Method for checkpointing in computer system under distributed processing environment

Claim 3

The method of claim 1, wherein the executing step also stops the normal processing of each process during the checkpointing processing.

Claim 4

The method of claim 2, further comprising the step of: resuming the inter-process communications and the normal processing of each process when the checkpointing processing for each process is finished.

Claim 5

The method of claim 1, further comprising the step of recording dirty pages to which a writing is carried out by the normal processing of each process since a last checkpoint; wherein the executing step executes the checkpointing processing for storing only the dirty pages recorded at the recording step into a checkpoint, in parallel to the normal processing.

Claim 6

The method of claim 4, further comprising the step of stopping the normal processing of each process when an access request to an unstored dirty page is issued by the normal processing of each process during the checkpointing processing for each process.

Claim 7

The method of claim 5, further comprising the step of resuming the normal processing for each process when the unstored dirty page is stored by the checkpointing processing.

Claim 8

The method of claim 4, wherein when an access request to an unstored dirty page is issued by the normal processing during the checkpointing processing, the checkpointing processing stores the unstored dirty page with priority over other dirty pages.

Claim 9

The method of claim 1, wherein the distributed processing system includes a checkpoint command process for controlling start and stopping of the checkpointing processing, the method further comprising the steps of: sending an inter-process stop signal to said checkpoint command process from each of said plurality of processes upon stopping of the inter-process communications at said corresponding one of said plurality of processes, wherein the normal processing of said each of said plurality of processes is continued until an intra-process stop signal is received by said each of said plurality of processes; and sending, from said checkpoint command process to said each of said plurality of processes, said intra-process stop signal when said inter-process stop signal has been received from every one of said plurality of processes.

Claim 16

The method of claim 11, wherein the executing step also stops the normal processing of each process during the checkpointing processing.

Claim 17

The method of claim 12, further comprising the step of: resuming the inter-process communications and the normal processing of each process when the checkpointing processing for each process is finished.

Claim 18

The method of claim 11, further comprising the step of recording dirty pages to which a writing is carried out by the normal processing of each process since a last checkpoint; wherein the executing step executes the checkpointing processing for storing only the dirty pages recorded at the recording step into a checkpoint, in parallel to the normal processing.

Claim 19

The method of claim 14, further comprising the step of stopping the normal processing of each process when an access request to an unstored dirty page is issued by the normal processing of each process during the checkpointing processing for each process.

Claim 20

The method of claim 15, further comprising the step of resuming the normal processing for each process when the unstored dirty page is stored by the checkpointing processing.

Claim 21

The method of claim 14, wherein when an access request to an unstored dirty page is issued by the normal processing during the checkpointing processing, the checkpointing processing stores the unstored dirty page with priority over other dirty pages.

Claim 26

A computer system, comprising: a plurality of process means operating under a distributed processing environment while carrying out inter-process communications; and checkpointing command means for issuing an inter-process communication stop command commanding a stopping of the inter-process communications of each process means, and issuing a checkpointing command commanding an execution of a checkpointing processing for each process means when responses to the inter-process communication stop commands from all process means are received; wherein each process means stops the inter-process communications of each process means while continuing a normal processing of each process means upon receiving the inter-process communication stop command and returns a response to the inter-process communication stop command when the inter-process communications of each process means are stopped, and starts an execution of the checkpointing processing for each process means upon receiving the checkpointing command.

Claim 27

A computer system, comprising: a plurality of process means operating under a distributed processing environment while carrying out inter-process communications; and checkpointing command means for issuing an inter-process communication stop command commanding a stopping of the inter-process communications of each process means, and issuing a checkpointing command commanding an execution of a checkpointing processing for each process means when responses to the inter-process communication stop commands from all process means are received; wherein each process means stops the inter-process communications of each process means while continuing a normal processing of each process means upon receiving the inter-process communication stop command and returns a response to the inter-process communication stop command when the inter-process communications of each process means are stopped, and starts an execution of the checkpointing processing for each process means upon receiving the checkpointing command.

Claim 28

The computer system of claim 20, further comprising: process management means, provided on each computer constituting the computer system, for managing the process means provided on each computer; and wherein the checkpointing command means is provided on one computer constituting the computer system, and issues the inter-process communication stop command and the checkpointing command for each process means through the process management means which manages each process means.

Claim 29

A computer system, comprising: a plurality of process means provided on computers, the process means being operating under a distributed processing environment while carrying out inter-process communications including inter-computer type inter-process communications to be carried out among the computers and intra-computer type inter-process communications to be carried out within one computer; and checkpointing command means for issuing an inter-computer type inter-process communication stop command commanding a stopping of the inter-computer type inter-process communications of each process means, and issuing a checkpointing command commanding an execution of a checkpointing processing for each process means and a stopping of the intra-computer type inter-process communications of each process means when responses to the inter-computer type inter-process communication stop commands from all process means are received; wherein each process means stops the inter-computer type inter-process communications of each process means while continuing the intra-computer type inter-process communications and a normal processing of each process means upon receiving the inter-computer type inter-process communication stop command and returns a response to the inter-computer type inter-process communication stop command when the inter-computer type inter-process communications of each process means are stopped, and starts an execution of the checkpointing processing for each process means while stopping the intra-computer type inter-process communications of each process means upon receiving the checkpointing command.

Claim 30

A computer system, comprising: a plurality of process means provided on computers, the process means being operating under a distributed processing environment while carrying out inter-process communications including inter-computer type inter-process communications to be carried out among the computers and intra-computer type inter-process communications to be carried out within one computer; and checkpointing command means for issuing an inter-computer type inter-process communication stop command commanding a stopping of the inter-computer type inter-process communications of each process means, and issuing a checkpointing command commanding an execution of a checkpointing processing for each process means and a stopping of the intra-computer type inter-process communications of each process means when responses to the inter-computer type inter-process communication stop commands from all process means are received; wherein each process means stops the inter-computer type inter-process communications of each process means while continuing the intra-computer type inter-process communications and a normal processing of each process means upon receiving the inter-computer type inter-process communication stop command and returns a response to the inter-computer type inter-process communication stop command when the inter-computer type inter-process communications of each process means are stopped, and starts an execution of the checkpointing processing for each process means while stopping the intra-computer type inter-process communications of each process means upon receiving the checkpointing command.