WO2021179522A1 - Système, procédé et appareil d'attribution de ressources informatiques, et dispositif informatique - Google Patents

Système, procédé et appareil d'attribution de ressources informatiques, et dispositif informatique Download PDF

Info

Publication number
WO2021179522A1
WO2021179522A1 PCT/CN2020/106078 CN2020106078W WO2021179522A1 WO 2021179522 A1 WO2021179522 A1 WO 2021179522A1 CN 2020106078 W CN2020106078 W CN 2020106078W WO 2021179522 A1 WO2021179522 A1 WO 2021179522A1
Authority
WO
WIPO (PCT)
Prior art keywords
distribution
node
result
task
allocation
Prior art date
Application number
PCT/CN2020/106078
Other languages
English (en)
Chinese (zh)
Inventor
周俊
Original Assignee
平安国际智慧城市科技股份有限公司
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 平安国际智慧城市科技股份有限公司 filed Critical 平安国际智慧城市科技股份有限公司
Publication of WO2021179522A1 publication Critical patent/WO2021179522A1/fr

Links

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F9/00—Arrangements for program control, e.g. control units
    • G06F9/06—Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
    • G06F9/46—Multiprogramming arrangements
    • G06F9/48—Program initiating; Program switching, e.g. by interrupt
    • G06F9/4806—Task transfer initiation or dispatching
    • G06F9/4843—Task transfer initiation or dispatching by program, e.g. task dispatcher, supervisor, operating system
    • G06F9/4881—Scheduling strategies for dispatcher, e.g. round robin, multi-level priority queues
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F9/00—Arrangements for program control, e.g. control units
    • G06F9/06—Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
    • G06F9/46—Multiprogramming arrangements
    • G06F9/50—Allocation of resources, e.g. of the central processing unit [CPU]
    • G06F9/5005—Allocation of resources, e.g. of the central processing unit [CPU] to service a request
    • G06F9/5011—Allocation of resources, e.g. of the central processing unit [CPU] to service a request the resources being hardware resources other than CPUs, Servers and Terminals
    • G06F9/5016—Allocation of resources, e.g. of the central processing unit [CPU] to service a request the resources being hardware resources other than CPUs, Servers and Terminals the resource being the memory
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F2209/00—Indexing scheme relating to G06F9/00
    • G06F2209/50—Indexing scheme relating to G06F9/50
    • G06F2209/5017—Task decomposition
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F2209/00—Indexing scheme relating to G06F9/00
    • G06F2209/50—Indexing scheme relating to G06F9/50
    • G06F2209/5018—Thread allocation

Definitions

  • This application relates to the field of artificial intelligence, in particular to a computing resource allocation system, method, device, computer equipment, and storage medium.
  • Computer resource allocation is an important content in the computer field.
  • the traditional computer resource allocation method is to obtain the computer task submitted by the user (for example, a main program) at a master node.
  • the master node divides the computer task into multiple subtasks (subthreads) and assigns them to other nodes to perform the tasks separately. Execute to achieve distributed thread division.
  • the inventor found that the traditional computer resource allocation method is based on the main node to complete the task division, so when the main node is too stressed, for example, when multiple users submit multiple computer tasks at the same time, the main node will not be able to smoothly Completing the task division, resulting in delays and other abnormal conditions, resulting in poor operation of the entire computer system.
  • the main purpose of this application is to provide a computing resource allocation system, method, device, computer equipment, and storage medium to ensure the smoothness of the overall system operation.
  • this application proposes a computing resource allocation system, including multiple resource allocation nodes, task acquisition nodes, and multiple task execution nodes, wherein the multiple resource allocation nodes include a first allocation node and a second allocation node.
  • the multiple resource allocation nodes include a first allocation node and a second allocation node.
  • the multiple task execution nodes are respectively connected to the task acquisition node in signal, n is an integer greater than 1, and i is greater than 1 and less than An integer equal to n;
  • the first distribution node is configured to, when the first distribution node is in a normal state, the first distribution node is in a signal connection with the task acquisition node through a preset first channel;
  • the second distribution node is configured to, when the first distribution node is not in a normal state, the second distribution node is in signal connection with the task acquisition node through a preset second channel, wherein the second distribution node is The channel includes the first channel;
  • the i-th distribution node is configured to, when the i-1th distribution node is not in a normal state, the i-th distribution node is in signal connection with the task acquisition node through a preset i-th channel, wherein the The i-th channel includes the first channel, the second channel, ..., the i-1th channel;
  • the nth distribution node is configured to, when the n-1th distribution node is not in a normal state, the nth distribution node is in signal connection with the task acquisition node through a preset nth channel, wherein the The nth channel includes the first channel, the second channel, ..., the n-1th channel;
  • the task acquisition node is configured to acquire a computer task input by a user, send the computer task to the multiple resource allocation nodes, receive the allocation result of the computer task by the multiple resource allocation nodes, and according to all The allocation result allocates the computer task to the multiple task execution nodes.
  • the present application provides a computing resource allocation method, which is applied to any one of the task acquisition nodes described above, and includes:
  • the multiple subtasks are sent to the multiple task execution nodes respectively.
  • the present application provides a computing resource allocation device, which is applied to the task acquisition node as described in any one of the above, and includes:
  • the computer task acquisition unit is used to acquire the computer task input by the user
  • a computer task sending unit configured to send the computer task to the first distribution node through the first channel
  • the receiving judgment unit is configured to judge whether the first distribution result sent by the first distribution node is received within a first preset time, wherein the first distribution result is that the first distribution node performs the computer task The result of the allocation;
  • the first allocation result judgment unit is configured to, if the first allocation result sent by the first allocation node is received within the first preset time, determine whether the first allocation result is based on a preset allocation result judgment method normal;
  • a split processing unit configured to perform split processing on the computer task according to the first assignment result if the first assignment result is normal, so as to obtain multiple subtasks
  • the subtask sending unit is configured to send the multiple subtasks to the multiple task execution nodes respectively.
  • the present application provides a computer device, including a memory and a processor, the memory stores a computer program, and the processor implements a computing resource allocation method when the computer program is executed, and the computing resource allocation method is applied to claim 1 or 2.
  • the task acquisition node includes: acquiring a computer task input by a user; sending the computer task to the first distribution node through the first channel; The first distribution result sent by the first distribution node, where the first distribution result is the result of the first distribution node distributing the computer task; if the first distribution result is received within the first preset time The first distribution result sent by the distribution node is determined according to the preset distribution result judgment method to determine whether the first distribution result is normal; if the first distribution result is normal, the computer will be processed according to the first distribution result.
  • the task is split and processed to obtain multiple subtasks; the multiple subtasks are respectively sent to the multiple task execution nodes.
  • the present application provides a computer-readable storage medium on which a computer program is stored, and when the computer program is executed by a processor, a method for allocating computing resources is implemented, and the method for allocating computing resources is applied to the task according to claim 1 or 2.
  • Obtaining a node includes: obtaining a computer task input by a user; sending the computer task to the first distribution node through the first channel; determining whether the first distribution node is received within a first preset time The first distribution result sent, where the first distribution result is the result of the first distribution node distributing the computer task; if the first distribution node receives the first distribution result within the first preset time For an allocation result, judge whether the first allocation result is normal according to the preset allocation result judgment method; if the first allocation result is normal, then split the computer task according to the first allocation result , Thereby obtaining multiple subtasks; sending the multiple subtasks to the multiple task execution nodes respectively.
  • the computing resource allocation system, method, device, computer equipment, and storage medium of the present application wherein the system includes multiple resource allocation nodes, task acquisition nodes, and multiple task execution nodes, wherein the multiple resource allocation nodes include a first allocation node , The second distribution node,..., the i-th distribution node,..., the n-th distribution node; the plurality of task execution nodes are respectively connected to the task acquisition node in signal, n is an integer greater than 1, and i is An integer greater than 1 and less than or equal to n; the i-th distribution node is configured to, when the i-1th distribution node is not in a normal state, the i-th distribution node communicates with the i-th channel through a preset A task acquisition node signal connection, wherein the i-th channel includes the first channel, the second channel, ..., the i-1th channel; the task acquisition node is configured to acquire a computer task input by a user, and The computer task is sent to the multiple resource allocation nodes, the
  • Figure 1 is a schematic block diagram of the structure of a computing resource allocation system according to an embodiment of this application;
  • FIG. 2 is a schematic flowchart of a computing resource allocation method according to an embodiment of this application.
  • FIG. 3 is a schematic block diagram of the structure of a computing resource allocation device according to an embodiment of the application.
  • FIG. 4 is a schematic block diagram of the structure of a computer device according to an embodiment of the application.
  • an embodiment of the present application provides a computing resource allocation system, including multiple resource allocation nodes, task acquisition nodes, and multiple task execution nodes, wherein the multiple resource allocation nodes include a first allocation node and a second allocation node.
  • the multiple resource allocation nodes include a first allocation node and a second allocation node.
  • the multiple task execution nodes are respectively connected to the task acquisition node in signal, n is an integer greater than 1, and i is greater than 1 and less than An integer equal to n;
  • the first distribution node is configured to, when the first distribution node is in a normal state, the first distribution node is in a signal connection with the task acquisition node through a preset first channel;
  • the second distribution node is configured to, when the first distribution node is not in a normal state, the second distribution node is in signal connection with the task acquisition node through a preset second channel, wherein the second distribution node is The channel includes the first channel;
  • the i-th distribution node is configured to, when the i-1th distribution node is not in a normal state, the i-th distribution node is in signal connection with the task acquisition node through a preset i-th channel, wherein the The i-th channel includes the first channel, the second channel, ..., the i-1th channel;
  • the nth distribution node is configured to, when the n-1th distribution node is not in a normal state, the nth distribution node is in signal connection with the task acquisition node through a preset nth channel, wherein the The nth channel includes the first channel, the second channel, ..., the n-1th channel;
  • the task acquisition node is configured to acquire a computer task input by a user, send the computer task to the multiple resource allocation nodes, receive the allocation result of the computer task by the multiple resource allocation nodes, and according to all The allocation result allocates the computer task to the multiple task execution nodes.
  • the computer resource allocation system of the present application adopts a special setting (see Figure 1), that is, the resource allocation nodes are arranged in a stepwise and hierarchical manner.
  • the lower-level allocation nodes are not in a normal state.
  • the resource allocation work is transferred to the high-level allocation node in time to ensure the normal operation of the system.
  • the adopted signal channel still includes the original signal channel, which saves the time of resetting.
  • the signal channel may be a physical signal channel (for example, a physical signal line, etc.), or a signal channel that does not have a physical signal (for example, a wireless signal channel).
  • the mutual inclusion relationship between the first channel, the second channel and the nth channel can be set by means of a signal switch, for example, the first channel and the second channel are divided by A signal switch is set at the intersection of a part of the channel, and the signal switch can control the direction of the signal (that is, when the first distribution node is in a normal state, the first distribution node is connected to the task acquisition node; when the first distribution node is not In the normal state, connect the non-first distribution node such as the second distribution node to the task acquisition node).
  • the second channel includes the first channel is understood to mean that the second channel includes all of the first channel, or the second channel includes at least a part of the first channel.
  • the first channel is, for example, a first routing structure (path) between a task acquisition node and the first distribution node, and the first routing structure may be a dynamic routing It can also be a static route.
  • a second routing structure is constructed on the basis of the first routing structure, so that the task acquisition node is signally connected to the second distribution node.
  • the second channel includes the first channel is understood to mean that the second channel includes all of the first channel, or the second channel includes at least a part of the first channel.
  • the node in this application can be any feasible node, such as an independent computer terminal.
  • the execution subject in this application is the task acquisition node, and the task acquisition node serves as the hub of the entire system. It is signal-connected to multiple task execution nodes, is signal-connected to multiple resource allocation nodes, and is also used to obtain computer tasks input by users. .
  • the task acquisition node divides the work that needs to be assigned to computer tasks, and multiple resource allocation nodes, so the actual load of the task acquisition node is not large, compared to the traditional solution where all the work is concentrated on one node Completed, the computing resource allocation system of this application is more stable.
  • the multiple task execution nodes in the present application receive the assigned subtasks, they can also perform data coordination among the multiple task execution nodes.
  • the implementation of the data coordination is, for example, monitoring the available memory data of the task execution node, and when the available memory data is less than a predetermined value, sending at least part of the received subtasks to other task execution nodes.
  • the memory of the first allocation node, the second allocation node,..., the i-th allocation node,..., the n-th allocation node in the computing resource allocation system of the present application increases sequentially.
  • special resource allocation nodes are arranged in a stepwise manner among each other, that is, to reduce the impact on the entire system when the allocation nodes are in an abnormal state.
  • the memory of multiple distribution nodes is increased sequentially, so that the computing power of multiple distribution nodes is sequentially increased. In this way, when the distribution node of the previous level is abnormal, the distribution node of the latter level is more likely to bear Assignment tasks that cannot be completed by the assignment node at the previous level.
  • the CPU frequencies corresponding to the first distribution node, the second distribution node, ..., the i-th distribution node, ..., and the n-th distribution node respectively increase in order.
  • the computing resource allocation system of the present application includes multiple resource allocation nodes, task acquisition nodes, and multiple task execution nodes, wherein the multiple resource allocation nodes include a first allocation node, a second allocation node,..., The i-th distribution node,..., the n-th distribution node; the multiple task execution nodes are respectively connected to the task acquisition node in signal, n is an integer greater than 1, and i is an integer greater than 1 and less than or equal to n;
  • the i-th distribution node is configured to, when the i-1th distribution node is not in a normal state, the i-th distribution node is in signal connection with the task acquisition node through a preset i-th channel, wherein the The i channel includes the first channel, the second channel, ..., the i-1th channel; the task acquisition node is configured to acquire a computer task input by a user, and send the computer task to the plurality of
  • the resource allocation node receives the allocation result of the computer task by the multiple resource allocation no
  • an embodiment of the present application provides a computing resource allocation method, which is applied to the task acquisition node described in any one of the foregoing (that is, any one of the foregoing embodiments), including:
  • This application can use any feasible method to obtain the computer task input by the user, for example, input through a preset input device, for example, using a physical interface such as a U disk interface for input, or input computer tasks through remote signal transmission.
  • the computer task is, for example, a complete program.
  • the first channel may be a signal channel with entities or a signal channel without entities.
  • the computer task is first sent to the first allocation node through the first channel.
  • the first distribution node will complete the distribution of the computer task within the first preset time and send it to the task acquisition node.
  • the first distribution node Under abnormal circumstances, for example, the first distribution node crashes due to overload, then the first distribution node cannot complete the distribution and sending of computer tasks within the first preset time. From this, it is possible to determine for the first time whether the first distribution node is normal.
  • the present application further performs a second judgment on the first allocation result sent by the first allocation node to determine whether the first allocation result is normal.
  • the judgment method may be any feasible method, such as judging whether the first distribution result conforms to a normal distribution, and if the first distribution result conforms to a normal distribution, it is determined that the first distribution result is normal.
  • the computer task is split and processed according to the first allocation result to obtain multiple subtasks, and the multiple subtasks are respectively sent to the multiple task execution nodes to complete the allocation of computing resources .
  • This application can be applied to any feasible scenario, for example: a user submits an application or task to the computing resource allocation system of this application, a task acquisition node in the computing resource allocation system receives the application or task, and the task acquisition node applies the application Or the task is sent to the first distribution node through the first channel; the first distribution node uses a preset task divider to divide, and then sends the result of the division to the task acquisition node; the task acquisition node according to the result of the division, the application or The task is specifically divided to obtain multiple subtasks; the task acquisition node sends the task acquisition node to multiple task execution nodes; the task execution node constructs threads corresponding to the subtasks and executes them in the thread pool.
  • Serialization is the process of converting data into a form that can be stored or transmitted. During serialization, data is written to temporary or persistent storage. Later, you can recreate the data by deserializing it from the storage area.
  • the task execution node can obtain specific subtasks through deserialization.
  • the status and progress of the execution of the subtask are also sent to the task acquisition node, so that the task acquisition node can grasp the running status of the task execution node in time.
  • the instructions for logging off and closing the computer task can be generated by the task execution node or the task acquisition node, thereby completing the distribution and final execution of the entire computer task.
  • step S2 of sending the computer task to the first distribution node through the first channel includes:
  • S201 Send an available memory acquisition request to the multiple task execution nodes respectively;
  • S203 Determine whether there is abnormal memory data with a value less than a preset memory threshold in the plurality of available memory data
  • the computer task is sent to the first distribution node through the first channel.
  • This application further uses abnormal memory data and abnormal task execution node elimination, and uses available memory data as a basis for computing resource allocation to improve the accuracy of allocation. Since this application uses available memory data, it can realize dynamic computing resource allocation, which is more flexible and more efficient than traditional static computing resource allocation. Among them, because the task execution node is also running at full capacity, the available memory is extremely small at this time, that is, abnormal memory data less than the preset memory threshold, so it should not be listed as the actual task execution node, so it should be excluded . Then, through the first channel, the computer task and the multiple normally available memory data are sent to the first allocation node, so that the first allocation node has an accurate basis for computing resource allocation.
  • the resource allocation nodes are arranged in a stepwise and hierarchical manner.
  • the first channel is modified to the second channel and passes through the second channel.
  • Channel sending the computer task to the second allocation node, so that the computer resource allocation can continue to be executed.
  • the second distribution result is completed in the second distribution node, but the splitting of computer tasks is completed in the task acquisition node, thereby avoiding the waste of additional signal transmission and further improving the overall Calculate the efficiency of resource allocation and execution.
  • the first allocation result includes the data allocation amount corresponding to the multiple task execution nodes, and the step S4 of judging whether the first allocation result is normal according to a preset allocation result judgment method ,include:
  • S402 Send an available memory acquisition request to the multiple task execution nodes respectively, and acquire multiple available memory data returned by the multiple task execution nodes respectively;
  • S403 Arrange the multiple task execution nodes in descending order according to the size of the multiple available memory data, so as to obtain a second sequence
  • S404 Calculate the Levinstein distance between the first sequence and the second sequence, and determine whether the Levinstein distance is greater than a preset distance threshold;
  • the first distribution result is the analysis result of how the first distribution node should distribute the computer tasks.
  • This application judges whether it is normal or not, so as to improve the overall fluency.
  • this application adopts descending order of multiple task execution nodes in the first allocation result according to the size of the data allocation amount Arrange to obtain the first sequence; according to the size of the multiple available memory data, arrange the multiple task execution nodes in descending order to obtain the second sequence; then calculate the first sequence and the second sequence Levinstein distance, and use Levinstein distance as the basis for judging whether it is normal.
  • the Levinstein distance is a distance that measures the degree of difference between two sequences, and refers to the minimum number of editing operations required to transform one sequence into another sequence.
  • the editing operations include replacing, inserting, and deleting.
  • the minimum value of the Levinstein distance is 0.
  • the Levinstein distance is 0, it indicates that the two sequences are exactly the same, that is, the division of multiple task execution nodes in the first allocation result and the available memory of multiple task execution nodes The data is completely corresponding, so it is judged that the first allocation result is normal; when the Levinstein distance is too large (that is, the Levinstein distance is greater than the preset distance threshold), it indicates that the two sequences are not the same, so the first allocation is judged
  • the result is not normal.
  • the preset distance threshold is, for example, 2-10, preferably 3-5.
  • the computing resource allocation method of the present application is applied to the task acquisition node in the computing resource allocation system described in any one of the foregoing, so as to pass the special setting in the computing resource allocation system, that is, the system includes multiple resource allocation nodes, A task acquisition node and a plurality of task execution nodes, wherein the plurality of resource allocation nodes include a first allocation node, a second allocation node, ..., the i-th allocation node, ..., the n-th allocation node;
  • Each task execution node is signally connected to the task acquisition node, n is an integer greater than 1, and i is an integer greater than 1 and less than or equal to n;
  • the i-th distribution node is configured to: When the node is not in a normal state, the i-th distribution node is signally connected to the task acquisition node through a preset i-th channel, where the i-th channel includes the first channel, the second channel, ..., Channel i-1; the task acquisition node is configured
  • an embodiment of the present application provides a computing resource allocation device, which is applied to the task acquisition node according to any one of the above, and includes:
  • the computer task obtaining unit 10 is used to obtain the computer task input by the user;
  • the computer task sending unit 20 is configured to send the computer task to the first distribution node through the first channel;
  • the receiving judging unit 30 is configured to judge whether a first distribution result sent by the first distribution node is received within a first preset time, wherein the first distribution result is the task of the first distribution node to the computer The result of the allocation;
  • the first allocation result judgment unit 40 is configured to, if the first allocation result sent by the first allocation node is received within the first preset time, judge the first allocation result according to a preset allocation result judgment method Is it normal
  • the split processing unit 50 is configured to, if the first distribution result is normal, perform split processing on the computer task according to the first distribution result, so as to obtain multiple subtasks;
  • the subtask sending unit 60 is configured to send the multiple subtasks to the multiple task execution nodes respectively.
  • the computer task sending unit 20 includes:
  • a memory acquisition request sending subunit configured to send available memory acquisition requests to the multiple task execution nodes respectively;
  • a memory data acquisition subunit for acquiring multiple available memory data returned by the multiple task execution nodes respectively;
  • An abnormal memory data judging subunit used for judging whether there is abnormal memory data with a value less than a preset memory threshold in the plurality of available memory data
  • the normal available memory data acquisition subunit is configured to remove the abnormal memory data to obtain a plurality of normal available memory data if there is abnormal memory data with a value less than a preset memory threshold in the plurality of available memory data, and The task execution node corresponding to the abnormal memory data is excluded from the multiple task execution nodes;
  • the computer task sending subunit is configured to send the computer task and the multiple normally available memory data to the first distribution node through the first channel.
  • the device includes:
  • the computer task resending unit is configured to send the computer task to the second channel through the second channel if the first distribution result sent by the first distribution node is not received within the first preset time Two distribution nodes;
  • the second allocation result receiving and judging unit is configured to judge whether a second allocation result sent by the second allocation node is received within a second preset time, wherein the second allocation result is that the second allocation node has The result of the assignment of the computer tasks;
  • a second allocation result judging unit configured to judge whether the second allocation result is normal if the second allocation result sent by the second allocation node is received within the second preset time
  • a plurality of subtask acquiring units are configured to, if the second distribution result is normal, split the computer task according to the second distribution result, thereby obtaining a plurality of subtasks, and send the plurality of subtasks to multiple subtasks respectively.
  • a task execution node configured to, if the second distribution result is normal, split the computer task according to the second distribution result, thereby obtaining a plurality of subtasks, and send the plurality of subtasks to multiple subtasks respectively.
  • the first allocation result includes data allocation amounts corresponding to the multiple task execution nodes, respectively, and the first allocation result judgment unit 40 includes:
  • the first sequence obtaining subunit is configured to sort the multiple task execution nodes in the first distribution result in descending order according to the size of the data distribution amount, so as to obtain the first sequence;
  • the available memory data obtaining subunit is configured to send available memory obtaining requests to the multiple task execution nodes, and obtain multiple available memory data returned by the multiple task execution nodes respectively;
  • the second sequence obtaining subunit is configured to arrange the plurality of task execution nodes in descending order according to the size of the plurality of available memory data, so as to obtain the second sequence;
  • a Levinstein distance calculation subunit configured to calculate the Levinstein distance between the first sequence and the second sequence, and determine whether the Levinstein distance is greater than a preset distance threshold
  • the first distribution result normal determination subunit is configured to determine that the first distribution result is normal if the Levinstein distance is not greater than a preset distance threshold.
  • the computing resource allocation device of the present application is applied to the task acquisition node in the computing resource allocation system described in any one of the foregoing, so as to pass the special setting in the computing resource allocation system, that is, the system includes multiple resource allocation nodes, A task acquisition node and a plurality of task execution nodes, wherein the plurality of resource allocation nodes include a first allocation node, a second allocation node, ..., the i-th allocation node, ..., the n-th allocation node;
  • Each task execution node is signally connected to the task acquisition node, n is an integer greater than 1, and i is an integer greater than 1 and less than or equal to n;
  • the i-th distribution node is configured to: When the node is not in a normal state, the i-th distribution node is signally connected to the task acquisition node through a preset i-th channel, where the i-th channel includes the first channel, the second channel, ..., Channel i-1; the task acquisition node is configured
  • an embodiment of the present application also provides a computer device.
  • the computer device may be a server, and its internal structure may be as shown in the figure.
  • the computer equipment includes a processor, a memory, a network interface, and a database connected through a system bus. Among them, the processor designed by the computer is used to provide calculation and control capabilities.
  • the memory of the computer device includes a non-volatile storage medium and an internal memory.
  • the non-volatile storage medium stores an operating system, a computer program, and a database.
  • the memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium.
  • the database of the computer equipment is used to store the data used in the computing resource allocation method.
  • the network interface of the computer device is used to communicate with an external terminal through a network connection.
  • the computer program is executed by the processor to realize a computing resource allocation method.
  • the foregoing processor executes the foregoing computing resource allocation method, wherein the steps included in the method respectively correspond to the steps of executing the computing resource allocation method of the foregoing embodiment, and will not be repeated here.
  • the computer device of the present application implements a computing resource allocation method, and the method is applied to the task acquisition node in the computing resource allocation system described in any one of the foregoing, so as to pass a special setting in the computing resource allocation system, that is, where
  • the system includes multiple resource allocation nodes, task acquisition nodes, and multiple task execution nodes, where the multiple resource allocation nodes include a first allocation node, a second allocation node,..., an i-th allocation node,..., The nth distribution node; the multiple task execution nodes are respectively connected to the task acquisition node in signal connection, n is an integer greater than 1, and i is an integer greater than 1 and less than or equal to n; the i-th distribution node is configured to, When the i-1th distribution node is not in a normal state, the i-th distribution node is signally connected to the task acquisition node through a preset i-th channel, wherein the i-th channel includes the first channel, The second channel,..., the i-1
  • An embodiment of the present application also provides a computer-readable storage medium on which a computer program is stored, and when the computer program is executed by a processor, a method for allocating computing resources is implemented, wherein the steps included in the method are respectively the same as those in performing the calculation in the foregoing embodiment.
  • the steps of the resource allocation method correspond to each other, so I won't repeat them here.
  • the computer-readable storage medium may be non-volatile or volatile.
  • the computer-readable storage medium of the present application implements a computing resource allocation method, and the method is applied to the task acquisition node in the computing resource allocation system described in any one of the foregoing, so as to pass special settings in the computing resource allocation system , That is, the system includes multiple resource allocation nodes, task acquisition nodes and multiple task execution nodes, wherein the multiple resource allocation nodes include a first allocation node, a second allocation node,..., the i-th allocation node,.
  • the nth distribution node the plurality of task execution nodes are respectively connected to the task acquisition node signal, n is an integer greater than 1, i is an integer greater than 1 and less than or equal to n; the i-th distribution node is It is configured that, when the i-1th distribution node is not in a normal state, the i-th distribution node is in signal connection with the task acquisition node through a preset i-th channel, wherein the i-th channel includes the A channel, a second channel, ..., the i-1th channel; the task acquisition node is configured to acquire a computer task input by a user, send the computer task to the plurality of resource allocation nodes, and receive According to the allocation result of the multiple resource allocation nodes to the computer task, the computer task is allocated to the multiple task execution nodes according to the allocation result. So as to ensure the smoothness of the overall system operation.
  • Non-volatile memory may include read only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory.
  • Volatile memory may include random access memory (RAM) or external cache memory.
  • RAM is available in many forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), dual-rate SDRAM (SSRSDRAM), enhanced SDRAM (ESDRAM), synchronous Link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.

Abstract

La présente invention se rapporte au domaine de l'intelligence artificielle, et concerne un système, un procédé et un appareil d'attribution de ressources informatiques, ainsi qu'un dispositif informatique et un support d'informations. Le système comprend une pluralité de nœuds d'attribution de ressources, un nœud d'obtention de tâche et une pluralité de nœuds d'exécution de tâche ; la pluralité de nœuds d'attribution de ressources comprenant un premier nœud d'attribution, ..., et un n-ième nœud d'attribution ; la pluralité de nœuds d'exécution de tâche sont signalés séparément au nœud d'obtention de tâche ; l'i-ème nœud d'attribution est configuré pour être signalé au nœud d'obtention de tâche au moyen d'un i-ème canal prédéfini lorsque l'(i-1)-ième nœud d'attribution n'est pas dans un état normal, et l'i-ème canal comprend un premier canal.., et an (i-1)-ième canal ; le nœud d'obtention de tâche est configuré pour obtenir une entrée de tâche informatique par un utilisateur, envoyer la tâche informatique à la pluralité de nœuds d'attribution de ressources, recevoir un résultat d'attribution, et attribuer la tâche informatique à la pluralité de nœuds d'exécution de tâche. Par conséquent, la régularité de fonctionnement de l'ensemble du système est assuré.
PCT/CN2020/106078 2020-03-13 2020-07-31 Système, procédé et appareil d'attribution de ressources informatiques, et dispositif informatique WO2021179522A1 (fr)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN202010177684.8A CN111506400A (zh) 2020-03-13 2020-03-13 计算资源分配***、方法、装置和计算机设备
CN202010177684.8 2020-03-13

Publications (1)

Publication Number Publication Date
WO2021179522A1 true WO2021179522A1 (fr) 2021-09-16

Family

ID=71875763

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2020/106078 WO2021179522A1 (fr) 2020-03-13 2020-07-31 Système, procédé et appareil d'attribution de ressources informatiques, et dispositif informatique

Country Status (2)

Country Link
CN (1) CN111506400A (fr)
WO (1) WO2021179522A1 (fr)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN115328663A (zh) * 2022-10-10 2022-11-11 亚信科技（中国）有限公司 基于PaaS平台进行资源调度的方法、装置、设备和存储介质

Families Citing this family (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113254199A (zh) * 2021-05-07 2021-08-13 埃森智能科技（深圳）有限公司 一种多任务同时处理方法、***及其设备

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104731663A (zh) * 2015-03-31 2015-06-24 北京奇艺世纪科技有限公司 任务处理方法及***
US20160239341A1 (en) * 2015-02-12 2016-08-18 International Business Machines Corporation Dynamic correlated operation management for a distributed computing system
CN106909451A (zh) * 2017-02-28 2017-06-30 郑州云海信息技术有限公司 一种分布式任务调度***及方法
CN107707592A (zh) * 2017-01-24 2018-02-16 贵州白山云科技有限公司 任务处理方法、节点及内容分发网络
CN108304255A (zh) * 2017-12-29 2018-07-20 北京城市网邻信息技术有限公司 分布式任务调度方法及装置、电子设备及可读存储介质
CN110673938A (zh) * 2019-09-23 2020-01-10 腾讯科技（深圳）有限公司 任务处理方法、***、服务器及存储介质

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20160239341A1 (en) * 2015-02-12 2016-08-18 International Business Machines Corporation Dynamic correlated operation management for a distributed computing system
CN104731663A (zh) * 2015-03-31 2015-06-24 北京奇艺世纪科技有限公司 任务处理方法及***
CN107707592A (zh) * 2017-01-24 2018-02-16 贵州白山云科技有限公司 任务处理方法、节点及内容分发网络
CN106909451A (zh) * 2017-02-28 2017-06-30 郑州云海信息技术有限公司 一种分布式任务调度***及方法
CN108304255A (zh) * 2017-12-29 2018-07-20 北京城市网邻信息技术有限公司 分布式任务调度方法及装置、电子设备及可读存储介质
CN110673938A (zh) * 2019-09-23 2020-01-10 腾讯科技（深圳）有限公司 任务处理方法、***、服务器及存储介质

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN115328663A (zh) * 2022-10-10 2022-11-11 亚信科技（中国）有限公司 基于PaaS平台进行资源调度的方法、装置、设备和存储介质
CN115328663B (zh) * 2022-10-10 2023-01-03 亚信科技（中国）有限公司 基于PaaS平台进行资源调度的方法、装置、设备和存储介质

Also Published As

Publication number Publication date
CN111506400A (zh) 2020-08-07

Similar Documents

Publication Publication Date Title
US11003507B2 (en) Mapreduce job resource sizing using assessment models
US8185908B2 (en) Dynamic scheduling in a distributed environment
CN105912399B (zh) 一种任务处理方法、装置及***
CN102835068B (zh) 用于管理***资源的重新分配的方法和装置
CN111367630A (zh) 一种基于云计算的多用户多优先级的分布式协同处理方法
CN109408590B (zh) 分布式数据库的扩容方法、装置、设备及存储介质
WO2021179522A1 (fr) Système, procédé et appareil d'attribution de ressources informatiques, et dispositif informatique
WO2024021489A1 (fr) Procédé et appareil de planification de tâches, et planificateur kubernetes
WO2021180092A1 (fr) Procédé et appareil de répartition des tâches
WO2023077750A1 (fr) Procédé et appareil d'attribution d'une tâche de calcul de réseau neuronal parmi des ressources hétérogènes, et dispositif
CN112463390A (zh) 一种分布式任务调度方法、装置、终端设备及存储介质
US8316375B2 (en) Load-balancing of processes based on inertia
WO2024022142A1 (fr) Procédé et appareil d'utilisation de ressources
US9672073B2 (en) Non-periodic check-pointing for fine granular retry of work in a distributed computing environment
CN117234733A (zh) 一种分布式***任务分配方法、***、存储介质及设备
CN110275777B (zh) 一种资源调度***
CN107645421A (zh) 一种分布式存储的iscsi协议实现方法
CN115469979B (zh) 量子控制***的调度装置、方法以及量子计算机
TW201627873A (zh) 用於在分散式計算中處理重發請求的方法與設備
CN113485828B (zh) 基于quartz的分布式任务调度***及方法
CN115292176A (zh) 一种压力测试方法、装置、设备及存储介质
KR102201651B1 (ko) 태스크 지역성과 다운스트림 상태를 고려한 확률 기반의 데이터 스트림 파티셔닝 방법 및 이를 기록한 기록매체
CN108809699B (zh) 一种实现大规模虚拟网络节点重复映射的方法
KR100795593B1 (ko) 동적 순위 매김 및 그룹핑을 이용한 그리드 컴퓨팅 시스템및 그 데이터 처리 방법
KR20200094838A (ko) 빅데이터 처리 장치 및 방법

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: 20924308

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

32PN Ep: public notification in the ep bulletin as address of the adressee cannot be established

Free format text: NOTING OF LOSS OF RIGHTS PURSUANT TO RULE 112(1) EPC (EPO FORM 1205 DATED 16/01/2023)

122 Ep: pct application non-entry in european phase

Ref document number: 20924308

Country of ref document: EP

Kind code of ref document: A1