CN-122019092-A - Computer cluster parallel processing method of reverse time migration program

CN122019092ACN 122019092 ACN122019092 ACN 122019092ACN-122019092-A

Abstract

The embodiment of the application discloses a computer cluster parallel processing method of a reverse time migration program, which comprises the steps that a main node determines a specified gun and distributes gun data processing tasks corresponding to the specified gun to a computing node according to a list file containing gun information to be processed, a mark file in a first global file catalog and a mark file in a second global file catalog in a global disk, wherein the first global file catalog is used for storing mark files of guns which are subjected to reverse time migration processing on each computing node, the second global file catalog is used for storing mark files of guns which are subjected to reverse time migration processing on each computing node, and the computing node calls a single gun reverse time migration program to execute reverse time migration processing on the corresponding specified gun according to the received gun data processing tasks and updates the mark files in the first global file catalog and/or the second global file catalog according to processing states. By the mode, the processing efficiency of the multi-gun reverse time migration task can be improved.

Inventors

ZHANG MINGQIANG
WANG XINLEI
WANG SHENGHOU
HUANG QUANCHENG
CHENG YAO
SUN LEIMING
JIAO XUMING
XU QIANG
WANG ZHENG
MA DEZHI
WANG WEI
WANG HAIKUN
RUAN FUMING

Assignees

中海油田服务股份有限公司

Dates

Publication Date: 20260512
Application Date: 20260127

Claims (10)

1. A computer cluster parallel processing method of a reverse time migration program, comprising: The method comprises the steps that a master node determines a specified cannon according to a list file containing cannon information to be processed in a global disk, a mark file in a first global file catalog and a mark file in a second global file catalog, and distributes cannon data processing tasks corresponding to the specified cannon to a computing node; The first global file catalogue is used for storing the mark files of the cannons which are subjected to reverse time migration processing on each computing node, and the second global file catalogue is used for storing the mark files of the cannons which are subjected to reverse time migration processing on each computing node; And the computing node responds to the received cannon data processing task, invokes a single cannon reverse time migration program to execute reverse time migration processing for the corresponding appointed cannon, and updates the mark files in the first global file catalog and/or the second global file catalog according to the processing state.
2. The method for parallel processing of computer clusters in a reverse time migration program of claim 1, further comprising: After the reverse time migration processing aiming at the designated cannon is completed, the computing node carries out partial superposition processing and gather extraction processing on the reverse time migration processing result to respectively obtain a node partial superposition result and a node gather extraction result; The method comprises the steps that a master node collects node part superposition results and node gather extraction results of all computing nodes, performs superposition processing on the collected node part superposition results to obtain global superposition results, and performs gather sorting processing on the collected node gather extraction results to obtain global gather extraction results.
3. The computer cluster parallel processing method of a reverse time migration program according to claim 1, wherein the determining, by the master node, a specified shot according to a list file containing shot information to be processed in a global disk, a flag file in a first global file directory, and a flag file in a second global file directory, and distributing a shot data processing task corresponding to the specified shot to a computing node further comprises: After any one of the plurality of computing nodes finishes reverse time migration processing, the master node determines a specified cannon according to the list file in the global disk, the mark file under the first global file directory and the mark file under the second global file directory, and distributes cannon data processing tasks corresponding to the specified cannon to the computing nodes.
4. The method for parallel processing of computer clusters in a reverse time migration program of claim 1, further comprising: under the condition that the gun data operation is interrupted, the master node deletes all the mark files under the first global file directory, and each computing node deletes temporary data generated by incomplete reverse time migration processing.
5. The method for parallel processing of computer clusters in a reverse time migration program according to claim 1, wherein the flag file of the gun is identified as a file name by the gun.
6. The method for computer cluster parallel processing of a reverse time migration program according to any one of claims 1 to 5, wherein the method further comprises: The method comprises the steps that a computing node creates an industrial area directory on a local disk, and creates a first local file directory and a second local file directory under the industrial area directory; the computing node stores the mark file of the specified shot which is subjected to reverse time migration processing in a first local file catalogue corresponding to the computing node, and the mark file of the specified shot which is subjected to reverse time migration processing is moved from the first local file catalogue corresponding to the computing node to a second local file catalogue.
7. The computer cluster parallel processing method of a reverse time migration program of claim 6, wherein the computing node performing the partial overlap-and-add processing and the gather extraction processing on the reverse time migration processing result further comprises: And the computing node judges whether a gun mark corresponding to any mark file in the second local file catalog exists in the mark array, if not, the computing node performs partial superposition processing and gather extraction processing on a reverse time migration processing result of the gun corresponding to the mark file, and records the gun mark corresponding to the mark file into the mark array after the processing is finished.
8. A computing device comprising a processor, a memory, a communication interface, and a communication bus, the processor, the memory, and the communication interface completing communication with each other over the communication bus; The memory is configured to store at least one executable instruction, where the executable instruction causes the processor to perform the operations corresponding to the computer cluster parallel processing method of the reverse time migration program according to any one of claims 1 to 7.
9. A computer storage medium having stored therein at least one executable instruction for causing a processor to perform operations corresponding to the computer cluster parallel processing method of a reverse time migration program according to any one of claims 1-7.
10. A computer program product comprising at least one executable instruction for causing a processor to perform operations corresponding to the computer cluster parallel processing method of a reverse time migration program according to any one of claims 1 to 7.

Description

Computer cluster parallel processing method of reverse time migration program Technical Field The embodiment of the application relates to the technical field of exploration geophysics, in particular to a computer cluster parallel processing method of a reverse time migration program. Background Reverse time migration is an important prestack depth migration imaging method in the field of seismic exploration, and is widely applied to complex structure imaging processing. The occupation and the calculated amount of the reverse time migration to the calculation resources are relatively high, a computer cluster is required to be used for processing in actual production, and the reverse time migration processing generally calls tens or hundreds of calculation nodes at a time and takes a plurality of hours to hundreds of hours to be different, so that the requirements on the stability, the flexibility and the efficiency of the parallel operation of the cluster are extremely high. Cluster parallel operation control relies on a deep understanding of the computer cluster hardware and software system, involves specialized programming techniques, and optimization experience based on large-scale practical production applications, which are often difficult for developers of reverse time migration method technology to adequately meet. Therefore, there is a need for a parallel transformation method of reverse time migration program computer clusters, which can be widely applied to various reverse time migration programs, various computer cluster hardware structures and various practical application conditions. Disclosure of Invention The present application has been made in view of the above problems, and it is an object of the present application to provide a computer cluster parallel processing method, a computing device, a computer storage medium and a computer program product of a reverse time migration program that overcomes or at least partially solves the above problems. According to an aspect of the embodiment of the present application, there is provided a computer cluster parallel processing method of a reverse time migration program, including: The method comprises the steps that a master node determines a specified cannon according to a list file containing cannon information to be processed in a global disk, a mark file in a first global file catalog and a mark file in a second global file catalog, and distributes cannon data processing tasks corresponding to the specified cannon to a computing node; The first global file catalogue is used for storing the mark files of the cannons which are subjected to reverse time migration processing on each computing node, and the second global file catalogue is used for storing the mark files of the cannons which are subjected to reverse time migration processing on each computing node; And the computing node responds to the received cannon data processing task, invokes a single cannon reverse time migration program to execute reverse time migration processing for the corresponding appointed cannon, and updates the mark files in the first global file catalog and/or the second global file catalog according to the processing state. Optionally, the method further comprises: After the reverse time migration processing aiming at the designated cannon is completed, the computing node carries out partial superposition processing and gather extraction processing on the reverse time migration processing result to respectively obtain a node partial superposition result and a node gather extraction result; The method comprises the steps that a master node collects node part superposition results and node gather extraction results of all computing nodes, performs superposition processing on the collected node part superposition results to obtain global superposition results, and performs gather sorting processing on the collected node gather extraction results to obtain global gather extraction results. Optionally, the determining, by the master node, a specified cannon according to the list file containing the cannon information to be processed in the global disk, the flag file in the first global file directory, and the flag file in the second global file directory, and distributing, to the computing node, a cannon data processing task corresponding to the specified cannon further includes: After any one of the plurality of computing nodes finishes reverse time migration processing, the master node determines a specified cannon according to the list file in the global disk, the mark file under the first global file directory and the mark file under the second global file directory, and distributes cannon data processing tasks corresponding to the specified cannon to the computing nodes. Optionally, the method further comprises: under the condition that the gun data operation is interrupted, the master node deletes all the mark files under the first global file directory, and each