[go: up one dir, main page]

WO2016015443A1 - Method, physical host and system for live migration of virtual machine - Google Patents

Method, physical host and system for live migration of virtual machine Download PDF

Info

Publication number
WO2016015443A1
WO2016015443A1 PCT/CN2014/095013 CN2014095013W WO2016015443A1 WO 2016015443 A1 WO2016015443 A1 WO 2016015443A1 CN 2014095013 W CN2014095013 W CN 2014095013W WO 2016015443 A1 WO2016015443 A1 WO 2016015443A1
Authority
WO
WIPO (PCT)
Prior art keywords
host
virtual machine
host adapter
data
adapter
Prior art date
Legal status (The legal status 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 status listed.)
Ceased
Application number
PCT/CN2014/095013
Other languages
French (fr)
Chinese (zh)
Inventor
黄凯明
仪长
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Huawei Technologies Co Ltd
Original Assignee
Huawei Technologies Co Ltd
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 Huawei Technologies Co Ltd filed Critical Huawei Technologies Co Ltd
Publication of WO2016015443A1 publication Critical patent/WO2016015443A1/en
Anticipated expiration legal-status Critical
Ceased legal-status Critical Current

Links

Images

Classifications

    • GPHYSICS
    • G06COMPUTING OR CALCULATING; COUNTING
    • G06FELECTRIC DIGITAL DATA PROCESSING
    • G06F9/00Arrangements for program control, e.g. control units
    • G06F9/06Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
    • G06F9/44Arrangements for executing specific programs
    • G06F9/455Emulation; Interpretation; Software simulation, e.g. virtualisation or emulation of application or operating system execution engines

Definitions

  • the present invention relates to the field of communications, and in particular, to a method, a physical host, and a system for virtual machine hot migration.
  • Computer virtualization technology is the current hot spot, and the virtual migration technology of virtual machine is a hot spot in virtualization technology. It can be said that the hot migration technology of virtual machine has become a test indicator to measure the advantages and disadvantages of computer virtualization technology.
  • the so-called virtual machine hot migration refers to the migration of the running virtual machine from the source physical host to the destination physical host while ensuring the normal operation of the service on the virtual machine.
  • a typical virtual machine application architecture that is commonly used today is a physical host that has a virtual machine manager (VMM) that manages virtual machine functions. Specifically, one or more virtual machines are managed by the VMM.
  • the operating system OS and various applications can be run on a virtual machine.
  • the hardware of a general physical host mainly includes: a part of hardware (such as CPU, memory, hard disk, etc.) that is strongly related to running a virtual machine, and a part of which is not directly used to run a virtual machine, but is used to perform some specific services, and the functions are relatively independent. Hardware (for example, Ethernet network card and host adapter).
  • the host adapter usually includes a single physical input SR-IOV (Single Root I/O Virtualization HBA/HCA standard implementation of a physical function PF and one or more virtual functions (VF), one of which The virtual function can be directly connected to a virtual machine corresponding to it.
  • SR-IOV Single Root I/O Virtualization HBA/HCA standard implementation of a physical function PF and one or more virtual functions (VF), one of which The virtual function can be directly connected to a virtual machine corresponding to it.
  • HBA/HCA Host Bus Adaptor
  • HBA/HCA pass-through virtual machine mode the virtual machine hot migration of the Small Computer System Interface (SCSI) service is not supported, that is, since the session control of the protocol stack is implemented by the HBA/HCA hardware, The session of the protocol stack cannot be migrated with the virtual machine.
  • SCSI Small Computer System Interface
  • SCSI has a very low tolerance for packet loss. Therefore, the continuous loss of the session context of the protocol stack saved on the HBA/HCA causes the interruption of the upper layer service and the virtual machine.
  • the hot migration is unsuccessful, so virtual machine hot migration is not possible in the HBA/HCA pass-through virtual machine mode.
  • the prior art allocates an emulated Ethernet network card device (Ethernet Network Interface Card) capable of running the virtual machine to the virtual machine running on the HBA/HCA of the source physical host through the multi-path technology, that is, through the hypervisor.
  • Ethernet Network Interface Card an emulated Ethernet network card device capable of running the virtual machine to the virtual machine running on the HBA/HCA of the source physical host through the multi-path technology, that is, through the hypervisor.
  • the source physical host can disconnect the HBA/HCA from the virtual machine, bind the virtual machine to the created emulation Eth NIC, and then pass The simulation path between the virtual machine and the simulated Eth NIC migrates the stored service data in the virtual machine to the destination physical host.
  • the inventor found that in the above-mentioned virtual machine migration process, since the source physical host is still unable to migrate the session connection context of the protocol stack saved on the HBA/HCA, the storage is lost in the HBA/HCA when the virtual machine is migrated.
  • the session context in the network which in turn causes the virtual machine to fail to complete the hot migration.
  • the embodiments of the present invention provide a method, a physical host, and a system for virtual machine hot migration, which are used to solve the problem that the hot migration of the virtual machine cannot be implemented in the HBA/HCA through virtual machine mode in the prior art.
  • an embodiment of the present invention provides a virtual machine hot migration method, where the first physical host includes a host and a host adapter connected to the host, where the host adapter is a match.
  • a single input and output virtualized SR-IOV standard hardware device including one or more physical functions PF and one or more virtual functions VF; the host runs a virtual machine and a virtual machine manager VMM, in the virtual machine Running with a guest operating system, the guest operating system passes the match
  • the pass-through technology of the SR-IOV standard directly accesses the VF;
  • the VMM also includes an emulated host adapter implemented in software, the method comprising:
  • the data corresponding to the virtual machine and the data corresponding to the emulated host adapter are migrated to the second physical host.
  • the data corresponding to the virtual machine includes data corresponding to the device simulated by the virtual machine and running data information; the data corresponding to the simulated host adapter includes the Simulate the data of the host adapter itself and the data run by the emulation adapter.
  • the host adapter includes an Internet Small Computer System Interface (ISCSI) HBA, and the interaction with the service sender enables The service sender suspends the transmission of service data including:
  • ISCSI Internet Small Computer System Interface
  • the service sender Transmitting, by the service sender, the transmission control protocol TCP packet information with the window field being 0, so that the service sender pauses the transmission of the service data; wherein the TCP packet information of the window domain is 0 is used to indicate the service transmission.
  • the party implements the suspension of the transmission of the service data within a preset time period.
  • the host adapter includes a Fibre Channel FC/Ethernet Fibre Channel FCoE HBA or an infinite bandwidth IB/RoCE HCA,
  • the interaction with the service sender, so that the service sender pauses the transmission of the service data specifically includes:
  • the sender of the service stops sending the credit response frame, so that the service sender pauses the transmission of the service data.
  • the method further includes:
  • the new virtual machine is bound to the emulated host adapter of the second physical host.
  • an embodiment of the present invention provides a first physical host, where the first physical host includes a host and a host adapter connected to the host, and the host adapter is a single input and output virtualized SR- A hardware device of the IOV standard, comprising one or more physical functions PF and one or more virtual functions VF; the host running a virtual machine and a virtual machine manager VMM, wherein the virtual machine runs a guest operating system, The guest operating system directly accesses the VF through a pass-through technology conforming to the SR-IOV standard;
  • the VMM further includes an emulated host adapter implemented in software; the host includes:
  • An interaction module configured to interact with a service sender, so that the service sender pauses the transmission of the service data
  • a copy module configured to copy data in the host adapter into the emulated host adapter
  • a binding module configured to release a binding between the virtual machine and the host adapter, and bind the virtual machine to the emulation host adapter;
  • a migration module configured to migrate data corresponding to the virtual machine and data corresponding to the emulated host adapter to the second physical host.
  • the data corresponding to the virtual machine includes data corresponding to the device simulated by the virtual machine and running data information; and the data corresponding to the simulated host adapter includes the Simulate the data of the host adapter itself and the data run by the emulation adapter.
  • the host adapter includes an Internet small computer system
  • the interface ISCSI HBA the interaction module is specifically configured to: continuously send, to the service sender, a transmission control protocol TCP packet information with a window domain of 0, so that the service sender pauses the transmission of the service data; wherein the window domain is 0.
  • the TCP packet information is used to instruct the service sender to implement the transmission suspension of the service data within a preset time period.
  • the host adapter includes a Fibre Channel FC/Ethernet Fibre Channel FCoE HBA or an infinite bandwidth IB/RoCE HCA
  • the interaction module is specifically configured to stop sending a credit response frame to the service sender, so that the service sender pauses the transmission of the service data.
  • the first physical host further includes:
  • An acquiring module configured to acquire data corresponding to the emulated host adapter of the second physical host and data corresponding to the virtual machine
  • the copying module is further configured to copy data corresponding to the emulated host adapter of the second physical host to a host adapter of the first physical host;
  • the binding unit is further configured to generate a corresponding new virtual machine according to the acquired data corresponding to the virtual machine, and bind the new virtual machine to an emulation host adapter of the second physical host.
  • a first physical host includes a memory, a processor, a host adapter inserted in the first physical host card slot, and an Ethernet network card, where:
  • the host adapter is configured to interact with a service sender, so that the service sender pauses the transmission of the service data.
  • the processor is configured to copy data in the host adapter into the emulation host adapter; unbind the virtual machine from the host adapter, and connect the virtual machine to the emulation host
  • the adapter performs binding; the data corresponding to the virtual machine and the data corresponding to the emulated host adapter are migrated to the second physical host.
  • the host adapter includes a host bus adapter HBA or a host channel adapter HCA; the data corresponding to the virtual machine includes data corresponding to the device simulated by the virtual machine and running data information; and the data corresponding to the simulated host adapter includes the emulation host adapter itself Data and data run by the emulation adapter.
  • the host adapter when the host adapter includes an Internet small computer system interface ISCSI HBA, the host adapter is specifically used to Transmitting, by the service sender, the transmission control protocol TCP packet information with the window field being 0, so that the service sender pauses the transmission of the service data; wherein the TCP packet information of the window domain is 0 is used to indicate the service transmission.
  • the party implements the suspension of the transmission of the service data within a preset time period.
  • the host adapter comprises a Fibre Channel FC/Ethernet Fibre Channel FCoE HBA or an infinite bandwidth IB/RoCE HCA Time
  • the host adapter is specifically configured to stop sending a credit response frame to the service sender, so that the service sender pauses the transmission of the service data.
  • the processor when the first physical host receives And the processor is further configured to: acquire the emulation host adapter of the second physical host, when the data corresponding to the emulated host adapter of the second physical host and the data corresponding to the virtual machine are sent by the second physical host Corresponding data and data corresponding to the virtual machine; copying data corresponding to the emulated host adapter of the second physical host to the host adapter of the first physical host, and generating corresponding new data according to the data corresponding to the virtual machine The virtual machine binds the new virtual machine to the emulated host adapter of the second physical host.
  • the method, device and system for virtual machine hot migration provided by the embodiment of the present invention, because the data in the host adapter is copied to the emulation host adapter, and the data corresponding to the virtual machine and the emulated host adapter are subsequently copied to the destination physical host. Therefore, there is no data loss in the process, so that the existing business can be uninterrupted.
  • FIG. 1 is a schematic diagram of an application architecture of a virtual machine according to an embodiment of the present invention.
  • FIG. 2 is a schematic flowchart diagram of a method for virtual machine hot migration according to an embodiment of the present invention
  • FIG. 3 is a schematic flowchart of another method for hot migration of a virtual machine according to an embodiment of the present invention.
  • FIG. 4 is a schematic flowchart diagram of still another method for virtual machine hot migration according to an embodiment of the present invention.
  • FIG. 5 is a schematic structural diagram of a first physical host according to an embodiment of the present invention.
  • FIG. 6 is a schematic structural diagram of another first physical host according to an embodiment of the present invention.
  • FIG. 7 is a schematic structural diagram of a virtual machine hot migration system according to an embodiment of the present invention.
  • the virtual machine application architecture applied in the embodiment of the present invention may be as shown in FIG. 1 .
  • the virtual machine application architecture that is, the VMM 11 having the function of managing virtual machines is installed on a physical host 1 in advance (in some other virtual machine application scenarios). Medium, with a similar function
  • the meta is also known as the hypervisor, and then the VMM manages one or more virtual machines (such as virtual machines 1-n in Figure 1), each of which can run the operating system OS and various applications.
  • the physical host 1 includes a part of hardware (for example, the Ethernet network card 12 and the host adapter 13) that is not directly used to run the virtual machine but performs some specific services and is relatively independent in function;
  • the hardware implementation can be based on the same hardware system (such as using the same PCB, the same processor to implement the network card 12 and the host adapter 13), or can also be based on different hardware systems (such as network cards) A single PCB and a separate processor are used, and the host adapter 13 uses another PCB separately and another separate processor).
  • the virtual machine running in the above architecture may lose the session connection context saved in the HBA/HCA during the hot migration.
  • the interruption of the upper layer service is that the session of the protocol stack (that is, the sum of the protocols in the network) is saved in the HBA/HCA, and the virtual machine guest operating system corresponding to the virtual machine cannot obtain the above session.
  • the HBA/HCA configured on the source physical host and the destination physical host must be identical.
  • the hardware parameters and slots of the HBA/HCA configured on the source physical host and the destination physical host must be the same.
  • the serial numbers of the VFs of the source physical host and the destination physical host must be the same, for example, when the VF 5 is on the source physical host.
  • VF 5 on the destination physical host must be in the idle state at the same time, so that VF 5 can be called when the virtual machine migrates from the source physical host to the destination physical host.
  • the virtual network application architecture includes a common network card, the ordinary network card only performs the task of sending and receiving network packets in the physical host, and the session context is maintained by the virtual machine guest operating system, and can be migrated with the virtual machine.
  • the HBA/HCA uninstalls the protocol stack, the session context cannot be migrated with the virtual machine, causing the interruption of the upper layer service. Therefore, the HBA/HCA is directly virtual In the mode of the machine, the virtual machine hot migration of the Small Computer System Interface (SCSI) service is not supported.
  • SCSI Small Computer System Interface
  • an embodiment of the present invention provides a new method, device, and system for virtual machine hot migration.
  • a first embodiment of the present invention provides a virtual machine hot migration method for migrating services in a virtual machine in a first physical host to a virtual machine of a second physical host.
  • an emulation host adapter is implemented in software in the VMM to simulate a real hardware host adapter.
  • the emulation host adapter may be pre-configured in the VMM or may be created by the VMM of the first physical host according to the hardware parameters of the host adapter in the first physical host, that is, the first physical host will be configured.
  • an emulated host adapter having the same descriptor information as the descriptor format and the control register stored in the host adapter according to the hardware parameters of the internal host adapter, the emulated host adapter using the same driver as the host adapter, and the After the emulation host adapter is loaded, it can provide the normal operating environment for the virtual machine just like the host adapter.
  • the specific creation process is a technology well known to those skilled in the art, and is not described in detail herein.
  • the emulation host adapter can be regarded as a software module added by the VMM to the prior art.
  • the embodiment of the present invention does not strictly describe each module except the emulation host adapter in the VMM.
  • VMM is used in place to replace other modules.
  • the operation of the emulation host adapter by the VMM can be understood as "the operation of the emulation host adapter by one or more other modules in the VMM", and those skilled in the art can understand that there will be a software module correspondingly. To perform an A operation on the emulated host adapter.
  • the method for providing a virtual machine hot migration includes the following steps:
  • the service sender here may be a second physical host that is connected to the first physical host through the network for service communication, for example, another server connected through Ethernet, or a terminal device.
  • the service data transmission is suspended mainly by the service sender, so that no subsequent service data is lost due to the migration of the virtual machine.
  • the first physical host interacts with the service sender
  • different interaction methods may be used according to different protocols.
  • the communication is performed based on the transmission control protocol TCP
  • the virtual machine of the first physical host may be used.
  • the VM responds to the peer by continuously sending a TCP acknowledgement message whose window field is 0 through the host adapter, so that the peer end suspends the transmission of the service data;
  • the VM in the first physical host may stop sending the credit response frame, and may also cause the peer to suspend data transmission;
  • the replication can be completed by the VMM in the first physical host.
  • the VMM can read the data in the host adapter through the interface provided by the host adapter, and then copy the read data to the emulated host adapter.
  • the data in the host adapter includes the receiving/transmitting queue of the host adapter and various hardware state information (such as session connection context information of the protocol stack).
  • the VMM completes the unbinding and rebinding.
  • binding is a term well known to those skilled in the art, and can be understood as associating one device with another device; otherwise, “unbinding” is to cancel the association between the two.
  • the original virtual machine is "bound" with the host adapter, and the service from the virtual machine is sent to other devices through the host adapter that is "bound” with it.
  • the virtual machine is unassociated from the host adapter and associated with the emulated host adapter, so that subsequent services from the virtual machine no longer communicate with the host adapter. And communicate with the emulated host adapter Letter so that no data is lost.
  • the first physical host After unbinding and rebinding, the first physical host removes the VF in the host adapter assigned to the virtual machine, and allocates the same VF in the emulated host adapter to the virtual machine.
  • the /output) channel is the emulation channel between the emulated host adapter and the virtual machine.
  • the data corresponding to the virtual machine and the data corresponding to the emulation host adapter are migrated to the second physical host.
  • the VMM completes the migration of data corresponding to the virtual machine and the self-defense host adapter.
  • the data corresponding to the virtual machine includes data corresponding to the virtual machine simulated device (such as simulated CPU, memory, hard disk, various peripherals, etc.) and running data information;
  • the data corresponding to the emulated host adapter includes the emulation host adapter data of the device itself and the data that the emulation adapter runs (such as data copied from the host adapter).
  • the migration of the two parts of data is the migration of software-level data.
  • the migration method is a technology well known to those skilled in the art.
  • the specific process may include: first migrating a part of data to a second physical host, and then comparing the first physical host. The difference between the data and the data in the second physical host, thereby updating the data in the second physical host; continuously iterating through the above method, and finally, when the difference is as small as a predetermined value, stopping the virtual machine, and migrating the data that needs to be updated last. Go to the second physical host.
  • the virtual machine needs to be stopped for a very short time, and the user is basically not aware of it, so that "hot migration" can be achieved.
  • the second physical host may restart the virtual machine and emulate the host adapter according to the migrated data, and notify the service sender to resend the service data to the second physical host, so that the second physical host may follow New business deal with.
  • the second physical host when the second physical host receives the data corresponding to the emulated host adapter sent by the first physical host, if the second physical host includes the host adapter (based on hardware), the second physical host receives The data corresponding to the emulated host adapter of the first physical host is completely copied to the host adapter of the first physical host, and the corresponding virtual machine is bound to perform the service; if the second physical host does not include the host adapter, Then, the VMM of the second physical host runs the emulated host adapter through software according to the received data, and performs with the virtual machine.
  • the first physical host can also serve as the destination host, and the second physical host serves as the source host.
  • the embodiment of the present invention further includes the following steps:
  • the virtual machine manager of the first physical host starts a replication function of the host adapter of the first physical host, thereby emulating the host adapter of the second physical host
  • the corresponding data is completely copied to the host adapter of the first physical host.
  • the VMM of the first physical host directly runs the emulated host adapter of the second physical host.
  • the host adapter or the emulated host adapter in the first physical host needs to perform a copy operation, that is, when the information is migrated, it needs to interact with the service sender. For example, when the emulation host adapter needs to send data in the host adapter stored in the emulated host adapter to the second physical host, the emulated host adapter running in the VMM of the first physical host needs to be performed by the VMM and the service sender. Interaction.
  • the virtual machine hot migration method provided by the embodiment of the present invention, since the data in the host adapter is copied to the emulation host adapter, and the virtual machine and the emulation are subsequently The data corresponding to the real host adapter is copied to the destination physical host. Therefore, no data is lost in the process, so that the virtual machine can be hot in the HBA/HCA pass-through virtual machine mode without interruption of the existing service. migrate.
  • the second embodiment of the present invention provides another method for hot migration of a virtual machine.
  • the steps are as follows. 101 is replaced with step 101a, or, when the host adapter is a Fibre Channel FC/Ethernet Fibre Channel FCoE HBA or an infinite bandwidth IB/RoCE HCA, step 101 is replaced with step 101b.
  • step 101 when step 101 is replaced with step 101a, the method includes the following steps:
  • the virtual machine of the first physical host can continuously send a TCP acknowledgement message with a window field of 0 through the host adapter to respond to the peer end by continuously sending a TCP acknowledgement message whose window field is 0, so that the peer end Suspend the transmission of business data.
  • the TCP message information of the window domain of 0 is communicated according to the transmission control protocol TCP, and is used to indicate that the service sender implements the transmission suspension of the service data within a preset time period.
  • the service sender after receiving the TCP packet information of the window domain 0, the service sender knows that the window domain in the first physical host is 0 (that is, there is no buffer in the first physical host), The TCP message should be sent and a pause of about 2 seconds is generated.
  • step 101 when step 101 is replaced with step 101b, the method includes the following steps:
  • the first physical host is based on the InfiniBand protocol, so that the virtual machine in the first physical host stops sending the credit response frame, so that the service sender does not receive the indication of the credit response frame, and the service sender pauses the data. send.
  • FIG. 3 Exemplarily, the embodiment corresponding to FIG. 3 is applied to a specific virtual machine. Use the architecture to specify.
  • the VMM in the source physical host allocates a hardware VMM to the virtual machine to allocate a VF of the hardware HBA to the virtual machine, and the virtual machine recognizes the hardware HBA VF.
  • the VF driver ie, the emulated HBA driver
  • SCSI Small Computer System Interface
  • the specific steps of performing hot migration from the source physical host to the destination physical host are as follows: First, the source physical host is based on the host in the first physical host through the VMM in the host.
  • the hardware parameters of the adapter are created, and an emulated HBA with the same hardware parameters as the hardware HBA of the source physical host is created on the source physical host.
  • the VMM virtual machine manager in the source physical host continuously sends the transmission control protocol TCP packet information of the window domain 0 to the service sender of the virtual machine through the hardware HBA, so that the service sender knows the first physical host. After the window field is 0 (that is, there is no buffer in the first physical host), the TCP message information is responded, thereby generating a transmission pause of about 2 seconds, so that the VMM will be the data in the hardware HBA on the source physical host. Completely copied to the emulated HBA.
  • the hardware HBA VF allocated to the virtual machine is removed, and the emulation HBA is allocated to the virtual machine so that the input/output of the virtual machine can be Take the simulation channel corresponding to the simulated HBA, so that the virtual machine can run normally on the simulated HBA.
  • the corresponding data of the emulated HBA stored on the source physical host emulates the host adapter data of the device itself and the data of the emulation adapter running (such as data copied from the host adapter) and the device simulated by the virtual machine (eg, The corresponding data and the running data information of the simulated CPU, memory, hard disk, various peripherals, etc.
  • the VMM in the destination physical host reassigns the migrated virtual machine to the same hardware HBA VF as the source physical host, and removes the simulated HBA so that the virtual machine's input/output goes to the hardware channel of the destination physical host.
  • the VMM in the destination physical host continuously sends the simulated HBA of the destination physical host to the service sender (for example, the source physical host).
  • the transmission control protocol TCP packet information of the window field is 0, so that the peer end stops transmitting data, and then the corresponding data of the emulation HBA is completely copied to the VF of the hardware HBA of the destination physical host.
  • the target physical host will run the simulated HBA directly through the VMM.
  • the data in the host adapter is copied to the emulation host adapter, and the data corresponding to the virtual machine and the emulation host adapter is subsequently copied to the destination physical host, In the process, there is no data loss, so that the hot migration of the virtual machine can be realized in the HBA/HCA through virtual machine mode without interruption of the existing service.
  • a third embodiment of the present invention provides a first physical host, where the first physical host includes a host and a host adapter connected to the host, and the host adapter is a hardware device conforming to a single input and output virtualized SR-IOV standard. Include one or more physical functions PF and one or more virtual functions VF; the host runs a virtual machine and a virtual machine manager VMM running a guest operating system, the guest operating system conforms to the SR-IOV standard The direct pass technology directly accesses the VF; at the same time, the above VMM also includes an emulated host adapter implemented in software. As shown in FIG. 5, the host 3 further includes: an interaction module 31, a replication module 32, a binding module 33, and a migration module 34, where:
  • the interaction module 31 is configured to interact with the service sender, so that the service sender pauses the transmission of the service data.
  • the copy module 32 is configured to copy data in the host adapter into the emulated host adapter.
  • the binding module 33 is configured to release the binding between the virtual machine and the host adapter, and bind the virtual machine to the emulated host adapter.
  • the migration module 34 is configured to migrate data corresponding to the virtual machine and data corresponding to the emulated host adapter to the second physical host.
  • the host adapter includes a host bus adapter HBA and a host channel adapter HCA; and the data corresponding to the virtual machine includes a virtual machine simulated device.
  • the corresponding data and the running data information; the data corresponding to the emulated host adapter includes the data of the emulated host adapter itself and the data of the emulation adapter running.
  • the interaction module 31 is specifically configured to: continuously send the transmission control protocol TCP packet information with the window domain 0 to the service sender, so that the service sender is suspended.
  • the sending of the service data, wherein the TCP message information of the window field is 0 is used to indicate that the service sender performs the transmission suspension of the service data within a preset time period.
  • the interaction module 31 is specifically configured to: stop transmitting a credit response frame to the service sender, so that the service sender Suspend the transmission of business data.
  • the first physical host receives the data corresponding to the simulated host adapter of the second physical host sent by the second physical host and the data corresponding to the virtual machine:
  • the first physical host further includes: an obtaining module 35, where:
  • the obtaining module 35 is configured to acquire data corresponding to the emulated host adapter of the second physical host and data corresponding to the virtual machine.
  • the copying module 32 is further configured to copy data corresponding to the emulated host adapter of the second physical host to the host adapter of the first physical host.
  • the binding unit 33 is further configured to generate a corresponding new virtual machine according to the data corresponding to the virtual machine, and bind the new virtual machine to the emulation host adapter of the second physical host.
  • the first physical host provided by the embodiment of the present invention copies the data in the host adapter to the emulated host adapter, and then copies the data corresponding to the virtual machine and the emulated host adapter to the destination physical host. Therefore, in the process, There is no data loss, so that the virtual machine can be hot-migrated in the HBA/HCA pass-through virtual machine mode without interrupting the existing services.
  • the fourth embodiment of the present invention shown in FIG. 6 provides a schematic structural diagram of another first physical host.
  • the first physical host 4 can be a server or a terminal device.
  • the first A physical host includes a memory 41, a processor 42, a host adapter 43 inserted in the physical host card slot, and an Ethernet network card 44.
  • the various components described above are coupled together by a bus system.
  • the bus may be an Industry Standard Architecture (ISA) bus, a Peripheral Component (PCI) bus, or an Extended Industry Standard Architecture (EISA) bus.
  • ISA Industry Standard Architecture
  • PCI Peripheral Component
  • EISA Extended Industry Standard Architecture
  • the bus can be divided into an address bus, a data bus, a control bus, and the like. For ease of representation, only one thick line is shown in Figure 6, but it does not mean that there is only one bus or one type of bus. among them:
  • the above-mentioned memory 41 may be any available medium that can be accessed by a computer, including but not limited to: read only memory (ROM), random access memory (RAM), or disk storage.
  • ROM read only memory
  • RAM random access memory
  • the storage device is used to provide instruction and data information to the processor, and the processor is connected through a bus.
  • the foregoing memory 41 stores: an operating system, an application, a VMM for managing virtual machine functions, an operating system for running the virtual machine, and various applications in the virtual machine, and program code for implementing the embodiment of the present invention. . among them,
  • the host adapter 43 described above is a hardware device conforming to a single input/output virtualized SR-IOV standard, and includes one or more physical functions PF and one or more virtual function VFs.
  • the processor 42 performs processing and transfer of messages between the devices by reading the information stored in the memory 41.
  • the host adapter 43 is configured to interact with the service sender, so that the service sender pauses the transmission of the service data.
  • a processor 42 for copying data in the host adapter into the emulation host adapter; and for unbinding the virtual machine from the host adapter, and binding the virtual machine to the emulated host adapter; The data corresponding to the virtual machine and the data corresponding to the emulated host adapter are migrated to the second physical host.
  • the above host adapter 43 includes a host bus adapter HBA and a host The channel adapter HCA; and the data corresponding to the virtual machine includes the data corresponding to the device simulated by the virtual machine and the running data information; the data corresponding to the simulated host adapter includes the data of the simulated host adapter itself and the data of the simulation adapter running.
  • the host adapter 43 when the host adapter 43 includes the Internet small computer system interface ISCSI HBA, the host adapter 43 is specifically configured to: continuously send the transmission control protocol TCP packet information of the window domain to 0 to the service sender, so that the service sender The transmission of the service data is suspended; wherein the TCP message information of the window field is 0, which is used to indicate that the service sender performs the transmission suspension of the service data within a preset time period.
  • the host adapter 43 when the host adapter 43 includes a Fibre Channel FC/Ethernet Fibre Channel FCoE HBA or an Infinite Bandwidth IB/RoCE HCA, the host adapter 43 is specifically configured to: stop transmitting a credit response frame to the service sender, so that the service is sent. The party suspends the transmission of business data.
  • the processor 42 is further configured to acquire the second Corresponding data of the simulated host adapter of the physical host and data corresponding to the virtual machine; and data for copying the emulated host adapter of the second physical host to the host adapter of the first physical host, and according to the The data corresponding to the virtual machine generates a corresponding new virtual machine, and the new virtual machine is bound to the emulated host adapter of the second physical host.
  • the first physical host provided by the embodiment of the present invention copies the data in the host adapter to the emulated host adapter, and then copies the data corresponding to the virtual machine and the emulated host adapter to the destination physical host. Therefore, in the process, There is no data loss, so that the virtual machine can be hot-migrated in the HBA/HCA pass-through virtual machine mode without interrupting the existing services.
  • a fifth embodiment of the present invention provides a virtual machine hot migration system.
  • the virtual machine hot migration system 5 includes a first physical host 51 and a second physical host 52, wherein the first physical host 51 is 2, 3, 4, 5, and 6 correspond to the first physical host shown in the embodiment.
  • the first physical host 51 is configured to interact with the service sender, so that the service sender pauses the transmission of the service data; copies the data in the host adapter to the emulation host adapter; and removes the binding between the virtual machine and the host adapter, and The virtual machine is bound to the emulation host adapter; the data corresponding to the virtual machine and the data corresponding to the emulated host adapter are migrated to the second physical host 52.
  • the host adapter includes a host bus adapter HBA and a host channel adapter HCA; and the data corresponding to the virtual machine includes data corresponding to the device simulated by the virtual machine and running data information; and the data corresponding to the simulated host adapter includes the simulation host.
  • the first physical host 51 is specifically configured to: continuously send, to the service sender, a transmission control protocol TCP packet information with a window domain of 0, so that the service is sent.
  • the party temporarily suspends the transmission of the service data; wherein the TCP message information of the window field is 0, which is used to indicate that the service sender implements the suspension of the service data within the preset time period.
  • the first physical host 51 is specifically configured to: stop transmitting a credit response frame to the service sender, so that the service is The sender suspends the transmission of business data.
  • the second physical host 52 is configured to obtain data corresponding to the emulated host adapter of the first physical host, and copy the data corresponding to the emulated host adapter of the first physical host to the host adapter of the second physical host.
  • the virtual machine hot migration system provided by the embodiment of the present invention copies the data in the host adapter to the emulated host adapter, and then copies the data corresponding to the virtual machine and the emulated host adapter to the destination physical host. Therefore, in this process, There is no data loss in the middle, so that the hot migration of the virtual machine can be realized in the HBA/HCA through virtual machine mode without interruption of the existing service.
  • the disclosed system, apparatus, and method may be implemented in other manners.
  • the device embodiments described above are merely illustrative.
  • the division of the modules or units is only a logical function division.
  • there may be another division manner for example, multiple units or components may be used. Combinations can be integrated into another system, or some features can be ignored or not executed.
  • the mutual coupling or direct coupling or communication connection shown or discussed may be an indirect coupling or communication connection through some interface, device or unit, and may be in an electrical, mechanical or other form.
  • each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit.
  • the above integrated unit can be implemented in the form of hardware or in the form of a software functional unit.
  • All or part of the steps of implementing the above method embodiments may be performed by hardware related to the program instructions.
  • the foregoing program may be stored in a computer readable storage medium, and when executed, the steps including the foregoing method embodiments are performed;
  • the foregoing storage medium includes: a U disk, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk, or an optical disk, and the like, which can store program codes.

Landscapes

  • Engineering & Computer Science (AREA)
  • Software Systems (AREA)
  • Theoretical Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • General Engineering & Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Computer And Data Communications (AREA)

Abstract

Embodiments of the present invention provide a method, a physical host and a system for live migration of a virtual machine, relating to the field of communications, for solving the problem existing in the prior art that live migration of a virtual machine can not be implemented in the mode of HBA/HCA directly accessing virtual machine. The method particularly comprises: interacting with a transaction sender to make the transaction sender pause sending transactional data; copying the data in a host adapter to an emulated host adapter; removing the binding between the virtual machine and the host adapter and binding the virtual machine to the emulated host adapter; and migrating the data corresponding to the virtual machine and the data corresponding to the emulated host adapter to a second physical host. The present invention is used for live migration of a virtual machine.

Description

一种虚拟机热迁移的方法、物理主机及系统Method, physical host and system for virtual machine hot migration

本申请要求于2014年7月31日提交中国专利局、申请号为201410373167.2、发明名称为“一种虚拟机热迁移的方法、物理主机及系统”的中国专利申请的优先权,其全部内容通过引用结合在本申请中。This application claims the priority of the Chinese patent application filed on July 31, 2014, the Chinese Patent Office, the application number is 201410373167.2, and the invention name is "a virtual machine thermal migration method, physical host and system". The citations are incorporated herein by reference.

技术领域Technical field

本发明涉及通信领域,尤其涉及一种虚拟机热迁移的方法、物理主机及系统。The present invention relates to the field of communications, and in particular, to a method, a physical host, and a system for virtual machine hot migration.

背景技术Background technique

计算机虚拟化技术是当前的热点,而虚拟机的热迁移技术则是虚拟化技术当中的热点,可以说,虚拟机的热迁移技术已经成为衡量计算机虚拟化技术优劣的一项检验指标。所谓虚拟机的热迁移是指在保证虚拟机上服务正常运行的同时,将正在运行的虚拟机从源物理主机迁移至目的物理主机中。Computer virtualization technology is the current hot spot, and the virtual migration technology of virtual machine is a hot spot in virtualization technology. It can be said that the hot migration technology of virtual machine has become a test indicator to measure the advantages and disadvantages of computer virtualization technology. The so-called virtual machine hot migration refers to the migration of the running virtual machine from the source physical host to the destination physical host while ensuring the normal operation of the service on the virtual machine.

现在常用的一种典型的虚拟机应用架构,即一台安装具有管理虚拟机功能的虚拟机管理器(Virtual Machine Manager,简称VMM)的物理主机,具体由VMM管理一个或多个虚拟机,每个虚拟机上可以运行操作系统OS以及各种应用。一般物理主机的硬件主要包括:一部分跟运行虚拟机强相关的硬件(例如CPU、内存、硬盘等),以及一部分不直接用于运行虚拟机,而用于执行一些特定业务,且功能相对独立的硬件(例如,以太网网卡和主机适配器)。而主机适配器通常包括基于单根输入输出虚拟化SR-IOV(Single Root I/O Virtualization HBA/HCA标准实现的一个物理功能PF以及一个或多个虚拟功能(Virtual Function,简称VF),其中,一个虚拟功能可以和与其对应的一个虚拟机进行直通操作。A typical virtual machine application architecture that is commonly used today is a physical host that has a virtual machine manager (VMM) that manages virtual machine functions. Specifically, one or more virtual machines are managed by the VMM. The operating system OS and various applications can be run on a virtual machine. The hardware of a general physical host mainly includes: a part of hardware (such as CPU, memory, hard disk, etc.) that is strongly related to running a virtual machine, and a part of which is not directly used to run a virtual machine, but is used to perform some specific services, and the functions are relatively independent. Hardware (for example, Ethernet network card and host adapter). The host adapter usually includes a single physical input SR-IOV (Single Root I/O Virtualization HBA/HCA standard implementation of a physical function PF and one or more virtual functions (VF), one of which The virtual function can be directly connected to a virtual machine corresponding to it.

当上述的虚拟机应用架构中的主机适配器为主机总线适配器(Host Bus Adaptor,简称HBA)/主机通道适配器(Host Channel Adaptor,简称HCA)时,对于运行在上述架构中的虚拟机来说,由 于HBA/HCA直通虚拟机的模式下不支持小型计算机系统接口(Small Computer System Interface,简称SCSI)业务不中断的虚拟机热迁移,即由于协议栈的会话控制是由HBA/HCA硬件实现的,且协议栈的会话无法随虚拟机进行迁移,同时SCSI对于报文丢失的容忍度极低,因此,连续丢失HBA/HCA上保存的协议栈的会话上下文,会导致上层业务的中断,造成虚拟机热迁移不成功,因此,在HBA/HCA直通虚拟机的模式下是无法实现虚拟机热迁移的。When the host adapter in the virtual machine application architecture is a Host Bus Adaptor (HBA)/Host Channel Adaptor (HCA), for the virtual machine running in the above architecture, In the HBA/HCA pass-through virtual machine mode, the virtual machine hot migration of the Small Computer System Interface (SCSI) service is not supported, that is, since the session control of the protocol stack is implemented by the HBA/HCA hardware, The session of the protocol stack cannot be migrated with the virtual machine. At the same time, SCSI has a very low tolerance for packet loss. Therefore, the continuous loss of the session context of the protocol stack saved on the HBA/HCA causes the interruption of the upper layer service and the virtual machine. The hot migration is unsuccessful, so virtual machine hot migration is not possible in the HBA/HCA pass-through virtual machine mode.

针对上述问题,现有技术通过多路径技术,即通过虚拟机管理程序为源物理主机的HBA/HCA上运行的虚拟机分配一个能够供虚拟机正常运行的仿真以太网网卡设备(Ethernet Network Interface Card,简称Eth NIC),使得虚拟机在进行热迁移时,源物理主机可以断开HBA/HCA与虚拟机间的绑定,并将该虚拟机与创建的仿真Eth NIC进行绑定,然后,通过虚拟机与该仿真Eth NIC间的仿真路径,将该虚拟机中的所存储的业务数据迁移到目的物理主机。In response to the above problem, the prior art allocates an emulated Ethernet network card device (Ethernet Network Interface Card) capable of running the virtual machine to the virtual machine running on the HBA/HCA of the source physical host through the multi-path technology, that is, through the hypervisor. (Eth NIC for short), when the virtual machine is hot migration, the source physical host can disconnect the HBA/HCA from the virtual machine, bind the virtual machine to the created emulation Eth NIC, and then pass The simulation path between the virtual machine and the simulated Eth NIC migrates the stored service data in the virtual machine to the destination physical host.

但是,发明人发现在上述的虚拟机迁移的过程中,由于源物理主机依旧无法迁移HBA/HCA上保存的协议栈的会话连接上下文,从而在虚拟机进行迁移时,丢失了存储在HBA/HCA中的会话上下文,进而导致虚拟机无法顺利完成热迁移。However, the inventor found that in the above-mentioned virtual machine migration process, since the source physical host is still unable to migrate the session connection context of the protocol stack saved on the HBA/HCA, the storage is lost in the HBA/HCA when the virtual machine is migrated. The session context in the network, which in turn causes the virtual machine to fail to complete the hot migration.

发明内容Summary of the invention

本发明的实施例提供一种虚拟机热迁移的方法、物理主机及系统,用于解决现有技术存在着的无法在HBA/HCA直通虚拟机的模式下实现虚拟机的热迁移的问题。The embodiments of the present invention provide a method, a physical host, and a system for virtual machine hot migration, which are used to solve the problem that the hot migration of the virtual machine cannot be implemented in the HBA/HCA through virtual machine mode in the prior art.

为达到上述目的,本发明的实施例采用如下技术方案:In order to achieve the above object, embodiments of the present invention adopt the following technical solutions:

第一方面,本发明实施例提供了一种虚拟机热迁移方法,用于第一物理主机,所述第一物理主机包括主机以及与所述主机相连的主机适配器,所述主机适配器为一个符合单根输入输出虚拟化SR-IOV标准的硬件设备,包括一个或多个物理功能PF以及一个或多个虚拟功能VF;所述主机运行有虚拟机以及虚拟机管理器VMM,所述虚拟机中运行有客户操作系统,所述客户操作系统通过符合 SR-IOV标准的直通技术直接访问所述VF;In a first aspect, an embodiment of the present invention provides a virtual machine hot migration method, where the first physical host includes a host and a host adapter connected to the host, where the host adapter is a match. A single input and output virtualized SR-IOV standard hardware device, including one or more physical functions PF and one or more virtual functions VF; the host runs a virtual machine and a virtual machine manager VMM, in the virtual machine Running with a guest operating system, the guest operating system passes the match The pass-through technology of the SR-IOV standard directly accesses the VF;

所述VMM中还包括用软件实现的仿真主机适配器,所述方法包括:The VMM also includes an emulated host adapter implemented in software, the method comprising:

与业务发送方进行交互,使得业务发送方暂停业务数据的发送;Interacting with the service sender, causing the service sender to suspend the transmission of the service data;

将所述主机适配器中的数据复制到所述仿真主机适配器中;Copying data from the host adapter into the emulated host adapter;

解除所述虚拟机与所述主机适配器间的绑定,并将所述虚拟机与所述仿真主机适配器进行绑定;Unbinding the virtual machine and the host adapter, and binding the virtual machine to the emulation host adapter;

将所述虚拟机对应的数据以及仿真主机适配器对应的数据迁移至所述第二物理主机。The data corresponding to the virtual machine and the data corresponding to the emulated host adapter are migrated to the second physical host.

在第一方面的第一种可能的实现方式中,所述虚拟机对应的数据包括所述虚拟机仿真出的设备对应的数据以及运行的数据信息;所述仿真主机适配器对应的数据包括所述仿真主机适配器本身数据以及所述仿真适配器运行的数据。In a first possible implementation manner of the first aspect, the data corresponding to the virtual machine includes data corresponding to the device simulated by the virtual machine and running data information; the data corresponding to the simulated host adapter includes the Simulate the data of the host adapter itself and the data run by the emulation adapter.

结合第一方面,或者第一方面第一种可能的实现方式,在第二种可能的实现方式中,所述主机适配器包括因特网小型计算机系统接口ISCSI HBA,所述与业务发送方进行交互,使得业务发送方暂停业务数据的发送包括:With reference to the first aspect, or the first possible implementation manner of the first aspect, in a second possible implementation, the host adapter includes an Internet Small Computer System Interface (ISCSI) HBA, and the interaction with the service sender enables The service sender suspends the transmission of service data including:

向所述业务发送方持续发送window域为0的传输控制协议TCP报文信息,使得业务发送方暂停业务数据的发送;其中所述window域为0的TCP报文信息用于指示所述业务发送方实现所述业务数据在预设时间段内的发送暂停。Transmitting, by the service sender, the transmission control protocol TCP packet information with the window field being 0, so that the service sender pauses the transmission of the service data; wherein the TCP packet information of the window domain is 0 is used to indicate the service transmission. The party implements the suspension of the transmission of the service data within a preset time period.

结合第一方面,或者第一方面第一种可能的实现方式,在第三种可能的实现方式中,所述主机适配器包括光纤通道FC/以太网光纤通道FCoE HBA或无限带宽IB/RoCE HCA,所述与业务发送方进行交互,使得业务发送方暂停业务数据的发送具体包括:With reference to the first aspect, or the first possible implementation manner of the first aspect, in a third possible implementation, the host adapter includes a Fibre Channel FC/Ethernet Fibre Channel FCoE HBA or an infinite bandwidth IB/RoCE HCA, The interaction with the service sender, so that the service sender pauses the transmission of the service data, specifically includes:

向所述业务发送方停止发送credit应答帧,使得业务发送方暂停业务数据的发送。The sender of the service stops sending the credit response frame, so that the service sender pauses the transmission of the service data.

结合第一方面,或者第一方面第一至第三种任意一种可能的实现方式,在第四种可能的实现方式中,所述方法还包括: With reference to the first aspect, or any one of the first to the third possible implementation manners of the first aspect, in a fourth possible implementation, the method further includes:

获取所述第二物理主机的仿真主机适配器对应的数据以及虚拟机对应的数据;Obtaining data corresponding to the emulated host adapter of the second physical host and data corresponding to the virtual machine;

将所述第二物理主机的仿真主机适配器对应的数据复制到所述第一物理主机的主机适配器中,并根据获取到的所述虚拟机对应的数据生成对应的新的虚拟机,将所述新的虚拟机与所述第二物理主机的仿真主机适配器进行绑定。Copying data corresponding to the emulated host adapter of the second physical host to the host adapter of the first physical host, and generating a corresponding new virtual machine according to the acquired data corresponding to the virtual machine, The new virtual machine is bound to the emulated host adapter of the second physical host.

第二方面,本发明实施例提供了一种第一物理主机,所述第一物理主机包括主机以及与所述主机相连的主机适配器,所述主机适配器为一个符合单根输入输出虚拟化SR-IOV标准的硬件设备,包括一个或多个物理功能PF以及一个或多个虚拟功能VF;所述主机运行有虚拟机以及虚拟机管理器VMM,所述虚拟机中运行有客户操作系统,所述客户操作系统通过符合SR-IOV标准的直通技术直接访问所述VF;In a second aspect, an embodiment of the present invention provides a first physical host, where the first physical host includes a host and a host adapter connected to the host, and the host adapter is a single input and output virtualized SR- A hardware device of the IOV standard, comprising one or more physical functions PF and one or more virtual functions VF; the host running a virtual machine and a virtual machine manager VMM, wherein the virtual machine runs a guest operating system, The guest operating system directly accesses the VF through a pass-through technology conforming to the SR-IOV standard;

所述VMM中还包括用软件实现的仿真主机适配器;所述主机包括:The VMM further includes an emulated host adapter implemented in software; the host includes:

交互模块,用于与业务发送方进行交互,使得业务发送方暂停业务数据的发送;An interaction module, configured to interact with a service sender, so that the service sender pauses the transmission of the service data;

复制模块,用于将所述主机适配器中的数据复制到所述仿真主机适配器中;a copy module, configured to copy data in the host adapter into the emulated host adapter;

绑定模块,用于解除所述虚拟机与所述主机适配器间的绑定,并将所述虚拟机与所述仿真主机适配器进行绑定;a binding module, configured to release a binding between the virtual machine and the host adapter, and bind the virtual machine to the emulation host adapter;

迁移模块,用于将所述虚拟机对应的数据以及仿真主机适配器对应的数据迁移至所述第二物理主机。And a migration module, configured to migrate data corresponding to the virtual machine and data corresponding to the emulated host adapter to the second physical host.

在第二方面的第一种可能的实现方式中,所述虚拟机对应的数据包括所述虚拟机仿真出的设备对应的数据以及运行的数据信息;所述仿真主机适配器对应的数据包括所述仿真主机适配器本身数据以及所述仿真适配器运行的数据。In a first possible implementation manner of the second aspect, the data corresponding to the virtual machine includes data corresponding to the device simulated by the virtual machine and running data information; and the data corresponding to the simulated host adapter includes the Simulate the data of the host adapter itself and the data run by the emulation adapter.

结合第二方面,或者第二方面第一种可能的实现方式,在第二种可能的实现方式中,所述主机适配器包括因特网小型计算机系统 接口ISCSI HBA,所述交互模块具体用于:向所述业务发送方持续发送window域为0的传输控制协议TCP报文信息,使得业务发送方暂停业务数据的发送;其中所述window域为0的TCP报文信息用于指示所述业务发送方实现所述业务数据在预设时间段内的发送暂停。With reference to the second aspect, or the first possible implementation manner of the second aspect, in a second possible implementation manner, the host adapter includes an Internet small computer system The interface ISCSI HBA, the interaction module is specifically configured to: continuously send, to the service sender, a transmission control protocol TCP packet information with a window domain of 0, so that the service sender pauses the transmission of the service data; wherein the window domain is 0. The TCP packet information is used to instruct the service sender to implement the transmission suspension of the service data within a preset time period.

结合第二方面,或者第二方面第一种可能的实现方式,在第三种可能的实现方式中,所述主机适配器包括光纤通道FC/以太网光纤通道FCoE HBA或无限带宽IB/RoCE HCA,所述交互模块具体用于:向所述业务发送方停止发送credit应答帧,使得业务发送方暂停业务数据的发送。With reference to the second aspect, or the first possible implementation manner of the second aspect, in a third possible implementation, the host adapter includes a Fibre Channel FC/Ethernet Fibre Channel FCoE HBA or an infinite bandwidth IB/RoCE HCA, The interaction module is specifically configured to stop sending a credit response frame to the service sender, so that the service sender pauses the transmission of the service data.

结合第二方面,或者第二方面第一至第三种任意一种可能的实现方式,在第四种可能的实现方式中,所述第一物理主机还包括:With reference to the second aspect, or any one of the first to the third possible implementation manners of the second aspect, in the fourth possible implementation, the first physical host further includes:

获取模块,用于获取所述第二物理主机的仿真主机适配器对应的数据以及虚拟机对应的数据;An acquiring module, configured to acquire data corresponding to the emulated host adapter of the second physical host and data corresponding to the virtual machine;

所述复制模块,还用于将所述第二物理主机的仿真主机适配器对应的数据复制到所述第一物理主机的主机适配器中;The copying module is further configured to copy data corresponding to the emulated host adapter of the second physical host to a host adapter of the first physical host;

所述绑定单元,还用于根据获取到的所述虚拟机对应的数据生成对应的新的虚拟机,将所述新的虚拟机与所述第二物理主机的仿真主机适配器进行绑定。The binding unit is further configured to generate a corresponding new virtual machine according to the acquired data corresponding to the virtual machine, and bind the new virtual machine to an emulation host adapter of the second physical host.

第三方面,提供一种第一物理主机,所述第一物理主机包括存储器、处理器、插在所述第一物理主机卡槽中的主机适配器及以太网网卡,其中:In a third aspect, a first physical host is provided, where the first physical host includes a memory, a processor, a host adapter inserted in the first physical host card slot, and an Ethernet network card, where:

所述主机适配器,用于与业务发送方进行交互,使得业务发送方暂停业务数据的发送。The host adapter is configured to interact with a service sender, so that the service sender pauses the transmission of the service data.

所述处理器,用于将所述主机适配器中的数据复制到所述仿真主机适配器中;解除所述虚拟机与所述主机适配器间的绑定,并将所述虚拟机与所述仿真主机适配器进行绑定;将所述虚拟机对应的数据以及仿真主机适配器对应的数据迁移至所述第二物理主机。The processor is configured to copy data in the host adapter into the emulation host adapter; unbind the virtual machine from the host adapter, and connect the virtual machine to the emulation host The adapter performs binding; the data corresponding to the virtual machine and the data corresponding to the emulated host adapter are migrated to the second physical host.

在第三方面的第一种可能的实现方式中,所述主机适配器包括 主机总线适配器HBA或主机通道适配器HCA;所述虚拟机对应的数据包括所述虚拟机仿真出的设备对应的数据以及运行的数据信息;所述仿真主机适配器对应的数据包括所述仿真主机适配器本身数据以及所述仿真适配器运行的数据。In a first possible implementation manner of the third aspect, the host adapter includes a host bus adapter HBA or a host channel adapter HCA; the data corresponding to the virtual machine includes data corresponding to the device simulated by the virtual machine and running data information; and the data corresponding to the simulated host adapter includes the emulation host adapter itself Data and data run by the emulation adapter.

根据第三方面的第一种可能的实现方式,在第三方面的第二种可能的实现方式中,当所述主机适配器包括因特网小型计算机系统接口ISCSI HBA时,所述主机适配器,具体用于向所述业务发送方持续发送window域为0的传输控制协议TCP报文信息,使得业务发送方暂停业务数据的发送;其中所述window域为0的TCP报文信息用于指示所述业务发送方实现所述业务数据在预设时间段内的发送暂停。According to a first possible implementation manner of the third aspect, in a second possible implementation manner of the third aspect, when the host adapter includes an Internet small computer system interface ISCSI HBA, the host adapter is specifically used to Transmitting, by the service sender, the transmission control protocol TCP packet information with the window field being 0, so that the service sender pauses the transmission of the service data; wherein the TCP packet information of the window domain is 0 is used to indicate the service transmission. The party implements the suspension of the transmission of the service data within a preset time period.

根据第一方面的第一种可能的实现方式,在第一方面的第三种可能的实现方式中,当所述主机适配器包括光纤通道FC/以太网光纤通道FCoE HBA或无限带宽IB/RoCE HCA时,According to a first possible implementation of the first aspect, in a third possible implementation of the first aspect, when the host adapter comprises a Fibre Channel FC/Ethernet Fibre Channel FCoE HBA or an infinite bandwidth IB/RoCE HCA Time,

所述主机适配器,具体用于向所述业务发送方停止发送credit应答帧,使得业务发送方暂停业务数据的发送。The host adapter is specifically configured to stop sending a credit response frame to the service sender, so that the service sender pauses the transmission of the service data.

结合第一方面或第一方面的第一种至第三种可能的实现方式中的任一种实现方式,在第一方面的第四种可能的实现方式中,当所述第一物理主机接收到所述第二物理主机发送的所述第二物理主机的仿真主机适配器对应的数据以及虚拟机对应的数据时;所述处理器,还用于:获取所述第二物理主机的仿真主机适配器对应的数据以及虚拟机对应的数据;将所述第二物理主机的仿真主机适配器对应的数据复制到所述第一物理主机的主机适配器中,并根据所述虚拟机对应的数据生成对应的新的虚拟机,将所述新的虚拟机与所述第二物理主机的仿真主机适配器进行绑定。With reference to the first aspect, or any one of the first to third possible implementation manners of the first aspect, in a fourth possible implementation manner of the first aspect, when the first physical host receives And the processor is further configured to: acquire the emulation host adapter of the second physical host, when the data corresponding to the emulated host adapter of the second physical host and the data corresponding to the virtual machine are sent by the second physical host Corresponding data and data corresponding to the virtual machine; copying data corresponding to the emulated host adapter of the second physical host to the host adapter of the first physical host, and generating corresponding new data according to the data corresponding to the virtual machine The virtual machine binds the new virtual machine to the emulated host adapter of the second physical host.

本发明的实施例提供的虚拟机热迁移的方法、装置及系统,由于主机适配器中的数据都复制到了仿真主机适配器,并后续又将虚拟机以及仿真主机适配器对应的数据复制到了目的物理主机,因此,在这过程中并没有数据丢失,从而可以在现有业务不中断的前提下, 在HBA/HCA直通虚拟机的模式下实现虚拟机的热迁移。The method, device and system for virtual machine hot migration provided by the embodiment of the present invention, because the data in the host adapter is copied to the emulation host adapter, and the data corresponding to the virtual machine and the emulated host adapter are subsequently copied to the destination physical host. Therefore, there is no data loss in the process, so that the existing business can be uninterrupted. Implement hot migration of virtual machines in the HBA/HCA pass-through virtual machine mode.

附图说明DRAWINGS

为了更清楚地说明本发明实施例或现有技术中的技术方案,下面将对实施例或现有技术描述中所需要使用的附图作简单地介绍,显而易见地,下面描述中的附图仅仅是本发明的一些实施例,对于本领域普通技术人员来讲,在不付出创造性劳动的前提下,还可以根据这些附图获得其他的附图。In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the embodiments or the description of the prior art will be briefly described below. Obviously, the drawings in the following description are only It is a certain embodiment of the present invention, and other drawings can be obtained from those skilled in the art without any creative work.

图1为本发明的实施例提供的一种虚拟机应用架构图;1 is a schematic diagram of an application architecture of a virtual machine according to an embodiment of the present invention;

图2为本发明的实施例提供的一种虚拟机热迁移的方法的流程示意图;FIG. 2 is a schematic flowchart diagram of a method for virtual machine hot migration according to an embodiment of the present invention;

图3为本发明的实施例提供的另一种虚拟机热迁移的方法的流程示意图;3 is a schematic flowchart of another method for hot migration of a virtual machine according to an embodiment of the present invention;

图4为本发明的实施例提供的又一种虚拟机热迁移的方法的流程示意图;FIG. 4 is a schematic flowchart diagram of still another method for virtual machine hot migration according to an embodiment of the present invention; FIG.

图5为本发明的实施例提供的一种第一物理主机的结构示意图;FIG. 5 is a schematic structural diagram of a first physical host according to an embodiment of the present invention;

图6为本发明的实施例提供的另一种第一物理主机的结构示意图;FIG. 6 is a schematic structural diagram of another first physical host according to an embodiment of the present invention;

图7为本发明的实施例提供的一种虚拟机热迁移系统的结构示意图。FIG. 7 is a schematic structural diagram of a virtual machine hot migration system according to an embodiment of the present invention.

具体实施方式detailed description

下面将结合本发明实施例中的附图,对本发明实施例中的技术方案进行清楚、完整地描述,显然,所描述的实施例仅仅是本发明一部分实施例,而不是全部的实施例。基于本发明中的实施例,本领域普通技术人员在没有做出创造性劳动前提下所获得的所有其他实施例,都属于本发明保护的范围。The technical solutions in the embodiments of the present invention are clearly and completely described in the following with reference to the accompanying drawings in the embodiments of the present invention. It is obvious that the described embodiments are only a part of the embodiments of the present invention, but not all embodiments. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative efforts are within the scope of the present invention.

本发明的实施例所应用的虚拟机应用架构可以如图1所示,该虚拟机应用架构,即预先在一台物理主机1上安装具有管理虚拟机功能的VMM11(在其他一些虚拟机应用场景中,具有类似功能的单 元也被称为Hypervisor),然后由VMM管理一个或多个虚拟机(如图1中的虚拟机1-n),每个虚拟机上可以运行操作系统OS以及各种应用。在本实施例中,物理主机1中包含有一部分不直接用于运行虚拟机,而用于执行一些特定业务,且功能相对独立的硬件(例如,以太网网卡12以及主机适配器13);其中,对于这些功能相对独立的硬件,在硬件实现上可以基于同一个硬件系统(如使用同一块PCB,同一个处理器来实现网卡12和主机适配器13),或者也可以基于不同的硬件系统(如网卡单独使用一块PCB以及单独的处理器,主机适配器13单独使用另一块PCB以及另一块单独的处理器)。The virtual machine application architecture applied in the embodiment of the present invention may be as shown in FIG. 1 . The virtual machine application architecture, that is, the VMM 11 having the function of managing virtual machines is installed on a physical host 1 in advance (in some other virtual machine application scenarios). Medium, with a similar function The meta is also known as the hypervisor, and then the VMM manages one or more virtual machines (such as virtual machines 1-n in Figure 1), each of which can run the operating system OS and various applications. In this embodiment, the physical host 1 includes a part of hardware (for example, the Ethernet network card 12 and the host adapter 13) that is not directly used to run the virtual machine but performs some specific services and is relatively independent in function; For these relatively independent hardware, the hardware implementation can be based on the same hardware system (such as using the same PCB, the same processor to implement the network card 12 and the host adapter 13), or can also be based on different hardware systems (such as network cards) A single PCB and a separate processor are used, and the host adapter 13 uses another PCB separately and another separate processor).

上述虚拟机的应用架构以及虚拟机基于SR-IOV对主机适配器中的虚拟功能进行直通操作的具体实现方法都为现有技术,本发明实施例不对此进行具体介绍。The application architecture of the above-mentioned virtual machine and the specific implementation method of the virtual machine performing the through-operation of the virtual function in the host adapter based on the SR-IOV are all the prior art, which is not specifically described in the embodiment of the present invention.

当上述的虚拟机应用架构中的主机适配器为HBA/HCA时,对于运行在上述架构中的虚拟机来说,虚拟机在进行热迁移时,会丢失HBA/HCA中保存的会话连接上下文,进而导致上层业务的中断,即由于协议栈(即网络中各层协议的总和)的会话是保存在HBA/HCA中,而虚拟机对应的虚拟机客户操作系统是无法获取上述的会话的。同时该虚拟机架构中的源物理主机和目的物理主机中的所配置的HBA/HCA需要完全一致,即源物理主机和目的物理主机中配置的HBA/HCA的硬件参数及插槽需要完全一致,例如当物理主机中的HBA/HCA采用单根输入输出虚拟化HBA/HCA SR-IOV模式时,源物理主机和目的物理主机的VF的序号必须相同,例如:当VF 5在源物理主机上被虚拟机使用时,目的物理主机上的VF 5必须同时处于空闲状态,以便虚拟机从源物理主机迁移至目的物理主机时,可以调用VF 5。再者,当该虚拟机应用架构中的包含普通网卡时,普通网卡在物理主机中仅承担网络报文收发的任务,而会话上下文由虚拟机客户操作系统进行维护,可以随虚拟机进行迁移,但由于HBA/HCA卸载了协议栈,进而造成会话上下文不可随虚拟机进行迁移,导致上层业务的中断。因此,HBA/HCA上直通虚拟 机的模式下是不支持小型计算机系统接口(Small Computer System Interface,简称SCSI)业务不中断的虚拟机热迁移的。When the host adapter in the virtual machine application architecture is an HBA/HCA, the virtual machine running in the above architecture may lose the session connection context saved in the HBA/HCA during the hot migration. The interruption of the upper layer service is that the session of the protocol stack (that is, the sum of the protocols in the network) is saved in the HBA/HCA, and the virtual machine guest operating system corresponding to the virtual machine cannot obtain the above session. The HBA/HCA configured on the source physical host and the destination physical host must be identical. The hardware parameters and slots of the HBA/HCA configured on the source physical host and the destination physical host must be the same. For example, when the HBA/HCA in the physical host adopts a single input and output virtualized HBA/HCA SR-IOV mode, the serial numbers of the VFs of the source physical host and the destination physical host must be the same, for example, when the VF 5 is on the source physical host. When the virtual machine is used, VF 5 on the destination physical host must be in the idle state at the same time, so that VF 5 can be called when the virtual machine migrates from the source physical host to the destination physical host. In addition, when the virtual network application architecture includes a common network card, the ordinary network card only performs the task of sending and receiving network packets in the physical host, and the session context is maintained by the virtual machine guest operating system, and can be migrated with the virtual machine. However, because the HBA/HCA uninstalls the protocol stack, the session context cannot be migrated with the virtual machine, causing the interruption of the upper layer service. Therefore, the HBA/HCA is directly virtual In the mode of the machine, the virtual machine hot migration of the Small Computer System Interface (SCSI) service is not supported.

为解决上述问题,本发明的实施例提供了一种新的虚拟机热迁移的方法、装置及系统。To solve the above problem, an embodiment of the present invention provides a new method, device, and system for virtual machine hot migration.

实施例一Embodiment 1

参见图2,本发明的实施例一提供了一种虚拟机热迁移的方法,用于将第一物理主机中的虚拟机中业务迁移到第二物理主机的虚拟机。本发明实施例中,在VMM中用软件实现一个仿真主机适配器,以模拟真正的硬件主机适配器。仿真主机适配器既可以预先配置在VMM中,也可以是要执行虚拟机热迁移时,由第一物理主机的VMM根据该第一物理主机中的主机适配器的硬件参数创建,即第一物理主机会根据该其内部的主机适配器的硬件参数创建出一个与该主机适配器中存储的描述符格式及控制寄存器的状态信息相同的仿真主机适配器,该仿真主机适配器与该主机适配器使用相同的驱动,且该仿真主机适配器在加载完驱动之后,可以与该主机适配器一样为虚拟机提供正常的运行环境。其中,具体的创建过程为本领域技术人员公知的技术,这里并不进行详细描述。Referring to FIG. 2, a first embodiment of the present invention provides a virtual machine hot migration method for migrating services in a virtual machine in a first physical host to a virtual machine of a second physical host. In the embodiment of the present invention, an emulation host adapter is implemented in software in the VMM to simulate a real hardware host adapter. The emulation host adapter may be pre-configured in the VMM or may be created by the VMM of the first physical host according to the hardware parameters of the host adapter in the first physical host, that is, the first physical host will be configured. Generating an emulated host adapter having the same descriptor information as the descriptor format and the control register stored in the host adapter according to the hardware parameters of the internal host adapter, the emulated host adapter using the same driver as the host adapter, and the After the emulation host adapter is loaded, it can provide the normal operating environment for the virtual machine just like the host adapter. The specific creation process is a technology well known to those skilled in the art, and is not described in detail herein.

本发明实施例中,仿真主机适配器可看成VMM相对于现有技术新增的一个软件模块,为了说明方便,本发明实施例并不对VMM中除仿真主机适配器外的各个其他模块进行严格说明,如无特殊说明,当涉及到其他模块与仿真主机适配器之间的交互时,统一使用VMM来代替其他模块。例如,“VMM对仿真主机适配器进行操作”可以理解为“VMM中的某一个或多个其他模块通过配合来对仿真主机适配器进行A操作”,本领域技术人员可以理解相应地会有一个软件模块来对仿真主机适配器进行A操作。In the embodiment of the present invention, the emulation host adapter can be regarded as a software module added by the VMM to the prior art. For convenience of description, the embodiment of the present invention does not strictly describe each module except the emulation host adapter in the VMM. Unless otherwise specified, when it comes to the interaction between other modules and the emulated host adapter, VMM is used in place to replace other modules. For example, "the operation of the emulation host adapter by the VMM" can be understood as "the operation of the emulation host adapter by one or more other modules in the VMM", and those skilled in the art can understand that there will be a software module correspondingly. To perform an A operation on the emulated host adapter.

基于上述应用场景,本发明实施例提供虚拟机热迁移方法包括如下步骤:Based on the foregoing application scenario, the method for providing a virtual machine hot migration according to the embodiment of the present invention includes the following steps:

101、与业务发送方进行交互,使得业务发送方暂停业务数据的发送。 101. Interact with the service sender, so that the service sender pauses the transmission of the service data.

这里的业务发送方可以是通过网络与第一物理主机连接以进行业务通信的第二物理主机,例如,通过以太网连接的另一台服务器,或者终端设备。这里主要通过业务发送方暂停业务数据发送,使得后续不会有业务数据因为虚拟机的迁移而丢失。The service sender here may be a second physical host that is connected to the first physical host through the network for service communication, for example, another server connected through Ethernet, or a terminal device. Here, the service data transmission is suspended mainly by the service sender, so that no subsequent service data is lost due to the migration of the virtual machine.

具体的,第一物理主机在与业务发送方进行交互时,可以根据不同的协议来使用不同的交互方法实现,例如,基于传输控制协议TCP进行通信时,可以由第一物理主机的虚拟机(VM)通过主机适配器持续发送window域为0的TCP确认报文来响应对端,从而使对端暂停业务数据的发送;Specifically, when the first physical host interacts with the service sender, different interaction methods may be used according to different protocols. For example, when the communication is performed based on the transmission control protocol TCP, the virtual machine of the first physical host may be used. The VM responds to the peer by continuously sending a TCP acknowledgement message whose window field is 0 through the host adapter, so that the peer end suspends the transmission of the service data;

或者,基于InfiniBand协议时,第一物理主机中的VM可以停止发送credit应答帧,也可以使得对端暂停数据的发送;Alternatively, when the InfiniBand protocol is used, the VM in the first physical host may stop sending the credit response frame, and may also cause the peer to suspend data transmission;

或者,采用其他协议时,也可以基于其他协议采用类似的方法来使得对端暂停发送数据。Alternatively, when other protocols are used, similar methods may be used based on other protocols to cause the peer to suspend sending data.

102、将主机适配器中的数据复制到仿真主机适配器中。102. Copy the data in the host adapter to the emulated host adapter.

具体的,可以由第一物理主机中的VMM来完成复制,例如,VMM可以通过主机适配器提供的接口读取主机适配器中的数据,然后将读取的数据复制到仿真主机适配器。Specifically, the replication can be completed by the VMM in the first physical host. For example, the VMM can read the data in the host adapter through the interface provided by the host adapter, and then copy the read data to the emulated host adapter.

其中,上述的主机适配器中的数据包括该主机适配器的接收/发送队列及各种硬件状态信息(如协议栈的会话连接上下文信息)。The data in the host adapter includes the receiving/transmitting queue of the host adapter and various hardware state information (such as session connection context information of the protocol stack).

103、解除虚拟机与主机适配器间的绑定,并将该虚拟机与仿真主机适配器进行绑定。103. Unbind the virtual machine from the host adapter and bind the virtual machine to the emulated host adapter.

具体的,由VMM来完成解绑以及重新绑定。其中,“绑定”为本领域技术人员所公知的术语,可以理解成将一个设备与另一个设备进行关联;反之,“解除绑定”是取消两者的关联性。本发明实施例中,原先虚拟机与主机适配器是“绑定”在一起的,来自虚拟机的业务会通过与之“绑定”的主机适配器发送到其他设备。通过本步骤的解除绑定以及与仿真主机适配器绑定,虚拟机解除了与主机适配器的关联,而与仿真主机适配器进行了关联,这样,后续来自虚拟机的业务便不再与主机适配器进行通信而与仿真主机适配器进行通 信,从而不会造成数据的丢失。Specifically, the VMM completes the unbinding and rebinding. Wherein, "binding" is a term well known to those skilled in the art, and can be understood as associating one device with another device; otherwise, "unbinding" is to cancel the association between the two. In the embodiment of the present invention, the original virtual machine is "bound" with the host adapter, and the service from the virtual machine is sent to other devices through the host adapter that is "bound" with it. Through the unbinding of this step and binding with the emulated host adapter, the virtual machine is unassociated from the host adapter and associated with the emulated host adapter, so that subsequent services from the virtual machine no longer communicate with the host adapter. And communicate with the emulated host adapter Letter so that no data is lost.

在解绑与重新绑定后,第一物理主机移除分配给虚拟机的主机适配器中的VF,并分配该仿真主机适配器中相同的VF给虚拟机,此时,该虚拟机的IO(输入/输出)通道为该仿真主机适配器与该虚拟机间的仿真通道。After unbinding and rebinding, the first physical host removes the VF in the host adapter assigned to the virtual machine, and allocates the same VF in the emulated host adapter to the virtual machine. At this time, the virtual machine's IO (input) The /output) channel is the emulation channel between the emulated host adapter and the virtual machine.

需要说明的是,由于是由VMM来完成,因此,可以做到让虚拟机不感知这种变化,即相对于虚拟机来说,虚拟机中“看到”的主机适配器其实并没有变化,虚拟机对主机适配器的操作也不发生变化。It should be noted that, because it is completed by the VMM, it is possible to make the virtual machine not perceive the change, that is, the host adapter "seeing" in the virtual machine does not change with respect to the virtual machine, and the virtual The operation of the machine to the host adapter does not change.

104、将虚拟机对应的数据以及仿真主机适配器对应的数据迁移至第二物理主机。104. The data corresponding to the virtual machine and the data corresponding to the emulation host adapter are migrated to the second physical host.

具体的,由VMM来完成虚拟机和防身主机适配器对应的数据的迁移。其中,虚拟机对应的数据包括虚拟机仿真出来的设备(如仿真出来的CPU、内存、硬盘、各种外设等)对应的数据以及运行的数据信息;Specifically, the VMM completes the migration of data corresponding to the virtual machine and the self-defense host adapter. The data corresponding to the virtual machine includes data corresponding to the virtual machine simulated device (such as simulated CPU, memory, hard disk, various peripherals, etc.) and running data information;

同理,仿真主机适配器对应的数据包括仿真主机适配器这个设备本身的数据以及仿真适配器运行的数据(如从主机适配器中复制过来的数据)。Similarly, the data corresponding to the emulated host adapter includes the emulation host adapter data of the device itself and the data that the emulation adapter runs (such as data copied from the host adapter).

这两部分数据的迁移为软件层面数据的迁移,该迁移方法为本领域技术人员所公知的技术,具体过程可以包括:先将一部分数据迁移到第二物理主机,然后比较第一物理主机中的数据跟第二物理主机中的数据的差别,从而更新第二物理主机中的数据;通过上述方法不断迭代,最后当差异小到一个预定的值后,停止虚拟机,将最后需要更新的数据迁移到第二物理主机。在这迁移过程中,虚拟机需要停止的时间非常短,用户基本不会感知,从而可以实现“热迁移”。The migration of the two parts of data is the migration of software-level data. The migration method is a technology well known to those skilled in the art. The specific process may include: first migrating a part of data to a second physical host, and then comparing the first physical host. The difference between the data and the data in the second physical host, thereby updating the data in the second physical host; continuously iterating through the above method, and finally, when the difference is as small as a predetermined value, stopping the virtual machine, and migrating the data that needs to be updated last. Go to the second physical host. During this migration process, the virtual machine needs to be stopped for a very short time, and the user is basically not aware of it, so that "hot migration" can be achieved.

在完成迁移后,第二物理主机可以根据迁移过来的数据重新启动虚拟机以及仿真主机适配器,并告知该业务发送方重新向该第二物理主机发送业务数据,从而由第二物理主机来对后续新业务进行 处理。After the migration is completed, the second physical host may restart the virtual machine and emulate the host adapter according to the migrated data, and notify the service sender to resend the service data to the second physical host, so that the second physical host may follow New business deal with.

在本实施例中,第二物理主机在接收到第一物理主机发送的仿真主机适配器对应的数据时,若该第二物理主机中包含主机适配器(基于硬件),则该第二物理主机将收到的第一物理主机的仿真主机适配器对应的数据完整的复制到该第一物理主机的主机适配器中,并绑定相应的虚拟机来执行业务;若该第二物理主机中未包含主机适配器,则该第二物理主机的VMM根据接收到的数据通过软件的方式来运行仿真主机适配器,并与虚拟机进行。In this embodiment, when the second physical host receives the data corresponding to the emulated host adapter sent by the first physical host, if the second physical host includes the host adapter (based on hardware), the second physical host receives The data corresponding to the emulated host adapter of the first physical host is completely copied to the host adapter of the first physical host, and the corresponding virtual machine is bound to perform the service; if the second physical host does not include the host adapter, Then, the VMM of the second physical host runs the emulated host adapter through software according to the received data, and performs with the virtual machine.

同理,第一物理主机也可以作为目的主机,而第二物理主机作为源主机,在这种情况下,本发明实施例还包括如下步骤:Similarly, the first physical host can also serve as the destination host, and the second physical host serves as the source host. In this case, the embodiment of the present invention further includes the following steps:

105、获取第二物理主机的仿真主机适配器对应的数据以及虚拟机对应的数据,将第二物理主机的仿真主机适配器对应的数据复制到第一物理主机的主机适配器中,并根据收到的虚拟机对应的数据生成对应的新的虚拟机,将新生成的虚拟机与仿真主机适配器进行绑定。105. Acquire data corresponding to the emulation host adapter of the second physical host and data corresponding to the virtual machine, and copy data corresponding to the emulated host adapter of the second physical host to the host adapter of the first physical host, and according to the received virtual The corresponding data of the machine generates a corresponding new virtual machine, and the newly generated virtual machine is bound to the emulation host adapter.

示例性的,若该第一物理主机中包含主机适配器,则该第一物理主机的虚拟机管理器启动该第一物理主机的主机适配器的复制功能,从而将该第二物理主机的仿真主机适配器对应的数据完整的复制到该第一物理主机的主机适配器中。但是,若该第一物理主机中未包含主机适配器,则该第一物理主机的VMM直接运行该第二物理主机的仿真主机适配器。Exemplarily, if the first physical host includes a host adapter, the virtual machine manager of the first physical host starts a replication function of the host adapter of the first physical host, thereby emulating the host adapter of the second physical host The corresponding data is completely copied to the host adapter of the first physical host. However, if the host adapter is not included in the first physical host, the VMM of the first physical host directly runs the emulated host adapter of the second physical host.

需要说明的是,第一物理主机中的主机适配器或仿真主机适配器需要进行复制操作,即进行信息迁移时,都需要与业务发送方发进行交互。例如,当仿真主机适配器需要将该仿真主机适配器中存储的主机适配器中的数据发送至第二物理主机时,该第一物理主机的VMM中运行的仿真主机适配器需要通过该VMM与业务发送方进行交互。It should be noted that the host adapter or the emulated host adapter in the first physical host needs to perform a copy operation, that is, when the information is migrated, it needs to interact with the service sender. For example, when the emulation host adapter needs to send data in the host adapter stored in the emulated host adapter to the second physical host, the emulated host adapter running in the VMM of the first physical host needs to be performed by the VMM and the service sender. Interaction.

本发明的实施例提供的虚拟机热迁移的方法中,由于主机适配器中的数据都复制到了仿真主机适配器,并后续又将虚拟机以及仿 真主机适配器对应的数据复制到了目的物理主机,因此,在这过程中并没有数据丢失,从而可以在现有业务不中断的前提下,在HBA/HCA直通虚拟机的模式下实现虚拟机的热迁移。In the virtual machine hot migration method provided by the embodiment of the present invention, since the data in the host adapter is copied to the emulation host adapter, and the virtual machine and the emulation are subsequently The data corresponding to the real host adapter is copied to the destination physical host. Therefore, no data is lost in the process, so that the virtual machine can be hot in the HBA/HCA pass-through virtual machine mode without interruption of the existing service. migrate.

实施例二Embodiment 2

本发明的实施例二提供另一种虚拟机热迁移的方法,本实施例在实施例一提供的步骤101至步骤105的基础上,在主机适配器为因特网小型计算机系统接口ISCSI HBA时,将步骤101替换为步骤101a,或者,在主机适配器为光纤通道FC/以太网光纤通道FCoE HBA或无限带宽IB/RoCE HCA时,将步骤101替换为步骤101b。The second embodiment of the present invention provides another method for hot migration of a virtual machine. In this embodiment, based on steps 101 to 105 provided in the first embodiment, when the host adapter is an Internet small computer system interface ISCSI HBA, the steps are as follows. 101 is replaced with step 101a, or, when the host adapter is a Fibre Channel FC/Ethernet Fibre Channel FCoE HBA or an infinite bandwidth IB/RoCE HCA, step 101 is replaced with step 101b.

如图3所示,当步骤101替换为步骤101a,该方法包括如下步骤:As shown in FIG. 3, when step 101 is replaced with step 101a, the method includes the following steps:

101a、向业务发送方持续发送window域为0的传输控制协议TCP报文信息。101a. Continuously send a transmission control protocol TCP packet information with a window domain of 0 to the service sender.

具体的,可以由第一物理主机的虚拟机通过主机适配器持续发送window域为0的TCP确认报文来响应对端持续发送window域为0的TCP确认报文来响应对端,从而使对端暂停业务数据的发送。其中,上述的window域为0的TCP报文信息是基于传输控制协议TCP进行通信的,用于指示该业务发送方实现该业务数据在预设时间段内的发送暂停。示例性的,业务发送方在接收到该window域为0的TCP报文信息后,获知该第一物理主机中的window域为0(即该第一物理主机中已无缓冲区)后,响应该TCP报文信息,并产生大约2秒的发送停顿。Specifically, the virtual machine of the first physical host can continuously send a TCP acknowledgement message with a window field of 0 through the host adapter to respond to the peer end by continuously sending a TCP acknowledgement message whose window field is 0, so that the peer end Suspend the transmission of business data. The TCP message information of the window domain of 0 is communicated according to the transmission control protocol TCP, and is used to indicate that the service sender implements the transmission suspension of the service data within a preset time period. Exemplarily, after receiving the TCP packet information of the window domain 0, the service sender knows that the window domain in the first physical host is 0 (that is, there is no buffer in the first physical host), The TCP message should be sent and a pause of about 2 seconds is generated.

如图4所示,当步骤101替换为步骤101b,该方法包括如下步骤:As shown in FIG. 4, when step 101 is replaced with step 101b, the method includes the following steps:

101b、向业务发送方停止发送credit应答帧。101b. Stop sending a credit response frame to the service sender.

具体的,第一物理主机基于InfiniBand协议,使得该第一物理主机中的虚拟机停止发送credit应答帧,以使业务发送方接收不到credit应答帧的指示,进而使得该业务发送方暂停数据的发送。Specifically, the first physical host is based on the InfiniBand protocol, so that the virtual machine in the first physical host stops sending the credit response frame, so that the service sender does not receive the indication of the credit response frame, and the service sender pauses the data. send.

示例性的,这里将图3对应的实施例应用于具体的虚拟机的应 用架构中进行具体说明。Exemplarily, the embodiment corresponding to FIG. 3 is applied to a specific virtual machine. Use the architecture to specify.

具体的,参照图1所示的虚拟机应用架构示意图,源物理主机中的VMM为虚拟机分配硬件VMM为虚拟机分配硬件HBA的一个VF给虚拟机,而虚拟机在识别到硬件HBA VF后,加载VF驱动(即仿真HBA驱动),使得硬件HBA VF成为虚拟机SCSI小型计算机系统接口(Small Computer System Interface,简称SCSI)操作的输入/输出通道。具体的,虚拟机在上述的网络环境下,从源物理主机向目的物理主机进行热迁移的具体步骤如下所示:首先,源物理主机通过其主机中的VMM根据该第一物理主机中的主机适配器的硬件参数创建,在源物理主机上创建一个与源物理主机的硬件HBA具有相同硬件参数的仿真HBA。其次,该源物理主机中的VMM虚拟机管理器通过硬件HBA向虚拟机的业务发送方持续发送window域为0的传输控制协议TCP报文信息,使得该业务发送方获知该第一物理主机中的window域为0(即该第一物理主机中已无缓冲区)后,响应该TCP报文信息,从而产生大约2秒的发送停顿,以便VMM将源物理主机上的硬件HBA中的数据的完整复制到仿真HBA中。而该源物理主机中的VMM在将硬件HBA的中的数据完整复制到仿真HBA后,移除分配给虚拟机的硬件HBA VF,同时分配仿真HBA给虚拟机,以便虚拟机的输入/输出可以走仿真HBA对应的仿真通道,使得虚拟机能够在仿真HBA上正常运行。然后,将该源物理主机上的仿真HBA存储的对应的数据仿真主机适配器这个设备本身的数据以及仿真适配器运行的数据(如从主机适配器中复制过来的数据)及虚拟机仿真出来的设备(如仿真出来的CPU、内存、硬盘、各种外设等)对应的数据以及运行的数据信息复制到目的物理主机上。最后,目的物理主机中的VMM为迁移过来的虚拟机重新分配与源物理主机相同的硬件HBA VF,并移除仿真HBA,使得虚拟机的输入/输出走该目的物理主机的硬件通道。此外,当目的物理主机中有可用的硬件HBA VF时,目的物理主机中的VMM会通过该目的物理主机的仿真HBA向业务发送方(如,源物理主机)持续发送 window域为0的传输控制协议TCP报文信息,以使对端停止发送数据,然后将仿真HBA的对应的数据完整复制到该目的物理主机的硬件HBA的VF中,当然,若该目的物理主机中没有可用的硬件HBA VF时,该目的物理主机会直接通过VMM运行该仿真HBA。Specifically, referring to the virtual machine application architecture diagram shown in FIG. 1, the VMM in the source physical host allocates a hardware VMM to the virtual machine to allocate a VF of the hardware HBA to the virtual machine, and the virtual machine recognizes the hardware HBA VF. The VF driver (ie, the emulated HBA driver) is loaded, so that the hardware HBA VF becomes an input/output channel of the virtual machine SCSI Small Computer System Interface (SCSI) operation. Specifically, in the foregoing network environment, the specific steps of performing hot migration from the source physical host to the destination physical host are as follows: First, the source physical host is based on the host in the first physical host through the VMM in the host. The hardware parameters of the adapter are created, and an emulated HBA with the same hardware parameters as the hardware HBA of the source physical host is created on the source physical host. The VMM virtual machine manager in the source physical host continuously sends the transmission control protocol TCP packet information of the window domain 0 to the service sender of the virtual machine through the hardware HBA, so that the service sender knows the first physical host. After the window field is 0 (that is, there is no buffer in the first physical host), the TCP message information is responded, thereby generating a transmission pause of about 2 seconds, so that the VMM will be the data in the hardware HBA on the source physical host. Completely copied to the emulated HBA. After the VMM in the source physical host completely copies the data in the hardware HBA to the emulation HBA, the hardware HBA VF allocated to the virtual machine is removed, and the emulation HBA is allocated to the virtual machine so that the input/output of the virtual machine can be Take the simulation channel corresponding to the simulated HBA, so that the virtual machine can run normally on the simulated HBA. Then, the corresponding data of the emulated HBA stored on the source physical host emulates the host adapter data of the device itself and the data of the emulation adapter running (such as data copied from the host adapter) and the device simulated by the virtual machine (eg, The corresponding data and the running data information of the simulated CPU, memory, hard disk, various peripherals, etc. are copied to the destination physical host. Finally, the VMM in the destination physical host reassigns the migrated virtual machine to the same hardware HBA VF as the source physical host, and removes the simulated HBA so that the virtual machine's input/output goes to the hardware channel of the destination physical host. In addition, when there is a hardware HBA VF available in the destination physical host, the VMM in the destination physical host continuously sends the simulated HBA of the destination physical host to the service sender (for example, the source physical host). The transmission control protocol TCP packet information of the window field is 0, so that the peer end stops transmitting data, and then the corresponding data of the emulation HBA is completely copied to the VF of the hardware HBA of the destination physical host. Of course, if the destination physical host When there is no hardware HBA VF available, the target physical host will run the simulated HBA directly through the VMM.

在本发明的实施例提供的虚拟机热迁移的方法中,由于主机适配器中的数据都复制到了仿真主机适配器,并后续又将虚拟机以及仿真主机适配器对应的数据复制到了目的物理主机,因此,在这过程中并没有数据丢失,从而可以在现有业务不中断的前提下,在HBA/HCA直通虚拟机的模式下实现虚拟机的热迁移。In the virtual machine hot migration method provided by the embodiment of the present invention, since the data in the host adapter is copied to the emulation host adapter, and the data corresponding to the virtual machine and the emulation host adapter is subsequently copied to the destination physical host, In the process, there is no data loss, so that the hot migration of the virtual machine can be realized in the HBA/HCA through virtual machine mode without interruption of the existing service.

实施例三Embodiment 3

本发明的实施例三提供一种第一物理主机,该第一物理主机包括主机以及与该主机相连的主机适配器,该主机适配器为一个符合单根输入输出虚拟化SR-IOV标准的硬件设备,包括一个或多个物理功能PF以及一个或多个虚拟功能VF;该主机运行有虚拟机以及虚拟机管理器VMM,该虚拟机中运行有客户操作系统,该客户操作系统通过符合SR-IOV标准的直通技术直接访问VF;同时上述的VMM中还包括用软件实现的仿真主机适配器。如图5所示,该主机3还包括:交互模块31、复制模块32、绑定模块33及迁移模块34,其中:A third embodiment of the present invention provides a first physical host, where the first physical host includes a host and a host adapter connected to the host, and the host adapter is a hardware device conforming to a single input and output virtualized SR-IOV standard. Include one or more physical functions PF and one or more virtual functions VF; the host runs a virtual machine and a virtual machine manager VMM running a guest operating system, the guest operating system conforms to the SR-IOV standard The direct pass technology directly accesses the VF; at the same time, the above VMM also includes an emulated host adapter implemented in software. As shown in FIG. 5, the host 3 further includes: an interaction module 31, a replication module 32, a binding module 33, and a migration module 34, where:

交互模块31,用于与业务发送方进行交互,使得业务发送方暂停业务数据的发送。The interaction module 31 is configured to interact with the service sender, so that the service sender pauses the transmission of the service data.

复制模块32,用于将主机适配器中的数据复制到仿真主机适配器中。The copy module 32 is configured to copy data in the host adapter into the emulated host adapter.

绑定模块33,用于解除虚拟机与主机适配器间的绑定,并将该虚拟机与该仿真主机适配器进行绑定。The binding module 33 is configured to release the binding between the virtual machine and the host adapter, and bind the virtual machine to the emulated host adapter.

迁移模块34,用于将虚拟机对应的数据以及仿真主机适配器对应的数据迁移至第二物理主机。The migration module 34 is configured to migrate data corresponding to the virtual machine and data corresponding to the emulated host adapter to the second physical host.

其中,上述的主机适配器包括主机总线适配器HBA和主机通道适配器HCA;而上述的虚拟机对应的数据包括虚拟机仿真出的设备 对应的数据以及运行的数据信息;仿真主机适配器对应的数据包括仿真主机适配器本身数据以及仿真适配器运行的数据。The host adapter includes a host bus adapter HBA and a host channel adapter HCA; and the data corresponding to the virtual machine includes a virtual machine simulated device. The corresponding data and the running data information; the data corresponding to the emulated host adapter includes the data of the emulated host adapter itself and the data of the emulation adapter running.

可选的,当上述的主机适配器包括因特网小型计算机系统接口ISCSI HBA时,交互模块31具体用于:向业务发送方持续发送window域为0的传输控制协议TCP报文信息,使得业务发送方暂停业务数据的发送;其中该window域为0的TCP报文信息用于指示上述的业务发送方实现业务数据在预设时间段内的发送暂停。Optionally, when the host adapter includes the Internet small computer system interface ISCSI HBA, the interaction module 31 is specifically configured to: continuously send the transmission control protocol TCP packet information with the window domain 0 to the service sender, so that the service sender is suspended. The sending of the service data, wherein the TCP message information of the window field is 0 is used to indicate that the service sender performs the transmission suspension of the service data within a preset time period.

可选的,当上述的主机适配器包括光纤通道FC/以太网光纤通道FCoE HBA或无限带宽IB/RoCE HCA时,交互模块31具体用于:向业务发送方停止发送credit应答帧,使得业务发送方暂停业务数据的发送。Optionally, when the foregoing host adapter includes a Fibre Channel FC/Ethernet Fibre Channel FCoE HBA or an infinite bandwidth IB/RoCE HCA, the interaction module 31 is specifically configured to: stop transmitting a credit response frame to the service sender, so that the service sender Suspend the transmission of business data.

可选的,当上述的第一物理主机接收到第二物理主机发送的第二物理主机的仿真主机适配器对应的数据以及虚拟机对应的数据时:Optionally, when the first physical host receives the data corresponding to the simulated host adapter of the second physical host sent by the second physical host and the data corresponding to the virtual machine:

如图5所示,该第一物理主机还包括:获取模块35,其中:As shown in FIG. 5, the first physical host further includes: an obtaining module 35, where:

获取模块35,用于获取第二物理主机的仿真主机适配器对应的数据以及虚拟机对应的数据。The obtaining module 35 is configured to acquire data corresponding to the emulated host adapter of the second physical host and data corresponding to the virtual machine.

复制模块32,还用于将第二物理主机的仿真主机适配器对应的数据复制到第一物理主机的主机适配器中。The copying module 32 is further configured to copy data corresponding to the emulated host adapter of the second physical host to the host adapter of the first physical host.

绑定单元33,还用于根据虚拟机对应的数据生成对应的新的虚拟机,将新的虚拟机与第二物理主机的仿真主机适配器进行绑定。The binding unit 33 is further configured to generate a corresponding new virtual machine according to the data corresponding to the virtual machine, and bind the new virtual machine to the emulation host adapter of the second physical host.

本发明的实施例提供的第一物理主机,由于主机适配器中的数据都复制到了仿真主机适配器,并后续又将虚拟机以及仿真主机适配器对应的数据复制到了目的物理主机,因此,在这过程中并没有数据丢失,从而可以在现有业务不中断的前提下,在HBA/HCA直通虚拟机的模式下实现虚拟机的热迁移。The first physical host provided by the embodiment of the present invention copies the data in the host adapter to the emulated host adapter, and then copies the data corresponding to the virtual machine and the emulated host adapter to the destination physical host. Therefore, in the process, There is no data loss, so that the virtual machine can be hot-migrated in the HBA/HCA pass-through virtual machine mode without interrupting the existing services.

实施例四Embodiment 4

如图6所示的本发明的实施例四提供另一种第一物理主机的结构示意图。该第一物理主机4可以是服务器,或者终端设备。该第 一物理主机包括存储器41、处理器42、插在该物理主机卡槽中的主机适配器43及以太网网卡44。上述的各个组件是通过总线系统耦合在一起。该总线可以是工业标准体系结构(Industry Standard Architecture,简称ISA)总线、外部设备互连(Peripheral Component,简称PCI)总线或扩展工业标准体系结构(Extended Industry Standard Architecture,简称EISA)总线等。该总线可以分为地址总线、数据总线、控制总线等。为便于表示,图6中仅用一条粗线表示,但并不表示仅有一根总线或一种类型的总线。其中:The fourth embodiment of the present invention shown in FIG. 6 provides a schematic structural diagram of another first physical host. The first physical host 4 can be a server or a terminal device. The first A physical host includes a memory 41, a processor 42, a host adapter 43 inserted in the physical host card slot, and an Ethernet network card 44. The various components described above are coupled together by a bus system. The bus may be an Industry Standard Architecture (ISA) bus, a Peripheral Component (PCI) bus, or an Extended Industry Standard Architecture (EISA) bus. The bus can be divided into an address bus, a data bus, a control bus, and the like. For ease of representation, only one thick line is shown in Figure 6, but it does not mean that there is only one bus or one type of bus. among them:

上述的存储器41可以是计算机能够存取的任何可用介质,包括但不限于:只读存储器(read only memory,简称ROM)、随机存储器(random access memory,简称RAM)、或磁盘存储(disk storage)、闪存、可编程只读存储器或电可擦写可编程存储器、寄存器等本领域熟悉的存储介质。存储设备用于向处理器提供指令和数据信息,和处理器通过总线连接。The above-mentioned memory 41 may be any available medium that can be accessed by a computer, including but not limited to: read only memory (ROM), random access memory (RAM), or disk storage. A storage medium familiar to the art, such as flash memory, programmable read only memory or electrically erasable programmable memory, registers, and the like. The storage device is used to provide instruction and data information to the processor, and the processor is connected through a bus.

上述的存储器41中存储:操作系统,应用程序,用于管理虚拟机功能的VMM,用于运行虚拟机的操作系统以及虚拟机中的各种应用,及用于实现本发明实施例的程序代码。其中,The foregoing memory 41 stores: an operating system, an application, a VMM for managing virtual machine functions, an operating system for running the virtual machine, and various applications in the virtual machine, and program code for implementing the embodiment of the present invention. . among them,

上述的主机适配器43为一个符合单根输入输出虚拟化SR-IOV标准的硬件设备,包括一个或多个物理功能PF以及一个或多个虚拟功能VF。The host adapter 43 described above is a hardware device conforming to a single input/output virtualized SR-IOV standard, and includes one or more physical functions PF and one or more virtual function VFs.

处理器42通过读取存储器41中存储的信息来进行各设备之间的消息的处理和传递。The processor 42 performs processing and transfer of messages between the devices by reading the information stored in the memory 41.

具体的,主机适配器43,用于与业务发送方进行交互,使得业务发送方暂停业务数据的发送。Specifically, the host adapter 43 is configured to interact with the service sender, so that the service sender pauses the transmission of the service data.

处理器42,用于将主机适配器中的数据复制到仿真主机适配器中;及用于解除虚拟机与主机适配器间的绑定,并将该虚拟机与该仿真主机适配器进行绑定;及用于将虚拟机对应的数据以及仿真主机适配器对应的数据迁移至第二物理主机。a processor 42 for copying data in the host adapter into the emulation host adapter; and for unbinding the virtual machine from the host adapter, and binding the virtual machine to the emulated host adapter; The data corresponding to the virtual machine and the data corresponding to the emulated host adapter are migrated to the second physical host.

其中,上述的主机适配器43包括主机总线适配器HBA和主机 通道适配器HCA;而上述的虚拟机对应的数据包括虚拟机仿真出的设备对应的数据以及运行的数据信息;仿真主机适配器对应的数据包括仿真主机适配器本身数据以及仿真适配器运行的数据。Wherein, the above host adapter 43 includes a host bus adapter HBA and a host The channel adapter HCA; and the data corresponding to the virtual machine includes the data corresponding to the device simulated by the virtual machine and the running data information; the data corresponding to the simulated host adapter includes the data of the simulated host adapter itself and the data of the simulation adapter running.

可选的,当上述的主机适配器43包括因特网小型计算机系统接口ISCSI HBA时,主机适配器43具体用于:向业务发送方持续发送window域为0的传输控制协议TCP报文信息,使得业务发送方暂停业务数据的发送;其中该window域为0的TCP报文信息用于指示上述的业务发送方实现业务数据在预设时间段内的发送暂停。Optionally, when the host adapter 43 includes the Internet small computer system interface ISCSI HBA, the host adapter 43 is specifically configured to: continuously send the transmission control protocol TCP packet information of the window domain to 0 to the service sender, so that the service sender The transmission of the service data is suspended; wherein the TCP message information of the window field is 0, which is used to indicate that the service sender performs the transmission suspension of the service data within a preset time period.

可选的,当上述的主机适配器43包括光纤通道FC/以太网光纤通道FCoE HBA或无限带宽IB/RoCE HCA时,主机适配器43具体用于:向业务发送方停止发送credit应答帧,使得业务发送方暂停业务数据的发送。Optionally, when the host adapter 43 includes a Fibre Channel FC/Ethernet Fibre Channel FCoE HBA or an Infinite Bandwidth IB/RoCE HCA, the host adapter 43 is specifically configured to: stop transmitting a credit response frame to the service sender, so that the service is sent. The party suspends the transmission of business data.

可选的,当上述的第一物理主机接收到第二物理主机发送的第二物理主机的仿真主机适配器对应的数据以及虚拟机对应的数据时,处理器42,还用于获取所述第二物理主机的仿真主机适配器对应的数据以及虚拟机对应的数据;及用于将所述第二物理主机的仿真主机适配器对应的数据复制到所述第一物理主机的主机适配器中,并根据所述虚拟机对应的数据生成对应的新的虚拟机,将所述新的虚拟机与所述第二物理主机的仿真主机适配器进行绑定。Optionally, when the first physical host receives the data corresponding to the simulated host adapter of the second physical host sent by the second physical host and the data corresponding to the virtual machine, the processor 42 is further configured to acquire the second Corresponding data of the simulated host adapter of the physical host and data corresponding to the virtual machine; and data for copying the emulated host adapter of the second physical host to the host adapter of the first physical host, and according to the The data corresponding to the virtual machine generates a corresponding new virtual machine, and the new virtual machine is bound to the emulated host adapter of the second physical host.

本发明的实施例提供的第一物理主机,由于主机适配器中的数据都复制到了仿真主机适配器,并后续又将虚拟机以及仿真主机适配器对应的数据复制到了目的物理主机,因此,在这过程中并没有数据丢失,从而可以在现有业务不中断的前提下,在HBA/HCA直通虚拟机的模式下实现虚拟机的热迁移。The first physical host provided by the embodiment of the present invention copies the data in the host adapter to the emulated host adapter, and then copies the data corresponding to the virtual machine and the emulated host adapter to the destination physical host. Therefore, in the process, There is no data loss, so that the virtual machine can be hot-migrated in the HBA/HCA pass-through virtual machine mode without interrupting the existing services.

实施例五Embodiment 5

本发明的实施例五提供一种虚拟机热迁移系统,如图7所示,该虚拟机热迁移系统5包括第一物理主机51和第二物理主机52,其中,该第一物理主机51为图2、3、4、5、6对应实施例所示的第一物理主机。 A fifth embodiment of the present invention provides a virtual machine hot migration system. As shown in FIG. 7, the virtual machine hot migration system 5 includes a first physical host 51 and a second physical host 52, wherein the first physical host 51 is 2, 3, 4, 5, and 6 correspond to the first physical host shown in the embodiment.

第一物理主机51,用于与业务发送方进行交互,使得业务发送方暂停业务数据的发送;将主机适配器中的数据复制到仿真主机适配器中;解除虚拟机与主机适配器间的绑定,并将该虚拟机与该仿真主机适配器进行绑定;将虚拟机对应的数据以及仿真主机适配器对应的数据迁移至第二物理主机52。The first physical host 51 is configured to interact with the service sender, so that the service sender pauses the transmission of the service data; copies the data in the host adapter to the emulation host adapter; and removes the binding between the virtual machine and the host adapter, and The virtual machine is bound to the emulation host adapter; the data corresponding to the virtual machine and the data corresponding to the emulated host adapter are migrated to the second physical host 52.

其中,上述的主机适配器包括主机总线适配器HBA和主机通道适配器HCA;而上述的虚拟机对应的数据包括虚拟机仿真出的设备对应的数据以及运行的数据信息;仿真主机适配器对应的数据包括仿真主机适配器本身数据以及仿真适配器运行的数据。The host adapter includes a host bus adapter HBA and a host channel adapter HCA; and the data corresponding to the virtual machine includes data corresponding to the device simulated by the virtual machine and running data information; and the data corresponding to the simulated host adapter includes the simulation host. The data of the adapter itself and the data that the emulation adapter runs.

可选的,当上述的主机适配器包括因特网小型计算机系统接口ISCSI HBA时,第一物理主机51具体用于:向业务发送方持续发送window域为0的传输控制协议TCP报文信息,使得业务发送方暂停业务数据的发送;其中该window域为0的TCP报文信息用于指示上述的业务发送方实现业务数据在预设时间段内的发送暂停。Optionally, when the foregoing host adapter includes an Internet small computer system interface ISCSI HBA, the first physical host 51 is specifically configured to: continuously send, to the service sender, a transmission control protocol TCP packet information with a window domain of 0, so that the service is sent. The party temporarily suspends the transmission of the service data; wherein the TCP message information of the window field is 0, which is used to indicate that the service sender implements the suspension of the service data within the preset time period.

可选的,当上述的主机适配器包括光纤通道FC/以太网光纤通道FCoE HBA或无限带宽IB/RoCE HCA时,第一物理主机51具体用于:向业务发送方停止发送credit应答帧,使得业务发送方暂停业务数据的发送。Optionally, when the foregoing host adapter includes a Fibre Channel FC/Ethernet Fibre Channel FCoE HBA or an Infinite Bandwidth IB/RoCE HCA, the first physical host 51 is specifically configured to: stop transmitting a credit response frame to the service sender, so that the service is The sender suspends the transmission of business data.

第二物理主机52,用于获取第一物理主机的仿真主机适配器对应的数据,并将第一物理主机的仿真主机适配器对应的数据复制到第二物理主机的主机适配器中。The second physical host 52 is configured to obtain data corresponding to the emulated host adapter of the first physical host, and copy the data corresponding to the emulated host adapter of the first physical host to the host adapter of the second physical host.

本发明的实施例提供的虚拟机热迁移系统,由于主机适配器中的数据都复制到了仿真主机适配器,并后续又将虚拟机以及仿真主机适配器对应的数据复制到了目的物理主机,因此,在这过程中并没有数据丢失,从而可以在现有业务不中断的前提下,在HBA/HCA直通虚拟机的模式下实现虚拟机的热迁移。The virtual machine hot migration system provided by the embodiment of the present invention copies the data in the host adapter to the emulated host adapter, and then copies the data corresponding to the virtual machine and the emulated host adapter to the destination physical host. Therefore, in this process, There is no data loss in the middle, so that the hot migration of the virtual machine can be realized in the HBA/HCA through virtual machine mode without interruption of the existing service.

所属领域的技术人员可以清楚地了解到,为描述的方便和简洁,仅以上述各功能模块的划分进行举例说明,实际应用中,可以根据需要而将上述功能分配由不同的功能模块完成,即将装置的内部结 构划分成不同的功能模块,以完成以上描述的全部或者部分功能。上述描述的系统,装置和单元的具体工作过程,可以参考前述方法实施例中的对应过程,在此不再赘述。It will be clearly understood by those skilled in the art that for the convenience and brevity of the description, only the division of each functional module described above is exemplified. In practical applications, the above function assignment can be completed by different functional modules as needed. Internal junction of the device The structure is divided into different functional modules to complete all or part of the functions described above. For the specific working process of the system, the device and the unit described above, reference may be made to the corresponding process in the foregoing method embodiments, and details are not described herein again.

在本申请所提供的几个实施例中,应该理解到,所揭露的系统,装置和方法,可以通过其它的方式实现。例如,以上所描述的装置实施例仅仅是示意性的,例如,所述模块或单元的划分,仅仅为一种逻辑功能划分,实际实现时可以有另外的划分方式,例如多个单元或组件可以结合或者可以集成到另一个系统,或一些特征可以忽略,或不执行。另一点,所显示或讨论的相互之间的耦合或直接耦合或通信连接可以是通过一些接口,装置或单元的间接耦合或通信连接,可以是电性,机械或其它的形式。In the several embodiments provided by the present application, it should be understood that the disclosed system, apparatus, and method may be implemented in other manners. For example, the device embodiments described above are merely illustrative. For example, the division of the modules or units is only a logical function division. In actual implementation, there may be another division manner, for example, multiple units or components may be used. Combinations can be integrated into another system, or some features can be ignored or not executed. In addition, the mutual coupling or direct coupling or communication connection shown or discussed may be an indirect coupling or communication connection through some interface, device or unit, and may be in an electrical, mechanical or other form.

另外,在本申请各个实施例中的各功能单元可以集成在一个处理单元中,也可以是各个单元单独物理存在,也可以两个或两个以上单元集成在一个单元中。上述集成的单元既可以采用硬件的形式实现,也可以采用软件功能单元的形式实现。In addition, each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above integrated unit can be implemented in the form of hardware or in the form of a software functional unit.

实现上述方法实施例的全部或部分步骤可以通过程序指令相关的硬件来完成,前述的程序可以存储于计算机可读取存储介质中,该程序在执行时,执行包括上述方法实施例的步骤;而前述的存储介质包括:U盘、移动硬盘、只读存储器(ROM,Read-Only Memory)、随机存取存储器(RAM,Random Access Memory)、磁碟或者光盘等各种可以存储程序代码的介质。All or part of the steps of implementing the above method embodiments may be performed by hardware related to the program instructions. The foregoing program may be stored in a computer readable storage medium, and when executed, the steps including the foregoing method embodiments are performed; The foregoing storage medium includes: a U disk, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk, or an optical disk, and the like, which can store program codes.

以上所述,以上实施例仅用以说明本申请的技术方案,而非对其限制;尽管参照前述实施例对本申请进行了详细的说明,本领域的普通技术人员应当理解:其依然可以对前述各实施例所记载的技术方案进行修改,或者对其中部分技术特征进行等同替换;而这些修改或者替换,并不使相应技术方案的本质脱离本申请各实施例技术方案的精神和范围。 The above embodiments are only used to explain the technical solutions of the present application, and are not limited thereto; although the present application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that they can still The technical solutions described in the embodiments are modified, or the equivalents of the technical features are replaced by the equivalents. The modifications and substitutions of the embodiments do not depart from the spirit and scope of the technical solutions of the embodiments of the present application.

Claims (10)

一种虚拟机热迁移方法,用于第一物理主机,所述第一物理主机包括主机以及与所述主机相连的主机适配器,所述主机适配器为一个符合单根输入输出虚拟化SR-IOV标准的硬件设备,包括一个或多个物理功能PF以及一个或多个虚拟功能VF;所述主机运行有虚拟机以及虚拟机管理器VMM,所述虚拟机中运行有客户操作系统,所述客户操作系统通过符合SR-IOV标准的直通技术直接访问所述VF;A virtual machine hot migration method is used for a first physical host, the first physical host includes a host and a host adapter connected to the host, and the host adapter is a single input and output virtualized SR-IOV standard Hardware device, comprising one or more physical functions PF and one or more virtual functions VF; the host running a virtual machine and a virtual machine manager VMM, wherein the virtual machine runs a guest operating system, the customer operation The system directly accesses the VF through a pass-through technology conforming to the SR-IOV standard; 其特征在于,所述VMM中还包括用软件实现的仿真主机适配器,所述方法包括:The VMM further includes an emulation host adapter implemented by software, and the method includes: 与业务发送方进行交互,使得业务发送方暂停业务数据的发送;Interacting with the service sender, causing the service sender to suspend the transmission of the service data; 将所述主机适配器中的数据复制到所述仿真主机适配器中;Copying data from the host adapter into the emulated host adapter; 解除所述虚拟机与所述主机适配器间的绑定,并将所述虚拟机与所述仿真主机适配器进行绑定;Unbinding the virtual machine and the host adapter, and binding the virtual machine to the emulation host adapter; 将所述虚拟机对应的数据以及仿真主机适配器对应的数据迁移至所述第二物理主机。The data corresponding to the virtual machine and the data corresponding to the emulated host adapter are migrated to the second physical host. 根据权利要求1所述的方法,其特征在于,所述虚拟机对应的数据包括所述虚拟机仿真出的设备对应的数据以及运行的数据信息;所述仿真主机适配器对应的数据包括所述仿真主机适配器本身数据以及所述仿真适配器运行的数据。The method according to claim 1, wherein the data corresponding to the virtual machine comprises data corresponding to the device simulated by the virtual machine and data information of the operation; and the data corresponding to the simulated host adapter includes the simulation The data of the host adapter itself and the data run by the emulation adapter. 根据权利要求1或2所述的方法,其特征在于,所述主机适配器包括因特网小型计算机系统接口ISCSI HBA,所述与业务发送方进行交互,使得业务发送方暂停业务数据的发送包括:The method according to claim 1 or 2, wherein the host adapter comprises an Internet Small Computer System Interface (ISCSI) HBA, and the interaction with the service sender causes the service sender to suspend the transmission of the service data, including: 向所述业务发送方持续发送window域为0的传输控制协议TCP报文信息,使得业务发送方暂停业务数据的发送;其中所述window域为0的TCP报文信息用于指示所述业务发送方实现所述业务数据在预设时间段内的发送暂停。Transmitting, by the service sender, the transmission control protocol TCP packet information with the window field being 0, so that the service sender pauses the transmission of the service data; wherein the TCP packet information of the window domain is 0 is used to indicate the service transmission. The party implements the suspension of the transmission of the service data within a preset time period. 根据权利要求1或2所述的方法,其特征在于,所述主机适配器包括光纤通道FC/以太网光纤通道FCoE HBA或无限带宽IB/RoCE HCA,所述与业务发送方进行交互,使得业务发送方暂停业 务数据的发送具体包括:The method according to claim 1 or 2, wherein the host adapter comprises a Fibre Channel FC/Ethernet Fibre Channel FCoE HBA or an infinite bandwidth IB/RoCE HCA, the interaction with the service sender, so that the service is sent Suspended business The delivery of data includes: 向所述业务发送方停止发送credit应答帧,使得业务发送方暂停业务数据的发送。The sender of the service stops sending the credit response frame, so that the service sender pauses the transmission of the service data. 根据权利要求1至4任一项所述的方法,其特征在于,所述方法还包括:The method according to any one of claims 1 to 4, further comprising: 获取所述第二物理主机的仿真主机适配器对应的数据以及虚拟机对应的数据;Obtaining data corresponding to the emulated host adapter of the second physical host and data corresponding to the virtual machine; 将所述第二物理主机的仿真主机适配器对应的数据复制到所述第一物理主机的主机适配器中,并根据获取到的所述虚拟机对应的数据生成对应的新的虚拟机,将所述新的虚拟机与所述第二物理主机的仿真主机适配器进行绑定。Copying data corresponding to the emulated host adapter of the second physical host to the host adapter of the first physical host, and generating a corresponding new virtual machine according to the acquired data corresponding to the virtual machine, The new virtual machine is bound to the emulated host adapter of the second physical host. 一种第一物理主机,所述第一物理主机包括主机以及与所述主机相连的主机适配器,所述主机适配器为一个符合单根输入输出虚拟化SR-IOV标准的硬件设备,包括一个或多个物理功能PF以及一个或多个虚拟功能VF;所述主机运行有虚拟机以及虚拟机管理器VMM,所述虚拟机中运行有客户操作系统,所述客户操作系统通过符合SR-IOV标准的直通技术直接访问所述VF;A first physical host, the first physical host includes a host and a host adapter connected to the host, the host adapter being a hardware device conforming to a single input and output virtualized SR-IOV standard, including one or more Physical function PF and one or more virtual functions VF; the host runs a virtual machine and a virtual machine manager VMM, wherein the virtual machine runs a guest operating system, and the guest operating system passes the SR-IOV standard Direct access technology directly accesses the VF; 其特征在于,所述VMM中还包括用软件实现的仿真主机适配器;所述主机包括:The VMM further includes an emulated host adapter implemented by software; the host includes: 交互模块,用于与业务发送方进行交互,使得业务发送方暂停业务数据的发送;An interaction module, configured to interact with a service sender, so that the service sender pauses the transmission of the service data; 复制模块,用于将所述主机适配器中的数据复制到所述仿真主机适配器中;a copy module, configured to copy data in the host adapter into the emulated host adapter; 绑定模块,用于解除所述虚拟机与所述主机适配器间的绑定,并将所述虚拟机与所述仿真主机适配器进行绑定;a binding module, configured to release a binding between the virtual machine and the host adapter, and bind the virtual machine to the emulation host adapter; 迁移模块,用于将所述虚拟机对应的数据以及仿真主机适配器对应的数据迁移至所述第二物理主机。And a migration module, configured to migrate data corresponding to the virtual machine and data corresponding to the emulated host adapter to the second physical host. 根据权利要求6所述的第一物理主机,其特征在于,所述虚拟机对应的数据包括所述虚拟机仿真出的设备对应的数据以及运行 的数据信息;所述仿真主机适配器对应的数据包括所述仿真主机适配器本身数据以及所述仿真适配器运行的数据。The first physical host according to claim 6, wherein the data corresponding to the virtual machine includes data corresponding to the device simulated by the virtual machine and running The data corresponding to the emulated host adapter includes data of the emulated host adapter itself and data of the emulation adapter running. 根据权利要6或7所述的第一物理主机,其特征在于,所述主机适配器包括因特网小型计算机系统接口ISCSI HBA,所述交互模块具体用于:向所述业务发送方持续发送window域为0的传输控制协议TCP报文信息,使得业务发送方暂停业务数据的发送;其中所述window域为0的TCP报文信息用于指示所述业务发送方实现所述业务数据在预设时间段内的发送暂停。The first physical host according to claim 6 or 7, wherein the host adapter comprises an Internet Small Computer System Interface (ISCSI) HBA, and the interaction module is specifically configured to: continuously send a window domain to the service sender The transmission control protocol TCP message information of 0 causes the service sender to suspend the transmission of the service data; wherein the TCP message information in which the window field is 0 is used to instruct the service sender to implement the service data in a preset time period. The transmission within is paused. 根据权利要求6或7所述的第一物理主机,其特征在于,所述主机适配器包括光纤通道FC/以太网光纤通道FCoE HBA或无限带宽IB/RoCE HCA,所述交互模块具体用于:向所述业务发送方停止发送credit应答帧,使得业务发送方暂停业务数据的发送。The first physical host according to claim 6 or 7, wherein the host adapter comprises a Fibre Channel FC/Ethernet Fibre Channel FCoE HBA or an infinite bandwidth IB/RoCE HCA, and the interaction module is specifically configured to: The service sender stops sending the credit response frame, so that the service sender pauses the transmission of the service data. 根据权利要求6至9任一项所述的第一物理主机,其特征在于,所述第一物理主机还包括:The first physical host according to any one of claims 6 to 9, wherein the first physical host further comprises: 获取模块,用于获取所述第二物理主机的仿真主机适配器对应的数据以及虚拟机对应的数据;An acquiring module, configured to acquire data corresponding to the emulated host adapter of the second physical host and data corresponding to the virtual machine; 所述复制模块,还用于将所述第二物理主机的仿真主机适配器对应的数据复制到所述第一物理主机的主机适配器中;The copying module is further configured to copy data corresponding to the emulated host adapter of the second physical host to a host adapter of the first physical host; 所述绑定单元,还用于根据获取到的所述虚拟机对应的数据生成对应的新的虚拟机,将所述新的虚拟机与所述第二物理主机的仿真主机适配器进行绑定。 The binding unit is further configured to generate a corresponding new virtual machine according to the acquired data corresponding to the virtual machine, and bind the new virtual machine to an emulation host adapter of the second physical host.
PCT/CN2014/095013 2014-07-31 2014-12-25 Method, physical host and system for live migration of virtual machine Ceased WO2016015443A1 (en)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201410373167.2 2014-07-31
CN201410373167.2A CN104199716B (en) 2014-07-31 2014-07-31 A kind of method of live migration of virtual machine, physical host and system

Publications (1)

Publication Number Publication Date
WO2016015443A1 true WO2016015443A1 (en) 2016-02-04

Family

ID=52085013

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2014/095013 Ceased WO2016015443A1 (en) 2014-07-31 2014-12-25 Method, physical host and system for live migration of virtual machine

Country Status (2)

Country Link
CN (1) CN104199716B (en)
WO (1) WO2016015443A1 (en)

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN109189557A (en) * 2018-09-03 2019-01-11 郑州云海信息技术有限公司 A virtual machine scheduling method and system for high network communication
CN114679461A (en) * 2022-03-30 2022-06-28 浙江大学 Cloud network session connectivity assurance method and system for virtual machine live migration scenarios
CN115202803A (en) * 2021-04-13 2022-10-18 超聚变数字技术有限公司 Fault processing method and device

Families Citing this family (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104199716B (en) * 2014-07-31 2018-01-16 华为技术有限公司 A kind of method of live migration of virtual machine, physical host and system
CN104753816A (en) * 2015-03-27 2015-07-01 华为技术有限公司 RDMA (remote direct memory access) connection message processing method and related device
CN109426547B (en) * 2017-09-04 2022-05-17 华为技术有限公司 Thermal migration method, device and system of virtual machine
CN111736943B (en) * 2019-03-25 2024-10-29 阿里巴巴集团控股有限公司 Migration method and system of virtual machine
CN113467791B (en) * 2021-06-25 2024-09-03 广东浪潮智慧计算技术有限公司 Deployment adjustment method, system, device and storage medium for applications in cloud platform
CN119336440A (en) * 2023-07-21 2025-01-21 华为技术有限公司 A virtual machine migration method, chip system and host
CN118295773B (en) * 2024-06-04 2024-08-23 浪潮云信息技术股份公司 Method, equipment and medium for thermally migrating information-based operating system across different CPU virtual machines

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20080189432A1 (en) * 2007-02-02 2008-08-07 International Business Machines Corporation Method and system for vm migration in an infiniband network
CN103201721A (en) * 2012-08-29 2013-07-10 华为技术有限公司 Virtual machine thermal migration system and method
US20130254321A1 (en) * 2012-03-26 2013-09-26 Oracle International Corporation System and method for supporting live migration of virtual machines in a virtualization environment
CN104199716A (en) * 2014-07-31 2014-12-10 华为技术有限公司 Virtual machine thermal migration method, physical host and system

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20080189432A1 (en) * 2007-02-02 2008-08-07 International Business Machines Corporation Method and system for vm migration in an infiniband network
US20130254321A1 (en) * 2012-03-26 2013-09-26 Oracle International Corporation System and method for supporting live migration of virtual machines in a virtualization environment
CN103201721A (en) * 2012-08-29 2013-07-10 华为技术有限公司 Virtual machine thermal migration system and method
CN104199716A (en) * 2014-07-31 2014-12-10 华为技术有限公司 Virtual machine thermal migration method, physical host and system

Cited By (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN109189557A (en) * 2018-09-03 2019-01-11 郑州云海信息技术有限公司 A virtual machine scheduling method and system for high network communication
CN115202803A (en) * 2021-04-13 2022-10-18 超聚变数字技术有限公司 Fault processing method and device
CN114679461A (en) * 2022-03-30 2022-06-28 浙江大学 Cloud network session connectivity assurance method and system for virtual machine live migration scenarios
CN114679461B (en) * 2022-03-30 2023-02-14 浙江大学 Method and system for ensuring cloud network session connectivity for virtual machine live migration scenarios

Also Published As

Publication number Publication date
CN104199716B (en) 2018-01-16
CN104199716A (en) 2014-12-10

Similar Documents

Publication Publication Date Title
CN104199716B (en) A kind of method of live migration of virtual machine, physical host and system
US9529773B2 (en) Systems and methods for enabling access to extensible remote storage over a network as local storage via a logical storage controller
US9430268B2 (en) Systems and methods for supporting migration of virtual machines accessing remote storage devices over network via NVMe controllers
US9294567B2 (en) Systems and methods for enabling access to extensible storage devices over a network as local storage via NVME controller
US9311230B2 (en) Local direct storage class memory access
KR101602519B1 (en) Virtualized storage assignment method
JP6055310B2 (en) Virtual memory target offload technology
US8830870B2 (en) Network adapter hardware state migration discovery in a stateful environment
CN103250395B (en) Asynchronous virtual machine replication method and device
US9588807B2 (en) Live logical partition migration with stateful offload connections using context extraction and insertion
US20180032249A1 (en) Hardware to make remote storage access appear as local in a virtualized environment
CN106557444B (en) Method and device for realizing SR-IOV network card and method and device for realizing dynamic migration
CN115221089A (en) Apparatus, method and computer program product for efficient software-defined network acceleration processing using storage local to a host
WO2017067404A1 (en) Method of creating mirror file in logical volume, method of initiating virtual machine, and server
CN113312143A (en) Cloud computing system, command processing method and virtualization simulation device
CN111988230A (en) Virtual machine communication method, device and system and electronic equipment
CN111782318A (en) Remote mount local disk to cloud desktop virtual machine shared access system and method
CN102917055A (en) Method and device of online migration of virtual machine, and terminal device
CN112422357A (en) State detection method and device for network communication path
JP5439435B2 (en) Computer system and disk sharing method in the computer system
US10120594B1 (en) Remote access latency in a reliable distributed computing system
CN110704163A (en) A server and its virtualized storage method and device
CN117135189A (en) Server access methods and devices, storage media, electronic equipment
US11055115B1 (en) Virtual machine conversion in distributed system of remote desktops
US11595464B2 (en) Migration of network file copy operations between host computing devices

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: 14898956

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 14898956

Country of ref document: EP

Kind code of ref document: A1