Supercomputer residual process clearing method, device, server and storage medium
By identifying and clearing the residual processes of computing nodes in the supercomputer system, the resource occupation problem caused by the compatibility issues of the job scheduling system and the network is solved, and the operation reliability and efficiency of user jobs are improved.
Patent Information
- Application Number
- CN202110706402.3
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2021-06-24
- Publication Date
- 2024-11-05
- Estimated Expiration
- 2041-06-24
AI Technical Summary
Due to the compatibility problem between the job scheduling system LSF and the system core network OPA in the supercomputer system, a large number of irregular processes remain on the computing nodes, occupying CPU and network resources, affecting the operation efficiency and reliability of new jobs.
By determining the current and historical job information, identifying the computing nodes to be detected, and clearing the residual processes according to the calibration node list, periodically clearing the residual processes to ensure the accurate job status and improving the reliability and efficiency of user jobs.
Effectively clear the residual processes of the computing node, avoid job errors and low computing efficiency caused by residual processes, improve the operation reliability of user jobs, save resources, and reduce the number of repeated detection and clearance.
Smart Images

Figure CN113553148B_ABST
Abstract
Description
Technical Field
[0001] The embodiments of the present invention relate to the field of computer technology, and in particular to a method, device, server and storage medium for clearing residual processes of a supercomputer. Background Art
[0002] The operation of the supercomputer system requires unified management and control of supercomputer resources through the job scheduling system LSF (Load Sharing Facility), but in reality, due to the compatibility issues between the job scheduling system LSF and the system core network OPA (Omni-Path Architecture), a large number of residual jobs are left on the computing nodes; the supercomputer computing nodes run for a long time, which will generate junk processes; residual processes will also be generated due to reasons such as irregular jobs submitted by users and irregular operations of users to clean up jobs; that is, irregular and abnormal processes, namely residual processes, appear on the computing nodes.
[0003] When the job scheduling system LSF detects that the user jobs on these computing nodes have been calculated, it will mark the node status of the computing nodes as idle and deliver new user jobs to these computing nodes. But in fact, the residual processes of these computing nodes will occupy CPU (Central Processing Unit) and network resources, so new jobs will cause job errors or low computing efficiency, or cannot run normally because of the residual processes occupying resources. In addition, some user jobs will generate new jobs during the calculation process. These new jobs can use the resources of other computing nodes without being scheduled by the job scheduling system LSF. This also makes the node status of the computing nodes recorded in the job scheduling system LSF different from the actual status of the computing nodes, causing the job scheduling system LSF to make errors when delivering new jobs. Summary of the invention
[0004] In view of this, an embodiment of the present invention provides a method, device, server and storage medium for clearing residual processes of a supercomputer, so as to clear residual processes of computing nodes and improve the reliability of user job operation.
[0005] In a first aspect, an embodiment of the present invention provides a method for clearing residual processes of a supercomputer, comprising:
[0006] Determine the current job information that is running at the current moment;
[0007] Acquire historical job information running at a historical moment, where the historical moment is separated from the current moment by a preset time period;
[0008] Determine a computing node to be detected according to the current job information and the historical job information;
[0009] The residual processes of the computing nodes to be detected are cleared according to the calibration node list.
[0010] Further, the current job information includes at least one current job number, the historical job information includes a maximum value of a historical running job number, and determining the computing node to be detected according to the current job information and the historical job information includes:
[0011] If the current job number is greater than the maximum value of the historical running job numbers, the computing node corresponding to the current job number is used as the computing node to be detected.
[0012] Further, determining the computing node to be detected according to the current job information and the historical job information further includes:
[0013] If the current job number is less than or equal to the maximum value of the historical running job number, determining the historical running status of the current job number;
[0014] If the historical running state is suspended, the computing node corresponding to the current job number is used as the computing node to be detected.
[0015] Furthermore, before clearing the residual process of the to-be-detected computing node according to the calibration node list, the method further includes:
[0016] Generate a calibration node list based on the normal job information submitted normally.
[0017] Further, clearing the residual process of the to-be-detected computing node according to the calibration node list includes:
[0018] Determine at least one to-be-detected user name corresponding to the to-be-detected computing node;
[0019] Determine a calibration user list according to the computing nodes to be detected and the calibration node list;
[0020] If the user name to be detected is not in the calibration user list, the process corresponding to the user name to be detected is used as a residual process;
[0021] Clear the residual process.
[0022] Further, after the residual processes of the to-be-detected computing nodes are cleared according to the calibration node list, the method further includes:
[0023] The at least one current job number and process information of the residual process are recorded.
[0024] In a second aspect, an embodiment of the present invention provides a supercomputer residual process clearing device, comprising:
[0025] The current information acquisition module is used to determine the current job information that is running at the current moment;
[0026] A historical information acquisition module, used to acquire historical job information running at historical moments, where the historical moments are separated from the current moment by a preset time period;
[0027] A to-be-detected node determination module, used to determine a to-be-detected computing node according to the current job information and the historical job information;
[0028] The residual process clearing module is used to clear the residual process of the computing node to be detected according to the calibration node list.
[0029] Further, the current job information includes at least one current job number, the historical job information includes a maximum value of a historical running job number, and the to-be-detected node determination module is specifically used to:
[0030] If the current job number is greater than the maximum value of the historical running job numbers, the computing node corresponding to the current job number is used as the computing node to be detected.
[0031] In a third aspect, an embodiment of the present invention provides a server, including:
[0032] one or more processors;
[0033] a storage device for storing one or more programs,
[0034] When the one or more programs are executed by the one or more processors, the one or more processors implement the supercomputer residual process clearing method provided by any embodiment of the present invention.
[0035] In a fourth aspect, an embodiment of the present invention provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the method for clearing residual processes of a supercomputer provided by any embodiment of the present invention.
[0036] The method for clearing residual processes of a supercomputer provided by an embodiment of the present invention determines the current job information being run at the current moment; obtains historical job information running at historical moments, wherein the historical moments are separated from the current moment by a preset time period; determines the computing node to be detected according to the current job information and the historical job information; and clears the residual processes of the computing node to be detected according to the calibration node list. The method realizes the clearing of residual processes of computing nodes, and the residual processes can be cleared periodically through a preset time period to ensure that user jobs will not run wrongly due to residual processes, thereby improving the reliability of user job operation; and performs residual process detection on the node where the newly added job is located through the current job information and the historical job information. If there is a residual process, it is cleared, that is, all computing nodes are not repeatedly detected, thereby improving the efficiency of clearing residual processes and saving resources. BRIEF DESCRIPTION OF THE DRAWINGS
[0037] Figure 1 A schematic diagram of a flow chart of a method for clearing residual processes of a supercomputer provided in Embodiment 1 of the present invention;
[0038] Figure 2 A schematic diagram of a flow chart of a method for clearing residual processes of a supercomputer provided in Embodiment 2 of the present invention;
[0039] Figure 3 A schematic diagram of the structure of a supercomputer residual process clearing device provided in Embodiment 3 of the present invention;
[0040] Figure 4 A schematic diagram of the structure of a server provided in Embodiment 4 of the present invention. DETAILED DESCRIPTION
[0041] The present invention will be further described in detail below in conjunction with the accompanying drawings and embodiments. It is to be understood that the specific embodiments described herein are only used to explain the present invention, rather than to limit the present invention. It should also be noted that, for ease of description, only parts related to the present invention, rather than all structures, are shown in the accompanying drawings.
[0042] It should be mentioned before discussing the exemplary embodiments in more detail that some exemplary embodiments are described as processes or methods depicted as flow charts. Although the flow charts describe the steps as sequential processes, many of the steps therein can be implemented in parallel, concurrently or simultaneously. In addition, the order of the steps can be rearranged. The process can be terminated when its operation is completed, but can also have additional steps not included in the accompanying drawings. The process can correspond to a method, function, procedure, subroutine, subprogram, etc.
[0043] In addition, the terms "first", "second", etc. may be used in this article to describe various directions, actions, steps or elements, but these directions, actions, steps or elements are not limited by these terms. These terms are only used to distinguish a first direction, action, step or element from another direction, action, step or element. The terms "first", "second", etc. should not be understood as indicating or implying relative importance or implicitly indicating the number of indicated technical features. Thus, the features defined as "first" and "second" may explicitly or implicitly include one or more of the features. In the description of the present invention, the meaning of "multiple" and "batch" is at least two, such as two, three, etc., unless otherwise clearly and specifically defined.
[0044] Embodiment 1
[0045] Figure 1 This is a flow chart of a method for removing residual processes of a supercomputer provided in the first embodiment of the present invention. This embodiment can be applied to the removal of residual processes on each computing node of a supercomputer system. Figure 1 As shown, the method for clearing residual processes of a supercomputer provided by the first embodiment of the present invention includes:
[0046] S110: Determine the current job information that is currently being executed.
[0047] Specifically, the current job information refers to the job information of the user job currently running on each computing node of the supercomputer system, such as the job number (JOB ID) of the user job.
[0048] S120, obtaining historical job information running at historical moments, where the historical moments are separated from the current moment by a preset time period.
[0049] Specifically, historical job information refers to job information of user jobs running on each computing node at a historical moment within a preset time period from the current moment. The preset time period is equivalent to the operating cycle of the supercomputer residual process clearing method provided in an embodiment of the present invention. For example, the current moment is 10:30, the preset time period is 5 minutes, and the historical moment is 10:25. The content of the historical job information is consistent with the content of the current job information. If the current job information includes the job number of the user job running at the current moment (referred to as the current job number), the historical job information includes the job number of the user job running at the historical moment (referred to as the historical job number).
[0050] S130. Determine a computing node to be detected according to the current job information and the historical job information.
[0051] Specifically, the newly added job information is determined according to the current job information and the historical job information, and the computing nodes corresponding to the newly added job information are used as the computing nodes to be detected, and the computing nodes to be detected are the computing nodes that need to clear the residual processes. For example, the current job numbers included in the current job information are: 04, 05, 06, 07, 08, 09, 10, 11, and the historical job numbers included in the historical job information are: 01, 02, 03, 04, 05, 06, then the newly added job numbers included in the newly added job information are: 07, 08, 09, 10, 11, and the computing nodes corresponding to the newly added job numbers 07, 08, 09, 10, 11 are used as the computing nodes to be detected.
[0052] S140: Clear the residual processes of the computing nodes to be detected according to the calibration node list.
[0053] Specifically, the calibration node list records the relevant information of the user jobs that are normally submitted and running on each computing node at the current moment, that is, normal job information, such as the normally submitted job number, calibration user name (user), calibration computing node, etc. First, determine the user name to be detected corresponding to the computing node to be detected and the calibration user name corresponding to the calibration computing node that is the same as the computing node to be detected in the calibration node list. The computing node is represented by the node number (node ID). The same as the computing node to be detected in the calibration node list means that the node ID of the calibration computing node is the same as the node ID of the computing node to be detected. Then determine whether the user name to be detected corresponding to the computing node to be detected is consistent with the calibration user name. If not, the process corresponding to the user name to be detected is used as a residual process and the residual process is cleared.
[0054] For example, the computing node to be detected is N40, and its corresponding user names to be detected include zhanga and zhangc. The calibration user names corresponding to the calibration computing node N40 in the calibration node list include zhanga and zhangb. If the user name to be detected zhanga is the same as the calibration user name zhanga, it is considered that the process corresponding to the user name to be detected zhanga is not a residual process. If the user name to be detected zhangc does not belong to any of the calibration user names, it is considered that the process corresponding to the user name to be detected zhangc is a residual process, and the process corresponding to the user name to be detected zhangc is cleared.
[0055] The method for clearing residual processes of a supercomputer provided in the first embodiment of the present invention determines the current job information being run at the current moment; obtains the historical job information running at the historical moment, and the historical moment is separated from the current moment by a preset time period; determines the computing node to be detected according to the current job information and the historical job information; and clears the residual processes of the computing node to be detected according to the calibration node list. The residual processes of the computing nodes are cleared, and the residual processes can be cleared periodically through the preset time period to ensure that the user's job will not run wrong due to the residual process, thereby improving the reliability of the user's job operation; and the node where the newly added job is located is detected for residual processes through the current job information and the historical job information. If there is a residual process, it is cleared, that is, all computing nodes are not repeatedly detected, thereby improving the efficiency of clearing residual processes and saving resources.
[0056] Embodiment 2
[0057] Figure 2 This is a flow chart of a method for removing residual processes of a supercomputer provided in the second embodiment of the present invention. This embodiment is a further refinement of the above embodiment. Figure 2 As shown, the method for clearing residual processes of a supercomputer provided by the second embodiment of the present invention includes:
[0058] S210: Determine the current job information currently being executed, where the current job information includes at least one current job number.
[0059] Specifically, the current job information refers to the job information of the user job currently running on each computing node of the supercomputer system, such as the job ID of the user job. In this embodiment, the current job information includes at least one current job ID, for example, the current job information includes: 05, 06, 07, 08, 09, 10, 11.
[0060] S220, obtaining historical job information running at historical moments, where the historical moments are separated from the current moment by a preset time period, and the historical job information includes a maximum value of a historical running job number.
[0061] In this embodiment, the historical operation information is the maximum value of the historical operation number. For example, if the historical operation numbers detected are: 01, 02, 03, 04, 05, 06, then the historical operation information is 06.
[0062] Optionally, the historical operation information may also be a historical operation number, such as the historical operation information includes: 01, 02, 03, 04, 05, 06.
[0063] S230: If the current job number is greater than the maximum value of the historical running job numbers, the computing node corresponding to the current job number is used as the computing node to be detected.
[0064] Specifically, the specific value of the supercomputer system job number is incremental, that is, the specific value of the new job number generated by the system will increase with each new user job. Therefore, if the current job number is greater than the maximum value of the historical running job number, it can be determined that the current job number is a new job number, and the computing node corresponding to the current job number (that is, the new job number) is used as the computing node to be detected.
[0065] Optionally, when the historical job information includes at least one historical job number, the current job number can be compared with the historical job number. If the current job number is different from all historical job numbers, the current job number is determined to be a new job number, and the computing node corresponding to the current job number is used as the computing node to be detected.
[0066] Optionally, when determining the computing node to be detected, it also includes: if the current job number is less than or equal to the maximum value of the historical running job number, then determining the historical running status of the current job number; if the historical running status is suspended, then taking the computing node corresponding to the current job number as the computing node to be detected.
[0067] In a supercomputer system, a user job needs to occupy a certain amount of computing resources to perform calculations. When a user job is submitted, if the system does not have enough computing resources to support the calculation of the user job, the user job will be suspended and will be run again when there are enough computing resources. For example, a user job requires 40 CPUs to be calculated, but there is only one idle computing node. Assuming that a computing node provides 20 CPUs, it is obvious that the computing node cannot meet the computing needs of the user job. In this case, the user job will be suspended and will be run again when there are two idle computing nodes.
[0068] Therefore, if the current job number is less than or equal to the historical job number, the historical running status of the current job number is queried to determine whether it was suspended due to insufficient computing resources at a historical moment, and can continue to run at the current moment due to sufficient computing resources. If the historical running status of the current job number is suspended, the computing node corresponding to the current job number is used as the computing node to be detected. Subsequently, the residual process of the computing node to be detected is cleared to avoid residual processes in the computing node assigned to the current job number, which may cause the job to fail to run normally.
[0069] S240. Generate a calibration node list according to the normal job information submitted normally.
[0070] Specifically, the calibration node list records the relevant information of the user jobs that are normally submitted and running on each computing node at the current moment, that is, normal job information, such as the normally submitted job number, calibration user name, node number (node ID) of the calibration computing node, etc.
[0071] S250: Determine at least one to-be-detected user name corresponding to the to-be-detected computing node.
[0072] Specifically, a computing node can run one user job or multiple user jobs. A user job corresponds to a job number and a user name. Therefore, the computing node to be detected includes at least one user name to be detected, such as the user names to be detected corresponding to the computing node to be detected N40 include zhanga and zhangc.
[0073] S260: Determine a calibration user list according to the computing nodes to be detected and the calibration node list.
[0074] Specifically, the calibration user list refers to all calibration user names corresponding to the calibration computing nodes that are the same as the computing node to be detected in the calibration node list. For example, if the computing node to be detected is N40, then in the calibration node list, the calibration user names corresponding to the calibration computing node N40 include zhanga and zhangb, that is, the calibration user list corresponding to the computing node to be detected includes zhanga and zhangb.
[0075] S270: If the user name to be detected is not in the calibration user list, the process corresponding to the user name to be detected is used as a residual process.
[0076] Specifically, if the user name to be detected is not in the calibration user list, it means that the user name to be detected is not a user name generated by a normally submitted user job, and it may be a residual process, so the process corresponding to the user name to be detected is taken as a residual process. For example, in the example of steps S205-S206, the user name to be detected zhanga is in the calibration user list, and the user name to be detected zhangc is not in the calibration user list, then the process corresponding to the user name to be detected zhangc is taken as a residual process.
[0077] S280: Clear the residual process.
[0078] S290. Record the at least one current job number and process information of the residual process.
[0079] Specifically, after the residual process is determined, the residual process is cleared, the computing node resources are released, and all current job numbers are recorded. The current job number is used as the historical job number for the next residual process detection. At the same time, the process information of the residual process, such as the user name, node ID, etc., is recorded to facilitate subsequent analysis of the cause of the residual process.
[0080] The supercomputer residual process clearing method provided in the second embodiment of the present invention realizes the clearing of residual processes of computing nodes. The residual processes can be periodically cleared through a preset time period to ensure that user jobs will not run incorrectly due to residual processes, thereby improving the reliability of user job operation; and residual process detection is performed on the node where the new job is located through current job information and historical job information. If there is a residual process, it is cleared, that is, repeated detection is not performed on all computing nodes, thereby improving the efficiency of clearing residual processes and saving resources.
[0081] Embodiment 3
[0082] Figure 3 This is a schematic diagram of the structure of a supercomputer residual process removal device provided in the third embodiment of the present invention. This embodiment can be applied to the cleaning of residual processes on each computing node of a supercomputer system. The supercomputer residual process removal device provided in this embodiment can implement the supercomputer residual process removal method provided in any embodiment of the present invention, and has the corresponding functional structure and beneficial effects of the implementation method. For the contents not described in detail in this embodiment, reference can be made to the description of any method embodiment of the present invention.
[0083] like Figure 3 As shown, the supercomputer residual process clearing device provided by the third embodiment of the present invention includes: a current information acquisition module 310, a historical information acquisition module 320, a to-be-detected node determination module 330 and a residual process clearing module 340, wherein:
[0084] The current information acquisition module 310 is used to determine the current job information that is currently running;
[0085] The historical information acquisition module 320 is used to acquire historical job information running at historical moments, where the historical moments are separated from the current moment by a preset time period;
[0086] The to-be-detected node determination module 330 is used to determine the to-be-detected computing node according to the current job information and the historical job information;
[0087] The residual process clearing module 340 is used to clear the residual process of the computing node to be detected according to the calibration node list.
[0088] Further, the current job information includes at least one current job number, the historical job information includes the maximum value of the historical running job number, and the to-be-detected node determination module 330 is specifically used for:
[0089] If the current job number is greater than the maximum value of the historical running job numbers, the computing node corresponding to the current job number is used as the computing node to be detected.
[0090] Furthermore, the to-be-detected node determination module 330 is further configured to:
[0091] If the current job number is less than or equal to the maximum value of the historical running job number, determining the historical running status of the current job number;
[0092] If the historical running state is suspended, the computing node corresponding to the current job number is used as the computing node to be detected.
[0093] Furthermore, it also includes:
[0094] The calibration node list generation module is used to generate a calibration node list according to the normal job information submitted normally.
[0095] Furthermore, the residual process clearing module 340 is specifically used for:
[0096] Determine at least one to-be-detected user name corresponding to the to-be-detected computing node;
[0097] Determine a calibration user list according to the computing nodes to be detected and the calibration node list;
[0098] If the user name to be detected is not in the calibration user list, the process corresponding to the user name to be detected is used as a residual process;
[0099] Clear the residual process.
[0100] Furthermore, it also includes:
[0101] An information recording module is used to record the at least one current job number and the process information of the residual process.
[0102] The supercomputer residual process clearing device provided in Example 3 of the present invention realizes the clearing of residual processes of computing nodes through the current information acquisition module, the historical information acquisition module, the node determination module to be detected and the residual process clearing module. The residual processes can be periodically cleared through a preset time period to ensure that user jobs will not run incorrectly due to residual processes, thereby improving the reliability of user job operation; and residual process detection is performed on the node where the new job is located through the current job information and the historical job information. If there is a residual process, it is cleared, that is, all computing nodes are not repeatedly detected, thereby improving the efficiency of residual process clearing and saving resources.
[0103] Embodiment 4
[0104] Figure 4 A schematic diagram of the structure of a server provided in Embodiment 4 of the present invention. Figure 4 A block diagram of an exemplary server 412 suitable for use in implementing embodiments of the present invention is shown. Figure 4The server 412 shown is only an example and should not bring any limitation to the functions and scope of use of the embodiments of the present invention.
[0105] like Figure 4 As shown, the server 412 is in the form of a general-purpose server. The components of the server 412 may include, but are not limited to: one or more processors 416, a storage device 428, and a bus 418 connecting different system components (including the storage device 428 and the processor 416).
[0106] Bus 418 represents one or more of several types of bus structures, including a storage device bus or storage device controller, a peripheral bus, an accelerated graphics port, a processor or a local bus using any of a variety of bus architectures. For example, these architectures include but are not limited to Industry Subversive Alliance (ISA) bus, Micro Channel Architecture (MAC) bus, Enhanced ISA bus, Video Electronics Standards Association (VESA) local bus and Peripheral Component Interconnect (PCI) bus.
[0107] The server 412 typically includes a variety of computer system readable media. These media can be any available media that can be accessed by the server 412, including volatile and non-volatile media, removable and non-removable media.
[0108] The storage device 428 may include computer system readable media in the form of volatile memory, such as random access memory (RAM) 430 and / or cache memory 432. The server 412 may further include other removable / non-removable, volatile / non-volatile computer system storage media. By way of example only, the storage system 434 may be used to read and write non-removable, non-volatile magnetic media ( Figure 4 not shown, usually called a "hard drive"). Although Figure 4Not shown in the figure, a disk drive for reading and writing a removable non-volatile disk (e.g., a "floppy disk"), and an optical disk drive for reading and writing a removable non-volatile optical disk, such as a read-only optical disk (Compact Disc Read-Only Memory, CD-ROM), a digital video disk (Digital Video Disc-Read Only Memory, DVD-ROM) or other optical media) may be provided. In these cases, each drive may be connected to the bus 418 via one or more data medium interfaces. The storage device 428 may include at least one program product having a set (e.g., at least one) of program modules that are configured to perform the functions of the various embodiments of the present invention.
[0109] A program / utility 440 having a set (at least one) of program modules 442 may be stored, for example, in storage device 428, such program modules 442 including, but not limited to, an operating system, one or more application programs, other program modules, and program data, each of which or some combination may include an implementation of a network environment. Program modules 442 generally perform the functions and / or methods of the embodiments described herein.
[0110] The server 412 may also communicate with one or more external devices 414 (e.g., keyboards, pointing terminals, displays 424, etc.), one or more terminals that enable users to interact with the server 412, and / or any terminals that enable the server 412 to communicate with one or more other computing terminals (e.g., network cards, modems, etc.). Such communication may be performed through an input / output (I / O) interface 422. Furthermore, the server 412 may also communicate with one or more networks (e.g., a local area network (LAN), a wide area network (WAN), and / or a public network, such as the Internet) through a network adapter 420. Figure 4 As shown, the network adapter 420 communicates with other modules of the server 412 via the bus 418. It should be understood that, although not shown in the figure, other hardware and / or software modules can be used in conjunction with the server 412, including but not limited to: microcode, terminal drivers, redundant processors, external disk drive arrays, disk arrays (Redundant Arrays of Independent Disks, RAID) systems, tape drives, and data backup storage systems.
[0111] The processor 416 executes various functional applications and data processing by running the program stored in the storage device 428, for example, implementing the supercomputer residual process clearing method provided by any embodiment of the present invention, which may include:
[0112] Determine the current job information that is running at the current moment;
[0113] Acquire historical job information running at a historical moment, where the historical moment is separated from the current moment by a preset time period;
[0114] Determine a computing node to be detected according to the current job information and the historical job information;
[0115] The residual processes of the computing nodes to be detected are cleared according to the calibration node list.
[0116] Embodiment 5
[0117] Embodiment 5 of the present invention further provides a computer-readable storage medium on which a computer program is stored. When the program is executed by a processor, the method for clearing residual processes of a supercomputer provided in any embodiment of the present invention is implemented. The method may include:
[0118] Determine the current job information that is running at the current moment;
[0119] Acquire historical job information running at a historical moment, where the historical moment is separated from the current moment by a preset time period;
[0120] Determine a computing node to be detected according to the current job information and the historical job information;
[0121] The residual processes of the computing nodes to be detected are cleared according to the calibration node list.
[0122] The computer storage medium of the embodiment of the present invention can adopt any combination of one or more computer-readable media. The computer-readable medium can be a computer-readable signal medium or a computer-readable storage medium. The computer-readable storage medium can be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device or device, or any combination of the above. More specific examples (non-exhaustive list) of computer-readable storage media include: an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In this document, a computer-readable storage medium can be any tangible medium containing or storing a program, which can be used by an instruction execution system, device or device or used in combination with it.
[0123] Computer-readable signal media may include data signals propagated in baseband or as part of a carrier wave, which carry computer-readable program code. Such propagated data signals may take a variety of forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination of the above. Computer-readable signal media may also be any computer-readable medium other than a computer-readable storage medium, which may send, propagate, or transmit a program for use by or in conjunction with an instruction execution system, apparatus, or device.
[0124] The program code embodied on the computer readable medium may be transmitted using any appropriate medium, including but not limited to wireless, wireline, optical fiber cable, RF, etc., or any suitable combination of the foregoing.
[0125] Computer program code for performing the operation of the present invention may be written in one or more programming languages or combinations thereof, including object-oriented programming languages (such as Java, Smalltalk, C++), and conventional procedural programming languages (such as "C" language or similar programming languages). The program code may be executed entirely on the user's computer, partially on the user's computer, as an independent software package, partially on the user's computer, partially on a remote computer, or entirely on a remote computer or terminal. In the case of a remote computer, the remote computer may be connected to the user's computer via any type of network (including a local area network (LAN) or a wide area network (WAN)), or may be connected to an external computer (e.g., using an Internet service provider to connect via the Internet).
[0126] Note that the above are only preferred embodiments of the present invention and the technical principles used. Those skilled in the art will understand that the present invention is not limited to the specific embodiments described herein, and that various obvious changes, readjustments and substitutions can be made by those skilled in the art without departing from the scope of protection of the present invention. Therefore, although the present invention has been described in more detail through the above embodiments, the present invention is not limited to the above embodiments, and may include more other equivalent embodiments without departing from the concept of the present invention, and the scope of the present invention is determined by the scope of the appended claims.
Claims
1. A method for removing residual processes of a supercomputer, characterized in that: include: Determine the current job information that is running at the current moment; Acquire historical job information running at a historical moment, where the historical moment is separated from the current moment by a preset time period; Determine a computing node to be detected according to the current job information and the historical job information; Clear the residual processes of the computing nodes to be detected according to the calibration node list; The current job information includes at least one current job number, the historical job information includes the maximum value of the historical running job number, and determining the computing node to be detected according to the current job information and the historical job information includes: If the current job number is greater than the maximum value of the historical running job number, the computing node corresponding to the current job number is used as the computing node to be detected; Before clearing the residual process of the to-be-detected computing node according to the calibration node list, the method further includes: Generate a calibration node list based on the normal job information submitted normally; The residual process of clearing the computing node to be detected according to the calibration node list includes: Determine at least one to-be-detected user name corresponding to the to-be-detected computing node; Determine a calibration user list according to the computing nodes to be detected and the calibration node list; If the user name to be detected is not in the calibration user list, the process corresponding to the user name to be detected is used as a residual process; Clear the residual process.
2. The method according to claim 1, characterized in that Determining the computing node to be detected according to the current job information and the historical job information further includes: If the current job number is less than or equal to the maximum value of the historical running job number, determining the historical running status of the current job number; If the historical running state is suspended, the computing node corresponding to the current job number is used as the computing node to be detected.
3. The method according to claim 1, characterized in that After the residual processes of the to-be-detected computing nodes are cleared according to the calibration node list, the method further includes: The at least one current job number and process information of the residual process are recorded.
4. A supercomputer residual process clearing device, characterized in that: include: The current information acquisition module is used to determine the current job information that is running at the current moment; A historical information acquisition module, used to acquire historical job information running at historical moments, where the historical moments are separated from the current moment by a preset time period; A to-be-detected node determination module, used to determine a to-be-detected computing node according to the current job information and the historical job information; A residual process clearing module, used to clear the residual process of the computing node to be detected according to the calibration node list; The current job information includes at least one current job number, the historical job information includes the maximum value of the historical running job number, and the to-be-detected node determination module is specifically used for: If the current job number is greater than the maximum value of the historical running job number, the computing node corresponding to the current job number is used as the computing node to be detected; Also includes: A calibration node list generation module is used to generate a calibration node list according to normal job information submitted normally; The residual process removal module is specifically used for: Determine at least one to-be-detected user name corresponding to the to-be-detected computing node; Determine a calibration user list according to the computing nodes to be detected and the calibration node list; If the user name to be detected is not in the calibration user list, the process corresponding to the user name to be detected is used as a residual process; Clear the residual process.
5. A server, characterized in that: include: one or more processors; a storage device for storing one or more programs, When the one or more programs are executed by the one or more processors, the one or more processors implement the supercomputer residual process clearing method as described in any one of claims 1-3.
6. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the program is executed by a processor, the method for clearing residual processes of a supercomputer as described in any one of claims 1 to 3 is implemented.
