统进入下一个阶段。6、运行(Run Time)阶段，此时系统的控制权从UEFI内核转交到OS Loader，随着OS Loader的运行，OS最终获得系统控制权。此时系统的控制权从UEFI内核转交到OS Loader，随着OS Loader的运行，OS最终获得系统控制权。本发明申请提供的信息收集模块即为运行在第5阶段操作系统加载前的阶段的应用，由此，实现了在操作系统启动前获取硬件属性并发送给裸金属管理器的方法，提高了裸金属服务器的发放效率。
在上述实施例的一种更为具体的实现中，裸金属服务器上的引导系统为UEFI BIOS的情况下，可以在UEFI已有的引导服务(Boot Service)或运行服务(Runtime Service)中增加信息收集接口，UEFI Application调用该信息收集接口以获取裸金属服务器的硬件属性。
UEFI新增信息收集接口描述可以包括：
将裸金属服务器的硬件属性传递给调用者(信息收集模块，如UEFI Application)，硬件属性可由一个典型的结构体封装，也可通过数组等方式传递，具体内容包括但不限于：CPU体系结构、逻辑CPU个数、硬盘大小、内存大小、MAC地址信息等。
UEFI新增信息收集接口定义可以包括：
typedef EFI_STATUS(EFIAPI*EFI_HARDWARE_INFO_GET)(OUT struct*hardware_info)；
struct hardware_info{
unsigned int arch；//体系结构
unsigned int vcpu_num；//系统逻辑CPU个数
unsigned int disk_size；//硬盘大小
unsigned int memory_size；//内存大小
struct mac_info client_macs；//MAC地址信息
…
}；
struct mac_info{
unsigned int port_num；//网络接口数量
unsigned int mac[1]；//具体MAC信息
}；
…
在上述实施例的一种更为具体的实现中，裸金属服务器12发送硬件属性的一种实现方式可以是采用DHCP协议发送，例如采用DHCP选项(DHCP option)携带裸金属服务器的硬件属性。在上述步骤510后，裸金属服务器12获取到裸金属管理器中的DHCP Server分配的IP地址，在信息收集模块获取到裸金属服务器的硬件属性之后，在步骤511中，信息收集模块可再次发送DHCP请求并将收集到的硬件属性构造在DHCP REQUEST报文中的option code中，裸金属管理器201中的发现模块2012(含PXE Server功能)接收到DHCP REQUEST消息后会根据配置将对应option code解析出来，并保存裸金属服务器的硬件属性。根据DHCP协议的option code的预留情况，可以利用预留的code93或code128-code254中的任意预留位来构造裸金属服务器的硬件属性字段，例如将code93定义为裸金属服务器的体系架构字段(client-architecture)，将code128定义为裸金属服务器的系统CPU个数字段(client-vcpu)，将code129定义为裸金属服务器的系统硬盘大小(client-disk)，将code130定义为裸金属服务器的系统内存大小(client-memory)，将code131定义为裸金属服务器的网卡信息(client-macs)等等。
如果采用DHCP协议发送硬件属性，裸金属服务器可以根据实际获取到硬件属性的时间来决定采用哪一条DHCP消息来携带硬件属性，在上述实施例中的步骤511中，信息收集模块再次发送DHCP请求以携带裸金属服务器的硬件属性，实际上，也可以在步骤507或步骤509中的消息中携带硬件属性。
另外，裸金属服务器12发送硬件属性的另一种实现方式可以是采用EFI的代表性状态传输(Representational State Transfer，REST)协议消息发送。在此种情况下，对端，即裸金属管理器201中的发现模块2012中包含解析REST协议的功能，用于对REST报文中携带的裸金属服务器的硬件属性进行接收和解析。
进一步，在本发明申请的一个更为优化的实施例中，除了在引导系统中新增信息收集接口之外，还可以再新增一个信息设置接口，该接口用于对裸金属服务器的硬件属性进行修改，例如，修改基板管理控制器(Baseboard Management Controller，BMC)的IP地址信息。与上述信息收集接口类似，该信息设置接口同样可以在UEFI已有的引导服务或运行服务中增加，相应地，裸金属管理器还可以用于向裸金属服务器发送硬件属性修改请求；信息收集模块还可以增加信息设置功能，用于调用该信息设置接口对裸金属服务器的硬件属性进行设置。信息设置接口具体定义可以如下：
typedef EFI_STATUS(EFIAPI*EFI_HARDWARE_INFO_SET)(IN struct*hardware_info)
本领域技术人员可以明了，上述信息设置接口除了可以与信息收集接口共同存在之外，也可以单独存在，也就是说，硬件属性的修改可以不依赖于硬件属性的收集，在操作系统启动之前，也可以采用本申请相关的技术方法实现硬件属性的直接设置，即裸金属管理器在收集到裸金属服务器的硬件属性之前，就修改裸金属服务器的硬件属性。
图6A为本申请的又一个具体实施例中裸金属管理器与裸金属服务器组成的计算机系统的简化方框图，图6B为在图6A所示系统中，裸金属管理器获取裸金属服务器的硬件属性的方法流程图。本实施例中，信息收集模块预先已经保存在裸金属服务器的ROM中，具体地，例如可以集成在引导系统中，也就是说，通过改进裸金属服务器的引导系统，使其新增信息收集功能模块。
在步骤601，裸金属管理器201使能裸金属服务器12的PXE启动方式，并给裸金属服务器12上电。步骤602，裸金属服务器12上电后，第二内部存储器403中的引导系统被加载到第一内部存储器402，裸金属服务器12的处理器401执行所述引导系统，所述引导系统被启动。引导系统发现PXE使能后，加载网卡404中的PXE客户端并运行该PXE客户端。进一步，本实施例中，裸金属服务器12还包括信息收集开关，该信息收集开关用于指示所述引导系统是否运行所述信息收集模块，引导系统根据该信息收集开关的指示确定是否加载所述信息收集模块。具体实现时，该信息收集开关可以由裸金属管理器通过BMC来进行控制开关，例如裸金属管理器确定需要裸金属服务器的硬件属性的时候，或者裸金属服务器待发放的时候，将所述信息收集开关打开。进一步，引导系统在PXE客户端运行到向所述裸金属管理器201中的TFTP服务器下载PXE操作系统引导模块之前，引导系统根据所述信息收集开关的指示运行所述信息收集模块，以完成裸金属服务器的硬件属性的发现。本实施例中，信息收集开关的指示为需要加载信息收集模块。因此当引导系统运行到操作系统加载前阶段，信息收集模块在引导系统的环境中开始运行，并执行裸金属服务器的硬件属性的收集工作，信息收集模块调用引导系统中的信息收集接口以获取裸金属服务器12的硬件属性。具体地，如果引导系统为UEFI BIOS，信息收集模块可以为UEFI Application，该UEFI Application在UEFI BIOS环境中运行。步骤603-606，由于PXE客户端的运行，裸金属服务器还要与裸金属管理器进行IP地址获取的流程，该流程在上述实施例中已经描述过，在此不再赘述。在获取到IP地址之后，裸金属服务器12在步骤607将信息收集阶段获取到的裸金属服务器的硬件属性发送给裸金属管理器201。
在本实施例中，信息收集模块除了存储位置不同之外，其它的属性以及具体实现方式和功能，例如向裸金属管理器发送硬件属性的方式，均可以与上述从裸金属管理器侧下载的实施方式一致，本实施例不再具体展开描述。
上述实施例中，信息收集模块的存储位置不在裸金属管理器侧，而是保存在裸金属服务器中，在引导系统运行时，在操作系统加载前阶段，该信息收集模块可以不用从远端下载即可加载执行，从而实现裸金属服务器硬件属性的发现。本实施例中由于不再需要从裸金属管理器下载信息收集模块，进一步节约了执行时间，提高了裸金属管理器获得裸金属服务器硬件属性的效率，从而也提高了裸金属管理器发放裸金属服务器的效率。
本申请中的裸金属服务器一般未安装操作系统，对于此类未安装操作系统的裸金属服务器，裸金属管理器可以直接使能PXE启动方式并上电裸金属服务器以实施后续的收集硬件属性并发送给裸金属管理器的方法，对于已经安装了操作系统的裸金属服务器，裸金属管理器可以修改裸金属服务器的启动模式之后再重启裸金属服务器。
最后，应了解上述实施例仅仅用于阐释，本申请的技术方案不限于此。尽管参考上述优选实施例对本申请进行详细描述，但是应了解，所属领域的技术人员可在不脱离本申请所附权利要求书的范围的情况下，做出各种修改、变更或替换。

Claims (40)

  1. 一种云计算系统，其特征在于，包括裸金属管理器和裸金属服务器；
    所述裸金属管理器，用于上电所述裸金属服务器，以及接收所述裸金属服务器通过网络发送的硬件属性；
    所述裸金属服务器包括引导系统，在所述裸金属服务器上电之后所述引导系统被启动；
    所述引导系统用于运行预启动执行环境(PXE)客户端并通过PXE启动方式引导操作系统启动，以及在所述操作系统启动之前运行信息收集模块，其中，所述信息收集模块，用于收集所述裸金属服务器的硬件属性，并将所述裸金属服务器的硬件属性通过网络发送给所述裸金属管理器。
  2. 如权利要求1所述的系统，其特征在于，所述裸金属管理器还用于使能所述裸金属服务器的PXE启动方式。
  3. 如权利要求1或2所述的系统，其特征在于，所述引导系统包括信息收集接口；
    所述信息收集模块用于调用所述信息收集接口以获取所述裸金属服务器的硬件属性。
  4. 如权利要求3所述的系统，其特征在于，所述引导系统还包括信息设置接口；
    所述裸金属管理器还用于通过网络向所述裸金属服务器发送硬件属性修改请求；所述信息收集模块还用于根据所述硬件属性修改请求，调用所述信息设置接口对所述裸金属服务器的硬件属性进行修改。
  5. 如权利要求1-4任一项所述的系统，其特征在于，所述信息收集模块预先保存在所述裸金属管理器；
    所述裸金属服务器中的PXE客户端用于获取为所述裸金属服务器分配的网际协议(IP)地址，使用所述IP地址与所述裸金属管理器进行通信，并从所述裸金属管理器下载所述信息收集模块。
  6. 如权利要求5所述的系统，其特征在于，所述裸金属管理器还用于向所述裸金属服务器发送所述信息收集模块的保存地址。
  7. 如权利要求6所述的系统，其特征在于，所述信息收集模块的保存地址为所述裸金属管理器中的文件传输协议(FTP)服务器的地址，所述信息收集模块代替PXE启动引导模块保存在所述FTP服务器。
  8. 如权利要求1-4任一项所述的系统，其特征在于，所述信息收集模块预先保存在所述裸金属服务器，所述裸金属服务器的引导系统在所述操作系统启动前阶段从所述只读存储器中加载并运行所述信息收集模块。
  9. 如权利要求8所述的系统，其特征在于，所述裸金属服务器还包括信息收集开关，所述信息收集开关用于指示所述引导系统是否运行所述信息收集模块；
    如果所述信息收集开关指示运行所述信息收集模块，所述裸金属服务器中的引导系统用于加载所述信息收集模块。
  10. 如权利要求9所述的系统，其特征在于，在所述PXE客户端运行之后并且在向所述裸金属管理器中的FTP服务器下载PXE启动引导模块之前，所述引导系统根据所述信息收集开关的指示运行所述信息收集模块。
  11. 如权利要求1-10任一项所述系统，其特征在于，所述裸金属服务器中的信息 收集模块用动态主机设定协议(DHCP)消息或代表性状态传输REST消息携带所述裸金属服务器的硬件属性。
  12. 一种裸金属服务器的硬件属性发现方法，其特征在于，应用于裸金属管理器，所述裸金属管理器对裸金属服务器进行管理；所述方法包括：上电所述裸金属服务器；以及，
    接收所述裸金属服务器通过网络发送的所述裸金属服务器的硬件属性，其中，所述裸金属服务器中的引导系统用于运行预启动执行环境PXE客户端并通过PXE启动方式引导操作系统启动，以及在所述裸金属服务器的操作系统启动之前运行信息收集模块，所述信息收集模块用于收集并通过网络发送所述裸金属服务器的硬件属性。
  13. 如权利要求12所述的方法，其特征在于，所述方法还包括：使能所述裸金属服务器的PXE启动方式。
  14. 如权利要求12或13所述的方法，其特征在于，所述方法还包括：
    通过网络向所述裸金属服务器发送硬件属性修改请求。
  15. 如权利要求12-14任一项所述的方法，其特征在于，所述裸金属管理器还用于保存所述信息收集模块；
    在所述接收所述裸金属服务器发送的所述裸金属服务器的硬件属性之前，所述方法还包括：
    通过网络向所述裸金属服务器提供所述信息收集模块。
  16. 如权利要求15所述的方法，其特征在于，所述方法还包括通知所述裸金属服务器所述信息收集模块的保存地址，所述信息收集模块的保存地址为所述裸金属管理中的文件传输协议FTP服务器的地址，所述信息收集模块代替PXE启动引导模块保存在所述FTP服务器。
  17. 如权利要求12-16任一项所述方法，其特征在于，所述接收所述裸金属服务器通过网络发送的所述裸金属服务器的硬件属性包括：
    接收所述裸金属服务器发送的动态主机设定协议(DHCP)消息或代表性状态传输(REST)消息，获取所述DHCP协议消息或所述REST消息中携带的所述裸金属服务器的硬件属性。
  18. 一种管理装置，其特征在于，用于对裸金属服务器进行管理；所述管理装置包括：
    存储设备，用于存储指令；
    以及
    至少一台处理器，与所述存储设备耦合；
    其中，当所述至少一台处理器执行所述指令时，所述指令致使所述处理器执行权利要求12-17任一项所述的方法。
  19. 一种裸金属管理器，其特征在于，用于对裸金属服务器进行管理，所述管理器包括：
    启动管理模块，用于上电所述裸金属服务器；
    发现模块，用于接收所述裸金属服务器通过网络发送的所述裸金属服务器的硬件属性，其中，所述裸金属服务器中的引导系统用于运行预启动执行环境(PXE)客户端并通过PXE启动方式引导操作系统启动，以及在所述裸金属服务器的操作系统启动之前运 行信息收集模块，所述信息收集模块用于收集并通过网络发送所述裸金属服务器的硬件属性。
  20. 如权利要求19所述的管理器，其特征在于，所述启动管理模块还用于使能所述裸金属服务器的PXE启动方式。
  21. 如权利要求19或20所述的管理器，其特征在于，所述发现模块，还用于通过网络向所述裸金属服务器发送硬件属性修改请求。
  22. 如权利要求20或21所述的管理器，其特征在于，所述发现模块还用于保存所述信息收集模块，以及通过网络向所述裸金属服务器提供所述信息收集模块。
  23. 如权利要求19-22所述的管理器，其特征在于，所述发现模块具体用于接收所述裸金属服务器发送的动态主机设定协议(DHCP)消息，通过所述发现模块中的DHCP服务器获取所述DHCP协议消息中携带的所述裸金属服务器的硬件属性；或者，所述发现模块具体用于接收所述裸金属服务器发送的代表性状态传输(REST)协议消息，通过所述发现模块中的解析模块获取所述REST协议消息中携带的所述裸金属服务器的硬件属性。
  24. 一种裸金属服务器的硬件属性发现方法，其特征在于，应用于裸金属服务器，裸金属管理器对裸金属服务器进行管理；所述方法包括：
    接收所述裸金属管理器的启动命令，根据所述启动命令进行上电操作，在所述裸金属服务器上电后，所述裸金属服务器中的引导系统被启动，所述引导系统通过预启动执行环境(PXE)启动方式引导操作系统启动；
    在操作系统启动之前，收集所述裸金属服务器的硬件属性，并将所述裸金属服务器的硬件属性通过网络发送给所述裸金属管理器。
  25. 如权利要求24所述的方法，其特征在于，在所述收集所述裸金属服务器的硬件属性之前还包括：
    获取为所述裸金属服务器分配的网际协议(IP)地址，使用所述IP地址与所述裸金属管理器进行通信，并从所述裸金属管理器下载信息收集模块，以及在所述引导系统中运行所述信息收集模块。
  26. 如权利要求24所述的方法，其特征在于，在所述收集所述裸金属服务器的硬件属性之前还包括：
    所述引导系统在所述操作系统启动前阶段从所述裸金属服务器中加载并运行信息收集模块。
  27. 如权利要求24至26任一项所述的方法，其特征在于，所述收集所述裸金属服务器的硬件属性包括：
    所述信息收集模块调用所述引导系统中的信息收集接口获取所述裸金属服务器的硬件属性。
  28. 如权利要求27所述的方法，其特征在于，所述方法还包括：
    接收所述裸金属管理器发送的硬件属性修改请求，根据所述硬件属性修改请求，所述信息收集模块调用所述裸金属服务器中的信息设置接口，对所述裸金属服务器的硬件属性进行修改。
  29. 如权利要求24-28任一项所述的方法，其特征在于，所述将所述裸金属服务器的硬件属性通过网络发送给所述裸金属管理器包括：
    将所述裸金属服务器的硬件属性携带在动态主机设定协议(DHCP)消息或者代表性状态传输(REST)协议消息中发送给所述裸金属管理器。
  30. 一种裸金属服务器，其特征在于，所述服务器包括：启动控制器、中央处理器、第一内部存储器、第二内部存储器以及网卡；
    所述启动控制器，用于接收所述裸金属管理器的启动命令，根据所述启动命令进行上电操作；
    所述第二内部存储器中保存有引导系统，所述裸金属服务器上电之后所述引导系统被加载到所述第一内部存储器中，所述中央处理器运行所述引导系统；所述引导系统用于运行预启动执行环境(PXE)客户端并通过PXE启动方式引导操作系统启动，以及在所述操作系统启动之前运行信息收集模块；
    所述信息收集模块，用于收集所述裸金属服务器的硬件属性，并将所述裸金属服务器的硬件属性通过所述网卡发送给所述裸金属管理器。
  31. 如权利要求30所述的裸金属服务器，其特征在于，所述引导系统包括信息收集接口；
    所述信息收集模块通过调用所述信息收集接口获取所述裸金属服务器的硬件属性。
  32. 如权利要求31所述的裸金属服务器，其特征在于，所述引导系统还包括信息设置接口；
    所述网卡还用于接收所述裸金属管理器发送的硬件属性修改请求，所述信息收集模块还用于根据所述硬件属性修改请求，调用所述信息设置接口对所述裸金属服务器的硬件属性进行修改。
  33. 如权利要求30-32任一项所述的裸金属服务器，其特征在于，所述信息收集模块预先保存在所述裸金属管理器；
    所述PXE客户端用于获取为所述裸金属服务器分配的网际协议(IP)地址，使用所述IP地址与所述裸金属管理器进行通信，并通过所述网卡从所述裸金属管理器下载所述信息收集模块。
  34. 如权利要求30-32任一项所述的裸金属服务器，其特征在于，所述信息收集模块预先保存在所述裸金属服务器的第二内部存储器中，所述裸金属服务器的引导系统用于在所述操作系统启动前阶段从所述第二内部存储器中加载并运行所述信息收集模块。
  35. 如权利要求34所述的裸金属服务器，其特征在于，所述裸金属服务器还包括信息收集开关，所述信息收集开关用于指示所述引导系统是否运行所述信息收集模块；
    如果所述信息收集开关指示运行所述信息收集模块，所述引导系统从所述第二内部存储器中加载所述信息收集模块到第一内部存储器中并运行所述信息收集模块。
  36. 如权利要求35所述的裸金属服务器，其特征在于，在所述PXE客户端运行后，以及所述PXE客户端向所述裸金属管理器中的FTP服务器下载PXE启动引导模块之前，所述引导系统根据所述信息收集开关的指示运行所述信息收集模块。
  37. 一种计算机程序产品，所述计算机程序产品保存在可读存储介质上，其特征在于，所述计算机程序产品包含指令，所述指令运行在计算机中的引导系统的运行环境下，并在所述计算机的操作系统启动前被运行，所述指令致使处理器执行以下操作：
    调用所述引导系统中的信息收集接口，收集所述裸金属服务器的硬件属性信息；以及，
    通过网络将所述裸金属服务器的硬件属性信息发送给裸金属管理器。
  38. 一种计算机系统，其特征在于，包括管理器与服务器；
    所述服务器包括引导系统，在所述服务器上电之后，所述引导系统被启动；所述引导系统，用于运行预启动执行环境PXE客户端并通过PXE启动方式引导操作系统启动，以及在所述操作系统启动之前运行信息收集模块；所述信息收集模块，用于收集所述服务器的硬件属性，并将所述服务器的硬件属性通过网络发送给所述管理器；
    所述管理器用于接收所述服务器发送的所述服务器的硬件属性。
  39. 如权利要求38所述的系统，其特征在于，所述信息收集模块预先保存在所述管理器；
    所述服务器中的PXE客户端用于获取为所述服务器分配的网际协议(IP)地址，使用所述IP地址与所述管理器进行通信，并从所述管理器下载所述信息收集模块。
  40. 如权利要求38所述的系统，其特征在于，所述信息收集模块预先保存在所述服务器中，所述服务器的引导系统在所述操作系统启动前阶段加载并运行所述信息收集模块。
PCT/CN2017/108949 2017-11-01 2017-11-01 裸金属服务器的硬件属性发现方法和系统 WO2019084853A1 (zh)

Priority Applications (4)

Application Number Priority Date Filing Date Title
CN201780089457.6A CN110495136B (zh) 2017-11-01 2017-11-01 裸金属服务器的硬件属性发现方法和系统
PCT/CN2017/108949 WO2019084853A1 (zh) 2017-11-01 2017-11-01 裸金属服务器的硬件属性发现方法和系统
EP17930891.1A EP3684009A4 (en) 2017-11-01 2017-11-01 PROCESS AND SYSTEM FOR DISCOVERING HARDWARE ATTRIBUTES OF A BARE METAL SERVER
US16/860,938 US20200257519A1 (en) 2017-11-01 2020-04-28 Method for discovering hardware attribute of bare metal server and system

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
PCT/CN2017/108949 WO2019084853A1 (zh) 2017-11-01 2017-11-01 裸金属服务器的硬件属性发现方法和系统

Related Child Applications (1)

Application Number Title Priority Date Filing Date
US16/860,938 Continuation US20200257519A1 (en) 2017-11-01 2020-04-28 Method for discovering hardware attribute of bare metal server and system

Publications (1)

Publication Number Publication Date
WO2019084853A1 true WO2019084853A1 (zh) 2019-05-09

Family

ID=66331246

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2017/108949 WO2019084853A1 (zh) 2017-11-01 2017-11-01 裸金属服务器的硬件属性发现方法和系统

Country Status (4)

Country Link
US (1) US20200257519A1 (zh)
EP (1) EP3684009A4 (zh)
CN (1) CN110495136B (zh)
WO (1) WO2019084853A1 (zh)

Cited By (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2021151352A1 (zh) * 2020-10-23 2021-08-05 平安科技（深圳）有限公司 裸金属服务交付方法、装置、计算机设备和存储介质
CN113746673A (zh) * 2021-08-24 2021-12-03 济南浪潮数据技术有限公司 裸金属服务器ipxe的部署方法、装置、设备和介质
CN114024853A (zh) * 2021-10-27 2022-02-08 浪潮电子信息产业股份有限公司 一种裸金属节点的硬件信息收集方法、系统及相关组件
WO2023050666A1 (zh) * 2021-09-30 2023-04-06 苏州浪潮智能科技有限公司 一种裸金属检测方法、系统及相关组件

Families Citing this family (35)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US11947489B2 (en) 2017-09-05 2024-04-02 Robin Systems, Inc. Creating snapshots of a storage volume in a distributed storage system
US11392363B2 (en) 2018-01-11 2022-07-19 Robin Systems, Inc. Implementing application entrypoints with containers of a bundled application
US11748203B2 (en) 2018-01-11 2023-09-05 Robin Systems, Inc. Multi-role application orchestration in a distributed storage system
US11582168B2 (en) 2018-01-11 2023-02-14 Robin Systems, Inc. Fenced clone applications
US11829478B2 (en) * 2019-01-08 2023-11-28 Oracle International Corporation Full server recovery architecture for cloud bare metal instances
US11086725B2 (en) 2019-03-25 2021-08-10 Robin Systems, Inc. Orchestration of heterogeneous multi-role applications
US11256434B2 (en) 2019-04-17 2022-02-22 Robin Systems, Inc. Data de-duplication
US11226847B2 (en) 2019-08-29 2022-01-18 Robin Systems, Inc. Implementing an application manifest in a node-specific manner using an intent-based orchestrator
US11520650B2 (en) 2019-09-05 2022-12-06 Robin Systems, Inc. Performing root cause analysis in a multi-role application
US11249851B2 (en) 2019-09-05 2022-02-15 Robin Systems, Inc. Creating snapshots of a storage volume in a distributed storage system
US11113158B2 (en) 2019-10-04 2021-09-07 Robin Systems, Inc. Rolling back kubernetes applications
US11347684B2 (en) 2019-10-04 2022-05-31 Robin Systems, Inc. Rolling back KUBERNETES applications including custom resources
US11403188B2 (en) 2019-12-04 2022-08-02 Robin Systems, Inc. Operation-level consistency points and rollback
CN111078305A (zh) * 2019-12-11 2020-04-28 秒针信息技术有限公司 信息采集方法、装置、服务器和信息管理系统
CN111324496B (zh) * 2020-02-20 2023-05-26 山东汇贸电子口岸有限公司 对裸机进行检测的方法及装置
CN111427749B (zh) * 2020-04-01 2023-07-11 山东汇贸电子口岸有限公司 一种openstack环境下针对ironic服务的监控工具及方法
US11108638B1 (en) * 2020-06-08 2021-08-31 Robin Systems, Inc. Health monitoring of automatically deployed and managed network pipelines
US11528186B2 (en) * 2020-06-16 2022-12-13 Robin Systems, Inc. Automated initialization of bare metal servers
CN111782283A (zh) * 2020-06-28 2020-10-16 北京思特奇信息技术股份有限公司 一种裸金属服务器的自动化管理方法、装置及介质
CN113965530B (zh) * 2020-07-01 2023-10-13 中移(苏州)软件技术有限公司 一种服务启动方法、装置和存储介质
US11740980B2 (en) 2020-09-22 2023-08-29 Robin Systems, Inc. Managing snapshot metadata following backup
CN114363295A (zh) * 2020-09-28 2022-04-15 华为云计算技术有限公司 一种租户服务器的管理方法及装置
US11743188B2 (en) 2020-10-01 2023-08-29 Robin Systems, Inc. Check-in monitoring for workflows
US11271895B1 (en) 2020-10-07 2022-03-08 Robin Systems, Inc. Implementing advanced networking capabilities using helm charts
US11456914B2 (en) 2020-10-07 2022-09-27 Robin Systems, Inc. Implementing affinity and anti-affinity with KUBERNETES
US11750451B2 (en) 2020-11-04 2023-09-05 Robin Systems, Inc. Batch manager for complex workflows
CN112328312B (zh) * 2020-11-06 2023-06-13 上海云轴信息科技有限公司 一种实现裸机设备无盘启动的方法与设备
US11556361B2 (en) 2020-12-09 2023-01-17 Robin Systems, Inc. Monitoring and managing of complex multi-role applications
CN112866017B (zh) * 2021-01-08 2022-07-08 苏州浪潮智能科技有限公司 配置裸金属服务器bmc ip地址的方法、系统、介质及设备
US11456987B1 (en) 2021-05-07 2022-09-27 State Farm Mutual Automobile Insurance Company Systems and methods for automatic internet protocol address management
CN113612702B (zh) * 2021-09-30 2021-12-24 苏州浪潮智能科技有限公司 一种实现裸金属检查流程的方法、系统、设备和存储介质
CN114157639B (zh) * 2021-11-26 2023-07-14 苏州浪潮智能科技有限公司 服务器智能网卡信息采集的方法、装置、设备及可读介质
CN114244703B (zh) * 2022-02-24 2022-05-10 苏州浪潮智能科技有限公司 一种裸金属服务器检查及部署方法、装置、设备、介质
CN115080479B (zh) * 2022-06-14 2024-03-26 阿里巴巴（中国）有限公司 传输方法、服务器、设备、裸金属实例及基板管理控制器
CN117093454A (zh) * 2023-10-17 2023-11-21 杭州优云科技有限公司 服务器信息管理方法、装置、计算机设备及存储介质

Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US8326972B2 (en) * 2008-09-26 2012-12-04 Red Hat, Inc. Methods and systems for managing network connections in a software provisioning environment
CN104767649A (zh) * 2015-04-30 2015-07-08 杭州华三通信技术有限公司 部署裸金属服务器的方法及装置
CN106888258A (zh) * 2017-02-22 2017-06-23 郑州云海信息技术有限公司 一种服务器信息录入方法和系统

Family Cites Families (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US9032200B2 (en) * 2013-03-14 2015-05-12 American Megatrends, Inc. Method of installing virtual service processor (VSP) to computing device and application of the same
CN106775797A (zh) * 2016-01-28 2017-05-31 新华三技术有限公司 一种操作系统安装状态通知方法及装置

Patent Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US8326972B2 (en) * 2008-09-26 2012-12-04 Red Hat, Inc. Methods and systems for managing network connections in a software provisioning environment
CN104767649A (zh) * 2015-04-30 2015-07-08 杭州华三通信技术有限公司 部署裸金属服务器的方法及装置
CN106888258A (zh) * 2017-02-22 2017-06-23 郑州云海信息技术有限公司 一种服务器信息录入方法和系统

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
See also references of EP3684009A4 *

Cited By (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2021151352A1 (zh) * 2020-10-23 2021-08-05 平安科技（深圳）有限公司 裸金属服务交付方法、装置、计算机设备和存储介质
CN113746673A (zh) * 2021-08-24 2021-12-03 济南浪潮数据技术有限公司 裸金属服务器ipxe的部署方法、装置、设备和介质
CN113746673B (zh) * 2021-08-24 2023-03-24 济南浪潮数据技术有限公司 裸金属服务器ipxe的部署方法、装置、设备和介质
WO2023050666A1 (zh) * 2021-09-30 2023-04-06 苏州浪潮智能科技有限公司 一种裸金属检测方法、系统及相关组件
CN114024853A (zh) * 2021-10-27 2022-02-08 浪潮电子信息产业股份有限公司 一种裸金属节点的硬件信息收集方法、系统及相关组件
CN114024853B (zh) * 2021-10-27 2024-05-28 广东浪潮智慧计算技术有限公司 一种裸金属节点的硬件信息收集方法、系统及相关组件

Also Published As

Publication number Publication date
CN110495136B (zh) 2021-11-19
EP3684009A1 (en) 2020-07-22
CN110495136A (zh) 2019-11-22
US20200257519A1 (en) 2020-08-13
EP3684009A4 (en) 2020-10-21

Similar Documents

Publication Publication Date Title
WO2019084853A1 (zh) 裸金属服务器的硬件属性发现方法和系统
US8533735B2 (en) System for execution context isolation in response to invoking a BIOS kernel function during a driver execution environment (DXE) phase of boot-up of a computer
US9158567B2 (en) System and method for reconfigurable network services using modified network configuration with modified bandwith capacity in dynamic virtualization environments
US10338935B2 (en) Customizing program logic for booting a system
US9996142B2 (en) Selective power management for pre-boot firmware updates
US9176744B2 (en) Quickly provisioning a virtual machine by identifying a path to a differential file during pre-boot
US7058797B2 (en) Use of off-motherboard resources in a computer system
US7243224B2 (en) Preboot execution bootloading
US20090150463A1 (en) Method of migration between virtual machine and physical machine and machine system thereof
JP2008305034A (ja) デバイス割り当て変更方法
JP6111181B2 (ja) 計算機の制御方法及び計算機
CN116243996B (zh) 业务的运行切换方法、装置、存储介质及电子装置
CN115357310A (zh) 系统启动方法、装置、电子设备和存储介质
US11797325B2 (en) Running arbitrary binaries as unikernels on embedded processors
JP2009211640A (ja) 無線ｌａｎを使用したネットワークブート方法およびネットワークブート用端末
CN109600439B (zh) 基于微服务的PaaS平台的部署方法及PaaS平台
US10572151B2 (en) System and method to allocate available high bandwidth memory to UEFI pool services
EP3543849A1 (en) Driver management method and host machine
US9544267B2 (en) Out-of band configuration of IP addresses
JP2003029975A (ja) Ｏｓ立上げ方法、ｏｓ立上げシステム及びｏｓ立上げのためのプログラム
JP5619999B2 (ja) ユーティリティプログラムを実行する方法、コンピュータシステム及びコンピュータプログラムプロダクト
US20240134654A1 (en) Network interface device booting one or more devices
US20240020103A1 (en) Parallelizing data processing unit provisioning
US20230325203A1 (en) Provisioning dpu management operating systems using host and dpu boot coordination
US20240036896A1 (en) Generating installation images based upon dpu-specific capabilities

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 17930891

Country of ref document: EP

Kind code of ref document: A1

ENP Entry into the national phase

Ref document number: 2017930891

Country of ref document: EP

Effective date: 20200417

NENP Non-entry into the national phase

Ref country code: DE