초록 ▼

A computer system is interrupted, and its entire state information is extracted as one or more checkpoints at one or more respective points during operation of the system. The checkpoint may be restored into the system at any later time, even multiple times, and it may also even be loaded into one o

A computer system is interrupted, and its entire state information is extracted as one or more checkpoints at one or more respective points during operation of the system. The checkpoint may be restored into the system at any later time, even multiple times, and it may also even be loaded into one or more other systems; all systems loaded with the same checkpoint will then execute from the same checkpointed state. The state extraction mechanism is preferably a virtual machine monitor, on which one or more virtual machines are installed, each virtual machine constituting an encapsulated, virtualized computer system whose states can be checkpointed under control of the virtual machine monitor. Checkpoints may be stored on a portable memory device or transmitted as a batch or dynamically over a network so that even virtual machines installed at different sites may execute from the same state.

대표청구항 ▼

1. A computer system comprising:a hardware processor;physical system storage;a virtual machine monitor functionally connected to the hardware processor and physical system storage;a virtual machine functionally connected to the virtual machine monitor and including:a virtual processor that has a plu

1. A computer system comprising:a hardware processor;physical system storage;a virtual machine monitor functionally connected to the hardware processor and physical system storage;a virtual machine functionally connected to the virtual machine monitor and including:a virtual processor that has a plurality of registers and executes a sequence of virtual processor instructions, virtual processor instructions being conveyed by the virtual machine monitor to the hardware processor, physical execution of the virtual processor instructions being carried out by the hardware processor;a virtual memory;a virtual operating system;at least one virtual peripheral device having a plurality of peripheral settings and peripheral device data;at least one application program loaded in the virtual machine and executing on the virtual machine under the control of the virtual operating system; in which:at the completion of execution of each processor instruction by the virtual processor, the virtual machine having a total state that includes state information of the virtual processor, of the virtual memory, of the virtual operating system, of each application program, and of each virtual peripheral device;the virtual machine monitor including a checkpointing mechanism:for interrupting the virtual processor;for sensing a checkpoint request for the virtual machine;for generating at least one checkpoint, each checkpoint comprising a list of the total state of the virtual machine corresponding to the respective checkpoint request, at the completion of execution of a respective virtual processor instructions;for storing each generated checkpoint;for sensing a state restoration request corresponding to one of the checkpoints; andupon sensing the state restoration request, for restoring into the virtual machine the total state corresponding to the checkpoint that in turn corresponds to the state restoration request, the virtual machine thereby resuming execution from the restored total state, that is, of the sequence of virtual processor instructions following the virtual processor instruction after whose completion the checkpoint was generated, identically as it would have if the virtual processor had not been interrupted;the physical system storage includes a checkpoint storage portion accessible only by the virtual machine monitor; andeach generated checkpoint is stored in the checkpoint storage portion, stored checkpoints thereby remaining unchanged by other access to the physical system storage, the total state of the virtual machine excluding the contents of the corresponding checkpoint storage portion. 2. A computer system comprising:a hardware processor;physical system storage;a virtual machine monitor functionally connected to the hardware processor and physical system storage;a virtual machine functionally connected to the virtual machine monitor and including:a virtual processor that has a plurality of registers and executes a sequence of virtual processor instructions, virtual processor instructions being conveyed by the virtual machine monitor to the hardware processor, physical execution of the virtual processor instructions being carried out by the hardware processor;a virtual memory;a virtual operating system;at least one application program executing on the virtual machine; in which:at the completion of execution of each processor instruction by the virtual processor, the virtual machine having a total state that includes state information of the virtual processor, of the virtual memory, of the virtual operating system, and of each application program;the virtual machine monitor including a checkpointing mechanism:for interrupting the virtual processor;for sensing a checkpoint request for the virtual machine; andfor generating at least one checkpoint, each checkpoint comprising a list of the total state of the virtual machine corresponding to the respective checkpoint request, at the completion of execution of a respective virtual proc essor instruction;in which;one of the virtual machines is a source virtual machine, for which at least one checkpoint is generated;the checkpointing mechanism is further provided for loading into at least one secondary virtual machine, which is different from the source virtual machine, the total state of the source virtual machine corresponding to one of the generated checkpoints. 3. A computer system as in claim 2, in which:the source virtual machine is located within a source computer; andthe secondary virtual machine is located within a secondary computer that contains a secondary virtual machine that is structurally substantially identical to the source virtual machine. 4. A computer system as in claim 3, in which the source and secondary computers have separate hardware processors and physical system storage. 5. A computer system as in claim 4, further including a transmission medium via which the checkpoint is transferred from the source virtual machine to the secondary virtual machine. 6. A computer system as in claim 5, in which the transmission medium is a portable memory device. 7. A computer system as in claim 5, in which the transmission medium is a network to which the source and secondary computers are connected. 8. A computer system as in claim 2, in which the virtual machine is independent of an underlying system hardware platform that includes the hardware processor and the physical system storage, the total state of any source virtual machine thereby being exportable to any other secondary virtual machine. 9. A computer system as in claim 2, further including a plurality of peripheral system hardware devices,in which:for each of the plurality of peripheral system hardware devices, the virtual machine monitor includes a respective device emulation module corresponding to a predetermined emulation of a representative, virtualized hardware device corresponding to the peripheral system hardware device;the device emulation modules form a device interface between the virtual machines and the peripheral system hardware devices;the virtual machine monitor further includes a conversion module converting device control signals and data from any virtual machine in a format of a representative virtualized hardware device into corresponding signals and data into a format of the corresponding peripheral system hardware device, and vice versa,each virtual machine thereby being independent of an underlying system hardware platform that includes the hardware processor, the physical system storage, and the peripheral system hardware devices, the total state of any source virtual machine thereby being exportable to any other secondary virtual machine. 10. In a computer system that includes:a hardware platform, which in turn includes a hardware processor and system storage;a plurality of virtual machines monitor functionally connected to the hardware platform;at least one virtual machine functionally connected to a virtual machine monitor and communicating with the hardware platform via the virtual machine monitor, each virtual machine having a virtual processor, which executes a series of virtual processor instructions, and a virtual memory; anda method comprising the following steps:in the virtual machine monitor, sensing at least one checkpoint request relating to a source virtual machine included among the plurality of virtual machines;upon sensing each checkpoint request, interrupting the virtual processor of the source virtual machine and generating a corresponding checkpoint, which comprises a list of a total state of the source virtual machine, including state information of the source virtual machine's virtual processor, virtual memory, and of any application running on the source virtual machine;loading into at least one secondary virtual machine, selected from among the plurality of virtual machines, a checkpoint corresponding to the total state of a different one of the virtual machines, that is, a source virtual ma chine;resuming unmodified execution of the source virtual machine after the step of generating each checkpoint, as well as starting execution of each secondary virtual machine from the checkpointed state of the source virtual machine, whereby the state of the source virtual machine is replicated in the secondary virtual machines. 11. A method as in claim 10, in which the source virtual machine and at least one of the secondary virtual machines are running on different virtual machine monitors. 12. A method as in claim 11, in which the different virtual machine monitors are installed on different hardware platforms. 13. A method as in claim 12, further including the step of transferring the checkpoint from the source virtual machine to each secondary virtual machine via a transmission medium. 14. A method as in claim 13, further including the step of storing at least one checkpoint on a portable memory device, which forms the transmission medium. 15. A method as in claim 13, in which the transmission medium is a network to which the hardware platforms on which the source virtual machine and each secondary machine are installed are connected for data transfer. 16. In a computer system that includes:a hardware platform, which in turn includes a hardware processor and system storage;a virtual machine monitor functionally connected to the hardware platform;at least one virtual machine functionally connected to the virtual machine monitor and communicating with the hardware platform via the virtual machine monitor, the virtual machine having a virtual processor, which executes a series of virtual processor instructions, and a virtual memory; andat least one application program running on at least one of the virtual machines;a method comprising the following steps:in the virtual machine monitor, sensing at least one checkpoint request;upon sensing each checkpoint request, interrupting the virtual processor of the virtual machine and generating a corresponding checkpoint, which comprises a list of a total state of the virtual machine, including state information of the virtual processor, the virtual memory, and of any application running on the virtual machinein which the steps of sensing each checkpoint request and generating a corresponding checkpoint includes the following sub-steps:generating an initial checkpoint including the total state of the virtual machine; andfor at least one subsequent checkpoint after the initial checkpoint, generating the subsequent checkpoint as a list of state changes relative to an immediately preceding generated checkpoint as well as link information to the immediately preceding generated checkpoint. 17. A method as in claim 16, further including the following steps:designating for deletion at least one checkpoint;if the initial checkpoint is designated for deletion, incorporating into each checkpoint immediately following the designated checkpoints all the total state information in the initial checkpoint;incorporating into each checkpoint immediately following the designated checkpoints all state change information in the checkpoints designated for deletion; anddeleting all designated checkpoints. 18. A method as in claim 16, further including the following steps:designating one of the subsequent checkpoints as a commit checkpoint;incorporating into the commit checkpoint all the total state information in the initial checkpoint as well as all state change information in all checkpoints between the initial checkpoint and the commit checkpoint, the commit checkpoint thereby becoming a new initial checkpoint. 19. In a computer system that includes:a hardware platform, which in turn includes a hardware processor and system storage;a virtual machine monitor functionally connected to the hardware platform;at least one virtual machine functionally connected to:the virtual machine monitor and communicating with the hardware platform via the virtual machine monitor, the virtual machine having a virtual proc essor, which executes a series of virtual processor instructions, and a virtual memory; andat least one application program running on at least one of the virtual machines;a method comprising the following steps:in the virtual machine monitor, sensing at least one checkpoint request;upon sensing each checkpoint request, interrupting the virtual processor of the virtual machine and generating a corresponding checkpoint, which comprises a list of a total state of the virtual machine, including state information of the virtual processor, the virtual memory, and of any application running on the virtual machine;generating and storing a plurality of checkpoints corresponding to the same virtual machine, each subsequently generated checkpoint being generated without changing any earlier generated checkpoint; andconcurrently loading the same stored checkpoint into a plurality of virtual machines, each virtual machine thereby being able to proceed from the loaded checkpoint independently of the other virtual machines. 20. A method as in claim 19, further including the step of restoring an arbitrary one of the stored checkpoints into an arbitrary one of the virtual machines. 21. In a computer system that includes:a hardware platform, which in turn includes a hardware processor and system storage;at least one peripheral system hardware device;a virtual machine monitor functionally connected to the hardware platform;a virtual machine functionally connected to the virtual machine monitor and communicating with the hardware platform via the virtual machine monitor, each virtual machine having a virtual processor, which executes a series of virtual processor instructions, and a virtual memory; andat least one application program running in the virtual machine;a method comprising the following steps:for each peripheral system hardware device, installing in the virtual machine monitor a respective device emulation module corresponding to a predetermined emulation of a representative, virtualized hardware device corresponding to the peripheral system hardware device;in the virtual machine monitor, converting device control signals and data from the virtual machine issued in a format of the respective representative virtualized hardware device into corresponding signals and data into a format of the corresponding peripheral system hardware device, and vice versa;the virtual machine thereby being independent of an underlying hardware platform, the total state of the virtual machine thereby being exportable to any other, secondary virtual machine;in the virtual machine monitor, sensing at least one checkpoint request relating to the virtual machine, which is a source virtual machine;upon sensing each checkpoint request, interrupting the virtual processor of the source virtual machine and generating a corresponding checkpoint, which comprises a list of a total state of the source virtual machine, including state information of the virtual processor, the virtual memory, and of any application running on the source virtual machine; andloading the checkpoint of the state of the source virtual machine into at least one other, secondary virtual machine, the secondary virtual machine thereby being able to proceed from the checkpoint independently of all other virtual machines.