[go: up one dir, main page]

CN119578850B - Data flow-based running context parameter transmission and task instance dynamic scheduling method - Google Patents

Data flow-based running context parameter transmission and task instance dynamic scheduling method Download PDF

Info

Publication number
CN119578850B
CN119578850B CN202510140290.8A CN202510140290A CN119578850B CN 119578850 B CN119578850 B CN 119578850B CN 202510140290 A CN202510140290 A CN 202510140290A CN 119578850 B CN119578850 B CN 119578850B
Authority
CN
China
Prior art keywords
task
parameter
information
data flow
instance
Prior art date
Legal status (The legal status 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 status listed.)
Active
Application number
CN202510140290.8A
Other languages
Chinese (zh)
Other versions
CN119578850A (en
Inventor
刘媛媛
陈廷权
原攀峰
陈廷梁
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Zhejiang Shuxin Network Co ltd
Original Assignee
Zhejiang Shuxin Network Co ltd
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 Zhejiang Shuxin Network Co ltd filed Critical Zhejiang Shuxin Network Co ltd
Priority to CN202510140290.8A priority Critical patent/CN119578850B/en
Publication of CN119578850A publication Critical patent/CN119578850A/en
Application granted granted Critical
Publication of CN119578850B publication Critical patent/CN119578850B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Classifications

    • GPHYSICS
    • G06COMPUTING OR CALCULATING; COUNTING
    • G06QINFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00Administration; Management
    • G06Q10/06Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
    • G06Q10/063Operations research, analysis or management
    • G06Q10/0631Resource planning, allocation, distributing or scheduling for enterprises or organisations
    • GPHYSICS
    • G06COMPUTING OR CALCULATING; COUNTING
    • G06FELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/20Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
    • G06F16/24Querying
    • G06F16/245Query processing
    • G06F16/2455Query execution
    • G06F16/24568Data stream processing; Continuous queries
    • YGENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y02TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02DCLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE
    • Y02D10/00Energy efficient computing, e.g. low power processors, power management or thermal management

Landscapes

  • Engineering & Computer Science (AREA)
  • Business, Economics & Management (AREA)
  • Human Resources & Organizations (AREA)
  • Theoretical Computer Science (AREA)
  • Strategic Management (AREA)
  • Entrepreneurship & Innovation (AREA)
  • Economics (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • General Engineering & Computer Science (AREA)
  • Development Economics (AREA)
  • Databases & Information Systems (AREA)
  • Educational Administration (AREA)
  • Data Mining & Analysis (AREA)
  • Game Theory and Decision Science (AREA)
  • Computational Linguistics (AREA)
  • Marketing (AREA)
  • Operations Research (AREA)
  • Quality & Reliability (AREA)
  • Tourism & Hospitality (AREA)
  • General Business, Economics & Management (AREA)
  • Debugging And Monitoring (AREA)

Abstract

The invention provides a data flow-based running context parameter transferring and task instance dynamic scheduling method, which relates to the technical field of data processing and comprises the steps of receiving a task scheduling request, and constructing a data flow dependency graph containing task nodes and data flow relation, and acquiring task instance running context information corresponding to each task node, wherein the task instance running context information comprises task instance identification, state information and input and output parameter information. And establishing a mapping relation between input and output parameters of the task instance by analyzing the data flow dependency graph and the runtime context information, and generating a parameter transfer rule. The method monitors task instance state information, when the state of an upstream task changes, an upstream output parameter is transmitted to a downstream input parameter according to a parameter transmission rule, and when the downstream input parameter is complete, the dynamic scheduling execution of the downstream task instance is triggered, so that the task scheduling driven by a data stream is realized, and the scheduling efficiency and the scheduling flexibility are improved.

Description

Data flow-based running context parameter transmission and task instance dynamic scheduling method
Technical Field
The invention relates to a data processing technology, in particular to a method for dynamically scheduling a running context parameter and a task instance based on a data stream.
Background
Currently, with the increasing size and complexity of data, data processing tasks are often broken down into multiple interdependent sub-tasks, forming a complex data stream. How to efficiently manage and schedule these subtasks, ensure the correct transfer of data between the different tasks, and achieve efficient execution of the entire data stream becomes an important challenge.
The existing task scheduling method mainly focuses on static scheduling, namely, the execution sequence and the dependency relationship of tasks are predefined before the tasks are executed. Such an approach is difficult to accommodate dynamic changes in the data stream, such as changes in the amount of output data or data format of upstream tasks, fluctuations in task execution time, and the like.
The prior art has the following defects and shortcomings:
1. and the dynamic data flow is difficult to process, and the static scheduling method is difficult to adapt to the dynamic change of the data flow, so that the task execution efficiency is low, and even the situation of failure of task execution occurs.
2. The parameter transfer efficiency is low, the existing parameter transfer mechanism generally depends on middleware such as shared storage or message queues, so that the cost of data transmission is increased, and the task execution efficiency is reduced.
3. The existing scheduling method generally lacks flexible scheduling strategies, and the execution sequence and the resource allocation of the tasks cannot be dynamically adjusted according to the actual execution conditions of the tasks, so that the resource utilization rate is low.
Disclosure of Invention
The embodiment of the invention provides a data flow-based running context parameter transmission and task instance dynamic scheduling method, which can solve the problems in the prior art.
In a first aspect of an embodiment of the present invention,
The method for dynamically scheduling the running context parameter and the task instance based on the data stream comprises the following steps:
Receiving a task scheduling request, wherein the task scheduling request comprises task identification information and data flow configuration information, constructing a data flow dependency graph based on the data flow configuration information, wherein the data flow dependency graph comprises a plurality of task nodes and data flow relationships among the task nodes, and acquiring corresponding task instance running context information for each task node in the data flow dependency graph, wherein the task instance running context information comprises task instance identification, task instance state information, task instance input parameter information and task instance output parameter information;
Based on the data flow direction relation between task nodes in the data flow dependency graph, analyzing the context information during the operation of the task instance, establishing a mapping relation between the input parameter information of the task instance and the output parameter information of the task instance, and generating a parameter transfer rule according to the mapping relation, wherein the parameter transfer rule defines a transfer mode of the output parameter information of an upstream task instance to the input parameter information of a downstream task instance;
and monitoring task instance state information of each task node in the data flow dependency graph, when detecting that the state information of an upstream task instance changes, transmitting output parameter information of the upstream task instance to input parameter information of a downstream task instance according to the parameter transmission rule, judging whether the downstream task instance meets the execution condition or not based on the integrity of the input parameter information of the downstream task instance, and when the execution condition is met, sending a task instance scheduling instruction to a task scheduling system to trigger dynamic scheduling execution of the downstream task instance.
Constructing a data flow dependency graph based on the data flow configuration information, wherein the data flow dependency graph comprises a plurality of task nodes and data flow relationships among the task nodes, and acquiring the corresponding task instance running context information for each task node in the data flow dependency graph comprises:
The data flow configuration information comprises task node information and data flow direction information, a node dependency relation matrix is generated based on the data flow configuration information, matrix elements in the node dependency relation matrix are used for representing data transfer relations among task nodes, the task nodes are subjected to topological ordering according to the node dependency relation matrix to obtain a task node sequence, and a data flow dependency graph is constructed based on the task node sequence;
Establishing a context information model of a task instance aiming at a task node in the data flow dependency graph, wherein the context information model of the task instance comprises task instance identification information, task instance running state information, task instance input parameter information and task instance output parameter information, generating a context acquisition function based on the context information model of the task instance, wherein the context acquisition function is used for acquiring context information of the task node under a specified time stamp, and calculating a minimum update time interval of the context information according to the context acquisition function;
And constructing a version vector of each task node in the data flow dependency graph, wherein the version vector comprises version information of related task nodes with data transfer relation with the task nodes, performing context information consistency check based on the version vector, determining that the context information consistency check passes when the version information of the related task nodes meets a preset time deviation requirement, updating a context information cache of the task nodes according to the minimum updating time interval, and rolling back the context information of the task nodes to a state corresponding to a latest valid check point when the task nodes are detected to be invalid.
Establishing a context information model of a task instance aiming at a task node in the data flow dependency graph, wherein the context information model of the task instance comprises task instance identification information, task instance running state information, task instance input parameter information and task instance output parameter information, generating a context acquisition function based on the context information model of the task instance, wherein the context acquisition function is used for acquiring context information of the task node under a specified timestamp, and calculating a minimum update time interval of the context information according to the context acquisition function comprises the following steps:
constructing a context information model of a task instance, wherein the context information model comprises instance identification information, task state information, input parameter information and output parameter information, constructing a state transition matrix aiming at the context information model, wherein the state transition matrix is used for representing the transition probability of the task state information, calculating the distribution parameter of state duration time based on the state transition matrix, and evaluating the stability index of the task state according to the distribution parameter;
constructing a context acquisition function based on the context information model, wherein the context acquisition function is used for acquiring instance identification information, task state information, input parameter information and output parameter information at a specified time point, analyzing the dependency relationship between the input parameter information and the output parameter information, establishing a parameter dependency relationship table, and calculating an integrity measurement value of the context information, wherein the integrity measurement value is the ratio of the number of effective input parameters to the total number of required input parameters;
Acquiring the parameter change rate of input parameter information and output parameter information, establishing an updating threshold function based on the parameter change rate, calculating a minimum updating time interval by the updating threshold function, and constructing a parameter time sequence correlation matrix, wherein the parameter time sequence correlation matrix is used for representing the time sequence correlation degree among parameters, establishing a parameter prediction model based on the parameter time sequence correlation matrix, and calculating a prediction precision evaluation value according to the parameter prediction model.
Based on the data flow direction relation between task nodes in the data flow dependency graph, analyzing the context information during the task instance running, establishing a mapping relation between task instance input parameter information and task instance output parameter information, and generating a parameter transfer rule according to the mapping relation comprises:
Constructing a data flow dependency graph, wherein the data flow dependency graph comprises a task node set, a data flow side set and a side weight matrix, the side weight matrix is used for representing the importance degree of the data flow side, and the data flow metric value between any two task nodes is calculated based on the side weight matrix and is related to the corresponding weight value in the side weight matrix and the data flow between the nodes;
Establishing a context incidence matrix between task nodes, wherein the context incidence matrix is used for describing the incidence degree between an output parameter set of a source task node and an input parameter set of a target task node, determining the correlation coefficient of any two task nodes based on the context incidence matrix, calculating the correlation coefficient based on the parameter covariance of the task nodes and the parameter standard deviation of the task nodes, and determining a parameter mapping function according to the correlation coefficient, wherein the parameter mapping function comprises a mapping weight coefficient and a parameter conversion function;
And constructing a parameter dependency relationship graph based on the parameter mapping function, calculating the dependency distance between task nodes in the parameter dependency relationship graph, calculating the dependency strength between parameters according to the dependency distance, wherein the dependency strength decays along with the increase of the dependency distance, analyzing the transfer characteristics between parameter sequences with the dependency relationship, generating a parameter transfer rule template according to the transfer characteristics, wherein the parameter transfer rule template comprises a source parameter set, a target parameter set, a triggering condition and a conversion action, and distributing priority to each parameter transfer rule based on the importance degree and the execution frequency of the rule.
Constructing a parameter dependency relationship graph based on the parameter mapping function, calculating the dependency distance between task nodes in the parameter dependency relationship graph, calculating the dependency strength between parameters according to the dependency distance, wherein the dependency strength decays along with the increase of the dependency distance, analyzing the transfer characteristic between parameter sequences with the dependency relationship, and generating a parameter transfer rule template according to the transfer characteristic comprises the following steps:
calculating the shortest path distance between task nodes based on the parameter dependency graph, wherein the shortest path distance is the minimum value of the sum of edge weights in paths connecting two task nodes, acquiring the type information of the task nodes and calculating the semantic distance between the task nodes, wherein the semantic distance is calculated based on a similarity function of the parameter types, and the shortest path distance and the semantic distance are combined through weight coefficients to obtain a comprehensive distance metric;
The method comprises the steps of calculating the dependence strength of parameters according to the comprehensive distance measurement value, wherein the dependence strength of the parameters adopts a product form of a basic strength function and a time sequence influence factor, the basic strength function decays exponentially along with the increase of the comprehensive distance measurement value, the time sequence influence factor is related to a time interval, a parameter transmission chain is built based on the dependence strength of the parameters, an integrity index and a reliability index of the parameter transmission chain are calculated, the integrity index is the minimum value of the dependence strength of adjacent parameters in the parameter transmission chain, the reliability index is calculated based on the error rate in the transmission process, and a parameter transmission rule template is generated based on the parameter transmission chain and comprises a source mode, a target mode, a conversion rule and constraint conditions.
Transmitting the output parameter information of the upstream task instance to the input parameter information of the downstream task instance according to the parameter transmission rule, judging whether the downstream task instance meets the execution condition based on the integrity of the input parameter information of the downstream task instance, and when the execution condition is met, transmitting a task instance scheduling instruction to a task scheduling system, wherein the triggering the dynamic scheduling execution of the downstream task instance comprises the following steps:
Acquiring an output parameter set of an upstream task instance, calculating accuracy and timeliness indexes of each parameter in the output parameter set, determining a parameter quality evaluation value based on the accuracy and the timeliness indexes, and converting the parameter quality evaluation value into a parameter availability index according to a preset weight coefficient;
Counting an input parameter set required by a downstream task instance, calculating a parameter integrity measurement value of an output parameter set of the upstream task instance and an input parameter set required by the downstream task instance, wherein the parameter integrity measurement value is a ratio of the number of actually received parameters to the number of parameters required by the task, and calculating a condition satisfaction degree based on the parameter integrity measurement value and the parameter availability index;
acquiring resource consumption information and task emergency degree information of a downstream task instance, calculating task basic priority based on the resource consumption information and the task emergency degree information, acquiring a load balancing factor of a current time point, wherein the load balancing factor is reduced along with the increase of the current load of a system, and multiplying the task basic priority by the load balancing factor to obtain task dynamic priority;
And carrying out matching degree evaluation on the available computing resources based on the task dynamic priority, wherein the matching degree evaluation comprises matching degree calculation of resource feature dimensions, generating a scheduling instruction comprising task identification, resource identification, priority information, execution constraint and parameter requirements, and carrying out resource availability and parameter completeness verification on the scheduling instruction.
In a second aspect of the embodiment of the present invention, a system for dynamically scheduling a task instance and a context transfer in a data flow based operation is provided, including:
A first unit, configured to receive a task scheduling request, where the task scheduling request includes task identification information and data flow configuration information, and construct a data flow dependency graph based on the data flow configuration information, where the data flow dependency graph includes a plurality of task nodes and a data flow relationship between task nodes, and obtain, for each task node in the data flow dependency graph, corresponding task instance runtime context information, where the task instance runtime context information includes task instance identifier, task instance state information, task instance input parameter information, and task instance output parameter information;
The second unit is used for analyzing the context information during the operation of the task instance based on the data flow relation between task nodes in the data flow dependency graph, establishing a mapping relation between the input parameter information of the task instance and the output parameter information of the task instance, and generating a parameter transfer rule according to the mapping relation, wherein the parameter transfer rule defines a transfer mode of the output parameter information of an upstream task instance to the input parameter information of a downstream task instance;
And the third unit is used for monitoring task instance state information of each task node in the data flow dependency graph, transmitting output parameter information of an upstream task instance to input parameter information of a downstream task instance according to the parameter transmission rule when detecting that the state information of the upstream task instance changes, judging whether the downstream task instance meets the execution condition or not based on the input parameter information integrity of the downstream task instance, and transmitting a task instance scheduling instruction to a task scheduling system when the execution condition is met to trigger dynamic scheduling execution of the downstream task instance.
Third aspect of embodiments of the invention
There is provided an electronic device including:
a processor;
a memory for storing processor-executable instructions;
Wherein the processor is configured to invoke the instructions stored in the memory to perform the method described previously.
In a fourth aspect of an embodiment of the present invention,
There is provided a computer readable storage medium having stored thereon computer program instructions which, when executed by a processor, implement the method as described above.
The beneficial effects of the application are as follows:
1. the dynamic scheduling of the task instance is realized, namely the execution of the downstream task is automatically triggered by monitoring the state change and the dependency relationship of the task instance, manual intervention is not needed, and the efficiency and the automation degree of task scheduling are improved.
2. And flexible parameter transmission, namely, establishing a parameter mapping relation and a transmission rule based on the data flow dependency graph and the running context information, realizing automatic transmission from the output parameters of the upstream task to the input parameters of the downstream task, avoiding complicated operation of manually configuring the parameters and reducing human errors.
3. The data flow processing efficiency is improved, namely, the downstream task is ensured to be executed under the condition of complete input parameters through a parameter transmission rule and a dynamic scheduling mechanism, invalid scheduling and waiting are avoided, and the overall efficiency of data flow processing is improved.
Drawings
FIG. 1 is a flow chart of a method for dynamic scheduling of context transfer and task instances during data flow based operation in accordance with an embodiment of the present invention;
FIG. 2 is a schematic diagram of a task context parameter definition based on a business process according to an embodiment of the present invention;
FIG. 3 is a diagram illustrating dynamic generation of sub-instances based on runtime context-based afferent parameter values in accordance with an embodiment of the present invention;
FIG. 4 is a schematic diagram illustrating an example of scheduling, allocating resources and executing resources by distributed scheduling according to an embodiment of the present invention;
FIG. 5 is a flow chart of an example of a dynamic scheduling task based on a data flow runtime context transfer in an embodiment of the present invention;
FIG. 6 is a schematic diagram of a dynamic scheduling system for data flow-based runtime context transfer and task instance in accordance with an embodiment of the present invention.
Detailed Description
For the purpose of making the objects, technical solutions and advantages of the embodiments of the present invention more apparent, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention, and it is apparent that the described embodiments are only some embodiments of the present invention, not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
The technical scheme of the invention is described in detail below by specific examples. The following embodiments may be combined with each other, and some embodiments may not be repeated for the same or similar concepts or processes.
Referring to fig. 1 to 5, a method for dynamically scheduling a context transfer and task instance based on a data stream in a runtime of an embodiment of the present invention includes:
S101, receiving a task scheduling request, wherein the task scheduling request comprises task identification information and data flow configuration information, constructing a data flow dependency graph based on the data flow configuration information, wherein the data flow dependency graph comprises a plurality of task nodes and data flow relationships among the task nodes, and acquiring corresponding task instance running context information for each task node in the data flow dependency graph, wherein the task instance running context information comprises task instance identification, task instance state information, task instance input parameter information and task instance output parameter information;
S102, analyzing the context information during the operation of the task instance based on the data flow relation between task nodes in the data flow dependency graph, establishing a mapping relation between the input parameter information of the task instance and the output parameter information of the task instance, and generating a parameter transfer rule according to the mapping relation, wherein the parameter transfer rule defines a transfer mode of the output parameter information of an upstream task instance to the input parameter information of a downstream task instance;
S103, task instance state information of each task node in the data flow dependency graph is monitored, when the state information of an upstream task instance is detected to change, output parameter information of the upstream task instance is transmitted to input parameter information of a downstream task instance according to the parameter transmission rule, whether the downstream task instance meets execution conditions is judged based on the integrity of the input parameter information of the downstream task instance, and when the execution conditions are met, a task instance scheduling instruction is sent to a task scheduling system to trigger dynamic scheduling execution of the downstream task instance.
In an optional implementation manner, a data flow dependency graph is constructed based on the data flow configuration information, the data flow dependency graph includes a plurality of task nodes and a data flow relation between the task nodes, and for each task node in the data flow dependency graph, acquiring the context information of the corresponding task instance during operation includes:
The data flow configuration information comprises task node information and data flow direction information, a node dependency relation matrix is generated based on the data flow configuration information, matrix elements in the node dependency relation matrix are used for representing data transfer relations among task nodes, the task nodes are subjected to topological ordering according to the node dependency relation matrix to obtain a task node sequence, and a data flow dependency graph is constructed based on the task node sequence;
Establishing a context information model of a task instance aiming at a task node in the data flow dependency graph, wherein the context information model of the task instance comprises task instance identification information, task instance running state information, task instance input parameter information and task instance output parameter information, generating a context acquisition function based on the context information model of the task instance, wherein the context acquisition function is used for acquiring context information of the task node under a specified time stamp, and calculating a minimum update time interval of the context information according to the context acquisition function;
And constructing a version vector of each task node in the data flow dependency graph, wherein the version vector comprises version information of related task nodes with data transfer relation with the task nodes, performing context information consistency check based on the version vector, determining that the context information consistency check passes when the version information of the related task nodes meets a preset time deviation requirement, updating a context information cache of the task nodes according to the minimum updating time interval, and rolling back the context information of the task nodes to a state corresponding to a latest valid check point when the task nodes are detected to be invalid.
The data flow dependency graph construction and context information management method is used for constructing a dependency graph of a data flow task and managing the runtime context information of a task instance so as to ensure data consistency and task reliability.
First, the data stream information is configured. The data stream configuration information is stored in JSON format and contains task node information and data stream information. The task node information includes a task ID, a task name, a task type, and the like. The data flow information describes the data transfer relationship between task nodes, e.g., the output of task a is the input of task B.
Then, a node dependency matrix is generated. And generating a node dependency matrix according to the data flow information in the data flow configuration information. The rows and columns of the matrix represent task nodes and the matrix elements represent data transfer relationships between the task nodes. If the output of task A is the input of task B, the element value of the corresponding position in the matrix is 1, otherwise it is 0.
Next, topology ordering is performed. And performing topological ordering on the task nodes according to the node dependency relation matrix to obtain a task node sequence. Topological ordering ensures that the order of task execution conforms to data dependencies. For example, in the above example, the task node sequence is task_a- > task_b- > task_c.
Subsequently, a dataflow dependency graph is constructed. And constructing a data flow dependency graph based on the task node sequence after topological sequencing. The data flow dependency graph shows the dependency relationship among task nodes in the form of a directed acyclic graph, wherein the nodes in the graph represent tasks, and the edges represent data flow directions.
And establishing a task instance context information model. And establishing a context information model of the task instance for each task node. The model contains task instance identification information (e.g., task instance ID), task instance running status information (e.g., running, success, failure), task instance input parameter information, and task instance output parameter information.
A context acquisition function is generated. A context acquisition function is generated based on the context information model of the task instance. The context acquisition function is used for acquiring the context information of the task node under the designated time stamp. A minimum update time interval of the context information is calculated. A minimum update time interval for the context information is calculated based on the context acquisition function. The time interval is used for controlling the frequency of updating the context information, so that performance waste caused by frequent updating is avoided. For example, the minimum update time interval may be determined according to the execution frequency of the task or the frequency of data change.
And constructing a version vector of the task node. A version vector is constructed for each task node. The version vector contains version information of the associated task node in data transfer relationship with the task node. The version information may be a time stamp or a version number.
And performing a context information consistency check. Context information consistency checking is performed based on the version vector. And when the version information of the related task node meets the preset time deviation requirement, determining that the context information consistency check passes. For example, if task B depends on task a and the difference in version timestamps of task a and task B does not exceed 1 minute, then the context information consistency check is considered to pass.
And updating the context information cache of the task node. And according to the minimum updating time interval, the context information cache of the task node is updated periodically.
And (5) performing failure processing on the task nodes. And when the failure of the task node is detected, the context information of the task node is rolled back to a state corresponding to the latest valid check point, so that the data consistency is ensured.
The application can realize the following steps:
And the data consistency is improved, namely the consistency check of the version vector and the context information ensures that the context information of the task node is kept consistent, and data collision and error are avoided. And the reliability of the task is enhanced, namely, the context information of the task node is rolled back to the latest effective check point through a task node failure processing mechanism, so that the reliable execution of the task is ensured. Optimizing resource utilization, namely controlling the update frequency of the context information through the minimum update time interval, avoiding resource waste caused by frequent update and improving the system efficiency.
In an alternative embodiment, for a task node in the data flow dependency graph, a context information model of a task instance is established, where the context information model of the task instance includes task instance identification information, task instance running state information, task instance input parameter information, and task instance output parameter information, and a context acquisition function is generated based on the context information model of the task instance, where the context acquisition function is used to acquire context information of the task node under a specified timestamp, and calculating a minimum update time interval of the context information according to the context acquisition function includes:
constructing a context information model of a task instance, wherein the context information model comprises instance identification information, task state information, input parameter information and output parameter information, constructing a state transition matrix aiming at the context information model, wherein the state transition matrix is used for representing the transition probability of the task state information, calculating the distribution parameter of state duration time based on the state transition matrix, and evaluating the stability index of the task state according to the distribution parameter;
constructing a context acquisition function based on the context information model, wherein the context acquisition function is used for acquiring instance identification information, task state information, input parameter information and output parameter information at a specified time point, analyzing the dependency relationship between the input parameter information and the output parameter information, establishing a parameter dependency relationship table, and calculating an integrity measurement value of the context information, wherein the integrity measurement value is the ratio of the number of effective input parameters to the total number of required input parameters;
Acquiring the parameter change rate of input parameter information and output parameter information, establishing an updating threshold function based on the parameter change rate, calculating a minimum updating time interval by the updating threshold function, and constructing a parameter time sequence correlation matrix, wherein the parameter time sequence correlation matrix is used for representing the time sequence correlation degree among parameters, establishing a parameter prediction model based on the parameter time sequence correlation matrix, and calculating a prediction precision evaluation value according to the parameter prediction model.
Aiming at task nodes in the data flow dependency graph, the embodiment provides a method for constructing a task instance context information model and calculating a minimum update time interval so as to improve task monitoring and management efficiency.
First, a context information model of the task instance is constructed. The model comprises four key information, namely instance identification information, task running state information, input parameter information and output parameter information. For example, for a task instance named "data cleaning", the identification information may be UUID, for example, "a1b2c3d4-e5f6-7890-1234-567890abcdef", the running status information may be "running", "success", "failure", etc., the input parameter information may be the data file name to be cleaned and the cleaning rule, and the output parameter information may be the cleaned data file name and the cleaning report.
Next, a state transition matrix is constructed for the context information model. The matrix is used to characterize the transition probabilities of the task state information. For example, the "in-flight" state may transition to a "success" or "failure" state, while the "success" state typically does not transition to other states. By counting a large amount of task instance state transition data, the state transition probability can be obtained, for example, the probability from "running" to "success" is 90%, and the probability from "failure" is 10%. The state transition probabilities for the "data cleansing" task instance are assumed to be 90% from "running" to "successful" and 10% to "failed", with the "successful" and "failed" states remaining unchanged.
Then, a distribution parameter of the state duration is calculated based on the state transition matrix. For example, the average duration, the longest duration, the shortest duration, etc. of each state may be calculated. These parameters may be used to evaluate the stability of the task state. For example, if the average duration of the "on-the-fly" state is long, it may indicate that the task is being performed inefficiently. Assuming that the average duration of the "on-the-fly" state for the "data cleansing" task instance is 10 minutes, the duration of the "success" and "failure" states can be considered to be infinitely long.
Then, a context acquisition function is constructed based on the context information model. The function is used for acquiring instance identification information, task state information, input parameter information and output parameter information at a specified time point. This information may be obtained, for example, by querying a database or calling an API interface.
Then, the dependency relationship between the input parameter information and the output parameter information is analyzed, and a parameter dependency relationship table is established. For example, the data file name after purging depends on the data file name to be purged and the purging rule. An integrity metric value of the context information is then calculated, the integrity metric value being defined as the ratio of the number of valid input parameters to the total number of required input parameters. For example, if the "data cleansing" task requires two input parameters, a data file name and cleansing rules, and only the data file name is actually provided, the integrity metric value is 50%.
And acquiring the parameter change rates of the input parameter information and the output parameter information. For example, the number of changes or the magnitude of the changes in each parameter over a period of time may be calculated.
An update threshold function is established based on the rate of change of the parameter. The function contains the adjustment coefficients and the base update time interval. For example, the adjustment coefficient may be set to 2 and the base update time interval to 1 minute. The update time interval will be shortened to 0.5 minutes when the rate of change of the parameter exceeds a certain threshold. Assuming that the input parameter "data file name" of the "data cleansing" task changes once per minute, the update time interval is 0.5 minutes.
Finally, a minimum update time interval is calculated. The time interval is used to determine when the context information needs to be updated. For example, if the minimum update time interval is 1 minute, it is necessary to call a context acquisition function to acquire the latest context information every 1 minute. And constructing a parameter time sequence correlation matrix, wherein the matrix is used for representing the time sequence correlation degree among the parameters. For example, a correlation coefficient of two parameters over time may be calculated. Then, a parameter prediction model is established based on the parameter timing correlation matrix. For example, a time series analysis method may be used to predict future values of the parameters. And finally, calculating a prediction accuracy evaluation value according to the parameter prediction model. For example, the prediction accuracy may be estimated using a mean square error. Assuming that the output parameter "amount of data after cleaning" of the "data cleaning" task is highly correlated with the input parameter "amount of data to be cleaned", a simple linear regression model can be built to predict the "amount of data after cleaning".
The application can realize the following steps:
The task monitoring efficiency is improved, namely task abnormality can be timely found and processed by acquiring the real-time context information of the task instance, and the task monitoring efficiency is improved. Optimizing resource scheduling, namely, by analyzing the distribution parameters of the task state duration, the stability and the resource demand of the task can be evaluated, so that the resource scheduling strategy is optimized. And the prediction accuracy is improved, namely, the future behavior of the task, such as the completion time or the resource consumption of the task, can be predicted by establishing a parameter prediction model, so that the prediction accuracy is improved.
In an alternative embodiment, based on a data flow direction relationship between task nodes in the data flow dependency graph, analyzing the context information during task instance operation, establishing a mapping relationship between task instance input parameter information and task instance output parameter information, and generating a parameter transfer rule according to the mapping relationship includes:
Constructing a data flow dependency graph, wherein the data flow dependency graph comprises a task node set, a data flow side set and a side weight matrix, the side weight matrix is used for representing the importance degree of the data flow side, and the data flow metric value between any two task nodes is calculated based on the side weight matrix and is related to the corresponding weight value in the side weight matrix and the data flow between the nodes;
Establishing a context incidence matrix between task nodes, wherein the context incidence matrix is used for describing the incidence degree between an output parameter set of a source task node and an input parameter set of a target task node, determining the correlation coefficient of any two task nodes based on the context incidence matrix, calculating the correlation coefficient based on the parameter covariance of the task nodes and the parameter standard deviation of the task nodes, and determining a parameter mapping function according to the correlation coefficient, wherein the parameter mapping function comprises a mapping weight coefficient and a parameter conversion function;
And constructing a parameter dependency relationship graph based on the parameter mapping function, calculating the dependency distance between task nodes in the parameter dependency relationship graph, calculating the dependency strength between parameters according to the dependency distance, wherein the dependency strength decays along with the increase of the dependency distance, analyzing the transfer characteristics between parameter sequences with the dependency relationship, generating a parameter transfer rule template according to the transfer characteristics, wherein the parameter transfer rule template comprises a source parameter set, a target parameter set, a triggering condition and a conversion action, and distributing priority to each parameter transfer rule based on the importance degree and the execution frequency of the rule.
In order to effectively manage and transfer parameter information in a complex workflow, the present embodiment provides a parameter transfer rule generation method based on data flow dependence and context information.
First, a data flow dependency graph is constructed. The graph contains a set of nodes representing individual tasks, and a set of directed edges representing the flow of data between tasks. Each edge is given a weight to indicate the importance of the data flow. For example, in one video processing flow, the weight of the side of the "video decoding" task directed to the "video encoding" task may be set to 0.9, which indicates that the decoding result is critical to the encoding process, and the weight of the side of the "subtitle" task directed to the "video encoding" task may be set to 0.5, which indicates that the influence of subtitle information on the encoding process is relatively small. By analyzing the weight of the edge and the data flow between the nodes, a data flow metric value between any two task nodes can be calculated, for example, if the "video decoding" task outputs 10MB data to the "video encoding" task, the data flow metric value between the two tasks is 0.9x10=9.
Next, a contextual relevance matrix between the task nodes is established. The matrix describes the degree of association between the output parameter set of the source task node and the input parameter set of the target task node. For example, the parameters output by the "video decoding" task include resolution, frame rate, and encoding format, while the input parameters for the "video encoding" task include resolution, frame rate, and code rate. By analyzing the relationship between these parameters, the correlation coefficient of the two task nodes can be determined. For example, resolution and frame rate are highly correlated, their correlation coefficients approach 1, while the correlation of coding format and code rate is lower, the coefficients approach 0. And determining a parameter mapping function according to the correlation coefficient. The function comprises a mapping weight coefficient and a parameter conversion function. For example, if the source task output resolution is 1920x1080 and the target task requires 720x480, the parameter transfer function may perform a corresponding scaling operation, and the mapping weight coefficient may be set according to the importance of the resolution.
Then, a parameter dependency graph is constructed based on the parameter mapping function. The dependency distance between task nodes in the graph is calculated, for example, if there are no other tasks between the "video decoding" task and the "video encoding" task, their dependency distance is 1, and if there is one "watermarking" task in the middle, the dependency distance is 2. And calculating the dependence intensity among the parameters according to the dependence distance, wherein the dependence intensity decays with the increase of the dependence distance. For example, the dependence intensity is 1 when the distance is 1, and 0.5 when the distance is 2. Analysis of transfer characteristics between parameter sequences having a dependency relationship, for example, resolution parameters need to be consistent throughout the video processing flow. And generating a parameter transfer rule template according to the transfer characteristic. The template contains a set of source parameters, a set of target parameters, trigger conditions, and a conversion action. For example, one rule template may be to pass the decoded resolution to the "video encoding" task and perform a corresponding scaling operation when the "video decoding" task is completed.
Finally, each parameter delivery rule is assigned a priority based on the importance level and execution frequency of the rule. For example, rules that guarantee consistency of video resolution are prioritized over subtitle-added rules. In the video processing flow, three parameters of resolution, frame rate and code rate are required to be transmitted, and priorities 1, 2 and 3 are respectively assigned according to the importance degree and the use frequency.
The application can realize the following steps:
And the parameter transmission efficiency is improved. By analyzing the dependency relationship between tasks and the association degree between parameters, the parameter transmission rule is automatically generated, and the complicated operation of manually configuring the parameters is avoided, so that the efficiency of parameter transmission is improved. Parameter delivery errors are reduced. By establishing the parameter dependency graph and the context correlation matrix, errors in the parameter transmission process, such as parameter type mismatch, parameter value deletion and the like, can be effectively identified and avoided, so that the accuracy of parameter transmission is improved. The flexibility of the workflow is enhanced. Through the parameter transfer rule template, the parameter transfer rule can be dynamically adjusted according to different task demands, so that the flexibility and adaptability of the workflow are enhanced, and for example, different coding parameters can be selected according to different video formats.
In an alternative embodiment, a parameter dependency graph is constructed based on the parameter mapping function, a dependency distance between task nodes in the parameter dependency graph is calculated, a dependency strength between parameters is calculated according to the dependency distance, the dependency strength decays along with the increase of the dependency distance, a transfer characteristic between parameter sequences with a dependency relationship is analyzed, and a parameter transfer rule template is generated according to the transfer characteristic, which comprises:
calculating the shortest path distance between task nodes based on the parameter dependency graph, wherein the shortest path distance is the minimum value of the sum of edge weights in paths connecting two task nodes, acquiring the type information of the task nodes and calculating the semantic distance between the task nodes, wherein the semantic distance is calculated based on a similarity function of the parameter types, and the shortest path distance and the semantic distance are combined through weight coefficients to obtain a comprehensive distance metric;
The method comprises the steps of calculating the dependence strength of parameters according to the comprehensive distance measurement value, wherein the dependence strength of the parameters adopts a product form of a basic strength function and a time sequence influence factor, the basic strength function decays exponentially along with the increase of the comprehensive distance measurement value, the time sequence influence factor is related to a time interval, a parameter transmission chain is built based on the dependence strength of the parameters, an integrity index and a reliability index of the parameter transmission chain are calculated, the integrity index is the minimum value of the dependence strength of adjacent parameters in the parameter transmission chain, the reliability index is calculated based on the error rate in the transmission process, and a parameter transmission rule template is generated based on the parameter transmission chain and comprises a source mode, a target mode, a conversion rule and constraint conditions.
The parameter dependence relation diagram construction and transmission rule template generation method is used for generating a parameter transmission rule template by analyzing dependence relations among parameters so as to improve the efficiency and accuracy of parameter transmission.
First, a parameter dependency graph is constructed. Each parameter is considered as a node in the graph, and if there is a dependency between two parameters, an edge is established between the two nodes. The weight of an edge represents the strength of the dependency relationship and may be determined by domain expert scoring or data driven methods. For example, if the parameters a and B together affect the result of a task, the edge weights may be set according to the extent to which they affect the result. Assuming that the degree of influence of the parameter a is 0.8 and the degree of influence of the parameter B is 0.6, the weight of the edge between them can be set to 0.8×0.6=0.48.
Next, the dependency distance between the task nodes is calculated. The dependency distance is used for measuring the strength of the dependency relationship between the parameters. First, the shortest path distance between task nodes, i.e. the minimum value of the sum of edge weights in the path connecting two task nodes, is calculated. For example, the shortest path from parameter a to parameter C passes through parameter B with an a-B edge weight of 0.48 and a B-C edge weight of 0.7, the shortest path distance of a-C is 0.48+0.7=1.18. Meanwhile, the type information of the task nodes, such as the data type, the value range and the like of the parameters, is acquired, and the semantic distance between the task nodes is calculated. The semantic distance is calculated based on a similarity function of the parameter types, e.g. if the data types of the two parameters are the same, their semantic distance is smaller. Assuming that the data type similarity of parameter a and parameter C is 0.9, their semantic distance may be set to 1-0.9=0.1. And then, combining the shortest path distance and the semantic distance through a weight coefficient to obtain a comprehensive distance measurement value. For example, if the shortest path distance is weighted to 0.6 and the semantic distance is weighted to 0.4, the combined distance metric value of a-C is 1.18×0.6+0.1×0.4=0.748.
Then, the dependency strength of the parameter is calculated from the integrated distance metric. The dependency strength of the parameters adopts the product form of a basic strength function and a time sequence influence factor. The base intensity function decays exponentially as the integrated distance measure increases, for example exp (-integrated distance measure) may be used as the base intensity function. The timing impact factor is related to the time interval, e.g. if the time interval of the two parameters is long, the dependency strength between them will decrease. Assuming that the time interval of the parameter a and the parameter C is 1 day, the timing influence factor may be set to 0.9. Assuming that the integrated distance metric value of a-C is 0.748, the dependency strength of a-C is exp (-0.748) 0.9=0.428.
Based on the dependency strength of the parameters, a parameter transfer chain is established. The parameter transfer chain indicates the order and dependency of parameter transfer. For example, if parameter a depends on parameter B, which depends on parameter C, a parameter transfer chain C- > B- > a may be established. And calculating an integrity index and a reliability index of the parameter transmission chain. The integrity index is the minimum value of the dependency strength of adjacent parameters in the parameter transfer chain. The reliability index is calculated based on the error rate during the transfer, e.g. the error rate of the transfer may be counted from historical data. Assuming that the dependency strength of C-B is 0.6 and the dependency strength of B-A is 0.428, the integrity index of the transfer chain is 0.428. Assuming that the error rate during transfer is 0.1, the reliability index of the transfer chain is 1-0.1=0.9.
Finally, a parameter delivery rule template is generated based on the parameter delivery chain. The parameter delivery rule template contains source patterns, target patterns, conversion rules, and constraints. The source mode represents an initial state of parameter delivery, the target mode represents a target state of parameter delivery, the transition rule describes how to transition the parameter from the source mode to the target mode, and the constraint limits the conditions of parameter delivery. For example, the range of the parameter C is [0, 1], the range of the parameter B is [1, 2], and a conversion rule of b=c+1 and a constraint of C > =0 can be set.
The application can realize the following steps:
And the parameter transmission efficiency is improved, namely unnecessary parameter transmission can be avoided by analyzing the dependency relationship among the parameters, so that the parameter transmission efficiency is improved. The parameter transmission accuracy is improved, namely the dependence strength of the parameters can be calculated more accurately by considering the type information and the time interval of the parameters, so that the parameter transmission accuracy is improved. The interpretive of the parameter transfer rule is enhanced, namely the parameter transfer rule template comprises a source mode, a target mode, a conversion rule and constraint conditions, and the logic of parameter transfer can be clearly described, so that the interpretive of the parameter transfer rule is enhanced.
In an alternative embodiment, the output parameter information of the upstream task instance is transferred to the input parameter information of the downstream task instance according to the parameter transfer rule, whether the downstream task instance meets the execution condition is judged based on the integrity of the input parameter information of the downstream task instance, and when the execution condition is met, a task instance scheduling instruction is sent to a task scheduling system, and the triggering of the dynamic scheduling execution of the downstream task instance includes:
Acquiring an output parameter set of an upstream task instance, calculating accuracy and timeliness indexes of each parameter in the output parameter set, determining a parameter quality evaluation value based on the accuracy and the timeliness indexes, and converting the parameter quality evaluation value into a parameter availability index according to a preset weight coefficient;
Counting an input parameter set required by a downstream task instance, calculating a parameter integrity measurement value of an output parameter set of the upstream task instance and an input parameter set required by the downstream task instance, wherein the parameter integrity measurement value is a ratio of the number of actually received parameters to the number of parameters required by the task, and calculating a condition satisfaction degree based on the parameter integrity measurement value and the parameter availability index;
acquiring resource consumption information and task emergency degree information of a downstream task instance, calculating task basic priority based on the resource consumption information and the task emergency degree information, acquiring a load balancing factor of a current time point, wherein the load balancing factor is reduced along with the increase of the current load of a system, and multiplying the task basic priority by the load balancing factor to obtain task dynamic priority;
And carrying out matching degree evaluation on the available computing resources based on the task dynamic priority, wherein the matching degree evaluation comprises matching degree calculation of resource feature dimensions, generating a scheduling instruction comprising task identification, resource identification, priority information, execution constraint and parameter requirements, and carrying out resource availability and parameter completeness verification on the scheduling instruction.
A task dynamic scheduling method based on parameter transfer and condition judgment is used for efficiently executing tasks in a distributed computing environment. The method is characterized in that whether the downstream task meets the execution condition is dynamically judged according to the output parameter information of the upstream task, and priority scheduling is carried out according to the system load condition, so that the resource utilization rate and the task execution efficiency are improved.
First, a set of output parameters for an upstream task instance is obtained. For example, the output parameters of an image recognition task may include the type of object identified, confidence, and image feature vector. Then, the accuracy and timeliness index of each output parameter is calculated separately. For example, confidence may be used as an accuracy indicator and image acquisition time may be used as a timeliness indicator. Assuming that the identified object type is "cat," the confidence is 95% and the image acquisition time is 1 minute ago.
Next, a parameter quality assessment value is determined based on the accuracy and timeliness index. For example, the confidence and timeliness indicators may be weighted and summed to obtain the parameter quality assessment value according to a preset rule. Assuming that the confidence weight is 0.8 and the timeliness weight is 0.2, the parameter quality evaluation value is 0.8×95% +0.2 (100% -1/60) ≡76.3%. Then, the parameter quality evaluation value is converted into a parameter availability index according to a preset weight coefficient. Assuming a weight coefficient of 0.9, the parameter availability index is 0.9×76.3+.68.7%.
Thereafter, the set of input parameters required for the downstream task instance is counted. For example, a target tracking task requires input of object type, image feature vector, and target initial position. Parameter integrity metric values are then calculated for the output parameter set of the upstream task instance and the input parameter set required for the downstream task instance. For example, if the upstream task provides object type and image feature vectors, while the target initial position is provided by other tasks, the parameter integrity metric value is 2/3≡66.7%. Then, a condition satisfaction is calculated based on the parameter integrity metric value and the parameter availability index. For example, the two may be weighted and summed to obtain the condition satisfaction. Assuming that the parameter integrity metric weight is 0.7 and the parameter availability index weight is 0.3, the condition satisfaction is 0.7×66.7% +0.3×68.7% ≡67.2%.
If the condition satisfaction reaches a preset threshold, for example 70%, the next step is performed. And acquiring resource consumption information and task emergency degree information of the downstream task instance. For example, the resource consumption information may include the number of CPU cores required, memory size, and GPU type, while the task urgency information may be evaluated based on the deadline and importance of the task. Assume that the target trace task requires 2 CPU cores, 4GB of memory, and one NVIDIA TESLA T GPU, with a task urgency of "high".
Task base priorities are calculated based on the resource consumption information and the task urgency information. For example, the task base priority may be obtained by weighted summing the resource consumption and the urgency according to a preset rule. Assuming that the resource consumption weight is 0.6 and the emergency degree weight is 0.4, the task base priority is 0.6 (2/8+4/16+1/1) +0.4×1=0.85, wherein the total resources of the CPU, the memory and the GPU are 8 cores, 16GB and 1, respectively.
And acquiring a load balancing factor of the current time point. The load balancing factor decreases as the current load of the system increases. Assuming that the current system load is high, the load balancing factor is 0.5. And multiplying the task basic priority by a load balancing factor to obtain the task dynamic priority. The task dynamic priority is 0.85 x 0.5=0.425.
And evaluating the matching degree of the available computing resources based on the task dynamic priority. The matching degree evaluation comprises matching degree calculation of the feature dimension of the resource. For example, matching is performed according to the type of GPU required for the task and the type of available GPUs. Scheduling instructions are generated that contain task identifications, resource identifications, priority information, execution constraints, and parameter requirements. For example, the scheduling instruction may contain the ID of the target trace task, the ID of the assigned GPU, the priority 0.425, the execution time limit of 1 hour, and the required input parameter information. And verifying the resource availability and parameter completeness of the scheduling instruction. And after the distributed resources are available and the parameters are complete, sending a task instance scheduling instruction to a task scheduling system, and triggering the dynamic scheduling execution of a downstream task instance.
The application can realize the following steps:
The resource utilization rate is improved, namely the calculation resource can be effectively utilized by dynamically scheduling according to the task demands and the system load, and the resource waste and the bottleneck are avoided. The task execution efficiency is improved, namely, urgent and important tasks can be executed preferentially through a priority scheduling and parameter prejudging mechanism, and the task waiting time is reduced, so that the overall task execution efficiency is improved. And the stability of the system is enhanced, namely, the overload of the system can be avoided and the stable operation of the system is ensured through a load balancing mechanism.
Fig. 6 is a schematic structural diagram of a dynamic scheduling system for context transfer and task instance based on data flow in the embodiment of the present invention, as shown in fig. 6, the system includes:
A first unit, configured to receive a task scheduling request, where the task scheduling request includes task identification information and data flow configuration information, and construct a data flow dependency graph based on the data flow configuration information, where the data flow dependency graph includes a plurality of task nodes and a data flow relationship between task nodes, and obtain, for each task node in the data flow dependency graph, corresponding task instance runtime context information, where the task instance runtime context information includes task instance identifier, task instance state information, task instance input parameter information, and task instance output parameter information;
The second unit is used for analyzing the context information during the operation of the task instance based on the data flow relation between task nodes in the data flow dependency graph, establishing a mapping relation between the input parameter information of the task instance and the output parameter information of the task instance, and generating a parameter transfer rule according to the mapping relation, wherein the parameter transfer rule defines a transfer mode of the output parameter information of an upstream task instance to the input parameter information of a downstream task instance;
And the third unit is used for monitoring task instance state information of each task node in the data flow dependency graph, transmitting output parameter information of an upstream task instance to input parameter information of a downstream task instance according to the parameter transmission rule when detecting that the state information of the upstream task instance changes, judging whether the downstream task instance meets the execution condition or not based on the input parameter information integrity of the downstream task instance, and transmitting a task instance scheduling instruction to a task scheduling system when the execution condition is met to trigger dynamic scheduling execution of the downstream task instance.
In a third aspect of an embodiment of the present invention,
There is provided an electronic device including:
a processor;
a memory for storing processor-executable instructions;
Wherein the processor is configured to invoke the instructions stored in the memory to perform the method described previously.
In a fourth aspect of an embodiment of the present invention,
There is provided a computer readable storage medium having stored thereon computer program instructions which, when executed by a processor, implement the method as described above.
The present invention may be a method, apparatus, system, and/or computer program product. The computer program product may include a computer readable storage medium having computer readable program instructions embodied thereon for performing various aspects of the present invention.
It should be noted that the above embodiments are merely for illustrating the technical solution of the present invention and not for limiting the same, and although the present invention has been described in detail with reference to the above embodiments, it should be understood by those skilled in the art that the technical solution described in the above embodiments may be modified or some or all of the technical features may be equivalently replaced, and these modifications or substitutions do not make the essence of the corresponding technical solution deviate from the scope of the technical solution of the embodiments of the present invention.

Claims (8)

1.基于数据流的运行时上下文传参及任务实例动态调度方法,其特征在于,包括:1. A method for dynamically scheduling task instances and contextual parameters at runtime based on data flow, characterized by comprising: 接收任务调度请求,所述任务调度请求包含任务标识信息和数据流配置信息,基于所述数据流配置信息构建数据流依赖图,所述数据流依赖图包含多个任务节点及任务节点之间的数据流向关系,针对所述数据流依赖图中的每个任务节点,获取对应的任务实例运行时上下文信息,所述任务实例运行时上下文信息包含任务实例标识、任务实例状态信息、任务实例输入参数信息和任务实例输出参数信息;Receive a task scheduling request, the task scheduling request includes task identification information and data flow configuration information, build a data flow dependency graph based on the data flow configuration information, the data flow dependency graph includes multiple task nodes and data flow direction relationships between task nodes, and for each task node in the data flow dependency graph, obtain corresponding task instance runtime context information, the task instance runtime context information includes task instance identification, task instance status information, task instance input parameter information, and task instance output parameter information; 基于所述数据流依赖图中任务节点之间的数据流向关系,分析所述任务实例运行时上下文信息,建立任务实例输入参数信息与任务实例输出参数信息之间的映射关系,根据所述映射关系生成参数传递规则,所述参数传递规则定义了上游任务实例的输出参数信息向下游任务实例的输入参数信息的传递方式;Based on the data flow relationship between the task nodes in the data flow dependency graph, the runtime context information of the task instance is analyzed, a mapping relationship between the input parameter information of the task instance and the output parameter information of the task instance is established, and a parameter transfer rule is generated according to the mapping relationship, wherein the parameter transfer rule defines the transfer method of the output parameter information of the upstream task instance to the input parameter information of the downstream task instance; 监控所述数据流依赖图中各任务节点的任务实例状态信息,当检测到上游任务实例的状态信息发生变化时,根据所述参数传递规则将上游任务实例的输出参数信息传递给下游任务实例的输入参数信息,基于所述下游任务实例的输入参数信息完整性判断下游任务实例是否满足执行条件,当满足执行条件时,向任务调度系统发送任务实例调度指令,触发下游任务实例的动态调度执行;Monitor the task instance status information of each task node in the data flow dependency graph, and when it is detected that the status information of the upstream task instance changes, pass the output parameter information of the upstream task instance to the input parameter information of the downstream task instance according to the parameter transfer rule, and judge whether the downstream task instance meets the execution condition based on the integrity of the input parameter information of the downstream task instance. When the execution condition is met, send a task instance scheduling instruction to the task scheduling system to trigger the dynamic scheduling execution of the downstream task instance; 基于所述数据流配置信息构建数据流依赖图,所述数据流依赖图包含多个任务节点及任务节点之间的数据流向关系,针对所述数据流依赖图中的每个任务节点,获取对应的任务实例运行时上下文信息包括:A data flow dependency graph is constructed based on the data flow configuration information, wherein the data flow dependency graph includes multiple task nodes and data flow relationships between the task nodes. For each task node in the data flow dependency graph, obtaining the corresponding task instance runtime context information includes: 所述数据流配置信息包含任务节点信息和数据流向信息,基于所述数据流配置信息生成节点依赖关系矩阵,所述节点依赖关系矩阵中的矩阵元素用于表征任务节点之间的数据传递关系,根据所述节点依赖关系矩阵对所述任务节点进行拓扑排序得到任务节点序列,基于所述任务节点序列构建数据流依赖图;The data flow configuration information includes task node information and data flow direction information, a node dependency matrix is generated based on the data flow configuration information, the matrix elements in the node dependency matrix are used to characterize the data transfer relationship between task nodes, the task nodes are topologically sorted according to the node dependency matrix to obtain a task node sequence, and a data flow dependency graph is constructed based on the task node sequence; 针对所述数据流依赖图中的任务节点,建立任务实例的上下文信息模型,所述任务实例的上下文信息模型包含任务实例标识信息、任务实例运行状态信息、任务实例输入参数信息和任务实例输出参数信息,基于所述任务实例的上下文信息模型生成上下文获取函数,所述上下文获取函数用于获取任务节点在指定时间戳下的上下文信息,根据所述上下文获取函数计算上下文信息的最小更新时间间隔;For the task nodes in the data flow dependency graph, a context information model of the task instance is established, the context information model of the task instance includes task instance identification information, task instance running status information, task instance input parameter information and task instance output parameter information, a context acquisition function is generated based on the context information model of the task instance, the context acquisition function is used to obtain the context information of the task node at a specified timestamp, and the minimum update time interval of the context information is calculated according to the context acquisition function; 针对所述数据流依赖图中的每个任务节点,构建任务节点的版本向量,所述版本向量包含与所述任务节点存在数据传递关系的相关任务节点的版本信息,基于所述版本向量进行上下文信息一致性检查,当所述相关任务节点的版本信息满足预设的时间偏差要求时确定上下文信息一致性检查通过,根据所述最小更新时间间隔更新任务节点的上下文信息缓存,并在检测到任务节点失效时,将所述任务节点的上下文信息回滚至最近有效检查点对应的状态。For each task node in the data flow dependency graph, a version vector of the task node is constructed, wherein the version vector includes version information of related task nodes that have a data transmission relationship with the task node, and a context information consistency check is performed based on the version vector. When the version information of the related task node meets a preset time deviation requirement, it is determined that the context information consistency check has passed, and the context information cache of the task node is updated according to the minimum update time interval. When a task node failure is detected, the context information of the task node is rolled back to a state corresponding to the most recent valid checkpoint. 2.根据权利要求1所述的方法,其特征在于,针对所述数据流依赖图中的任务节点,建立任务实例的上下文信息模型,所述任务实例的上下文信息模型包含任务实例标识信息、任务实例运行状态信息、任务实例输入参数信息和任务实例输出参数信息,基于所述任务实例的上下文信息模型生成上下文获取函数,所述上下文获取函数用于获取任务节点在指定时间戳下的上下文信息,根据所述上下文获取函数计算上下文信息的最小更新时间间隔包括:2. The method according to claim 1 is characterized in that, for the task node in the data flow dependency graph, a context information model of the task instance is established, the context information model of the task instance includes task instance identification information, task instance running status information, task instance input parameter information and task instance output parameter information, a context acquisition function is generated based on the context information model of the task instance, the context acquisition function is used to obtain the context information of the task node at a specified timestamp, and the minimum update time interval of the context information calculated according to the context acquisition function includes: 构建任务实例的上下文信息模型,所述上下文信息模型包含实例标识信息、任务状态信息、输入参数信息和输出参数信息,针对所述上下文信息模型构建状态迁移矩阵,所述状态迁移矩阵用于表征任务状态信息的迁移概率,基于所述状态迁移矩阵计算状态持续时间的分布参数,根据所述分布参数评估任务状态的稳定性指标;Constructing a context information model of a task instance, wherein the context information model includes instance identification information, task state information, input parameter information, and output parameter information; constructing a state transition matrix for the context information model, wherein the state transition matrix is used to characterize the transition probability of task state information; calculating a distribution parameter of a state duration based on the state transition matrix; and evaluating a stability index of the task state according to the distribution parameter; 基于所述上下文信息模型构建上下文获取函数,所述上下文获取函数用于获取指定时间点下的实例标识信息、任务状态信息、输入参数信息和输出参数信息,分析所述输入参数信息和所述输出参数信息之间的依赖关系,建立参数依赖关系表,计算上下文信息的完整性度量值,所述完整性度量值为有效输入参数数量与所需输入参数总数的比值;Constructing a context acquisition function based on the context information model, wherein the context acquisition function is used to obtain instance identification information, task status information, input parameter information, and output parameter information at a specified time point, analyze the dependency relationship between the input parameter information and the output parameter information, establish a parameter dependency relationship table, and calculate an integrity metric value of the context information, wherein the integrity metric value is a ratio of the number of valid input parameters to the total number of required input parameters; 获取输入参数信息和输出参数信息的参数变化速率,基于所述参数变化速率建立更新阈值函数,所述更新阈值函数包含调节系数和基础更新时间间隔,计算最小更新时间间隔,构建参数时序相关性矩阵,所述参数时序相关性矩阵用于表征参数之间的时序关联程度,基于所述参数时序相关性矩阵建立参数预测模型,根据所述参数预测模型计算预测精度评估值。Obtain the parameter change rate of the input parameter information and the output parameter information, establish an update threshold function based on the parameter change rate, the update threshold function includes an adjustment coefficient and a basic update time interval, calculate the minimum update time interval, construct a parameter timing correlation matrix, the parameter timing correlation matrix is used to characterize the degree of timing correlation between parameters, establish a parameter prediction model based on the parameter timing correlation matrix, and calculate a prediction accuracy evaluation value according to the parameter prediction model. 3.根据权利要求1所述的方法,其特征在于,基于所述数据流依赖图中任务节点之间的数据流向关系,分析所述任务实例运行时上下文信息,建立任务实例输入参数信息与任务实例输出参数信息之间的映射关系,根据所述映射关系生成参数传递规则包括:3. The method according to claim 1 is characterized in that, based on the data flow relationship between the task nodes in the data flow dependency graph, the runtime context information of the task instance is analyzed, a mapping relationship between the input parameter information of the task instance and the output parameter information of the task instance is established, and the parameter transfer rule is generated according to the mapping relationship, comprising: 构建数据流依赖图,所述数据流依赖图包含任务节点集合、数据流向边集合和边权重矩阵,所述边权重矩阵用于表征数据流向边的重要程度,基于所述边权重矩阵计算任意两个任务节点之间的数据流向度量值,所述数据流向度量值与所述边权重矩阵中对应的权重值及节点间数据流量相关;Construct a data flow dependency graph, wherein the data flow dependency graph includes a task node set, a data flow edge set and an edge weight matrix, wherein the edge weight matrix is used to characterize the importance of the data flow edge, and a data flow metric value between any two task nodes is calculated based on the edge weight matrix, wherein the data flow metric value is related to the corresponding weight value in the edge weight matrix and the data flow between the nodes; 建立任务节点之间的上下文关联矩阵,所述上下文关联矩阵用于描述源任务节点的输出参数集合与目标任务节点的输入参数集合之间的关联程度,基于所述上下文关联矩阵确定任意两个任务节点的相关性系数,所述相关性系数基于任务节点的参数协方差与任务节点的参数标准差计算得到,根据所述相关性系数确定参数映射函数,所述参数映射函数包含映射权重系数和参数转换函数;Establishing a context association matrix between task nodes, the context association matrix is used to describe the degree of association between the output parameter set of the source task node and the input parameter set of the target task node, determining the correlation coefficient of any two task nodes based on the context association matrix, the correlation coefficient is calculated based on the parameter covariance of the task nodes and the parameter standard deviation of the task nodes, and determining a parameter mapping function based on the correlation coefficient, the parameter mapping function includes a mapping weight coefficient and a parameter conversion function; 基于所述参数映射函数构建参数依赖关系图,计算所述参数依赖关系图中任务节点之间的依赖距离,根据所述依赖距离计算参数间的依赖强度,所述依赖强度随依赖距离增大而衰减,分析具有依赖关系的参数序列间的传递特性,根据所述传递特性生成参数传递规则模板,所述参数传递规则模板包含源参数集合、目标参数集合、触发条件和转换动作,基于规则的重要程度和执行频率为每条参数传递规则分配优先级。A parameter dependency graph is constructed based on the parameter mapping function, the dependency distance between task nodes in the parameter dependency graph is calculated, the dependency strength between parameters is calculated based on the dependency distance, the dependency strength decays as the dependency distance increases, the transfer characteristics between parameter sequences with dependencies are analyzed, a parameter transfer rule template is generated based on the transfer characteristics, the parameter transfer rule template includes a source parameter set, a target parameter set, a trigger condition and a conversion action, and a priority is assigned to each parameter transfer rule based on the importance and execution frequency of the rule. 4.根据权利要求3所述的方法,其特征在于,基于所述参数映射函数构建参数依赖关系图,计算所述参数依赖关系图中任务节点之间的依赖距离,根据所述依赖距离计算参数间的依赖强度,所述依赖强度随依赖距离增大而衰减,分析具有依赖关系的参数序列间的传递特性,根据所述传递特性生成参数传递规则模板包括:4. The method according to claim 3 is characterized in that, based on the parameter mapping function, a parameter dependency graph is constructed, the dependency distance between task nodes in the parameter dependency graph is calculated, the dependency strength between parameters is calculated according to the dependency distance, the dependency strength decays as the dependency distance increases, the transfer characteristics between parameter sequences with dependency relationships are analyzed, and the parameter transfer rule template is generated according to the transfer characteristics, comprising: 基于所述参数依赖图计算任务节点之间的最短路径距离,所述最短路径距离为连接两个任务节点的路径中边权重之和的最小值,获取任务节点的类型信息并计算任务节点之间的语义距离,所述语义距离基于参数类型的相似度函数计算得到,将所述最短路径距离和所述语义距离通过权重系数组合得到综合距离度量值;Based on the parameter dependency graph, the shortest path distance between task nodes is calculated, the shortest path distance is the minimum value of the sum of edge weights in the path connecting two task nodes, the type information of the task nodes is obtained and the semantic distance between the task nodes is calculated based on the similarity function of the parameter type, and the shortest path distance and the semantic distance are combined by a weight coefficient to obtain a comprehensive distance measurement value; 根据所述综合距离度量值计算参数的依赖强度,所述参数的依赖强度采用基础强度函数与时序影响因子的乘积形式,其中所述基础强度函数随综合距离度量值增大呈指数衰减,所述时序影响因子与时间间隔相关,基于所述参数的依赖强度建立参数传递链,计算所述参数传递链的完整性指标和可靠性指标,所述完整性指标为参数传递链中相邻参数的依赖强度的最小值,所述可靠性指标基于传递过程中的错误率计算;基于所述参数传递链生成参数传递规则模板,所述参数传递规则模板包含源模式、目标模式、转换规则和约束条件。The dependency strength of the parameters is calculated according to the comprehensive distance metric value, and the dependency strength of the parameters is in the form of the product of a basic strength function and a timing influence factor, wherein the basic strength function decays exponentially with the increase of the comprehensive distance metric value, and the timing influence factor is related to the time interval. A parameter transfer chain is established based on the dependency strength of the parameters, and an integrity index and a reliability index of the parameter transfer chain are calculated, wherein the integrity index is the minimum value of the dependency strength of adjacent parameters in the parameter transfer chain, and the reliability index is calculated based on the error rate during the transfer process; a parameter transfer rule template is generated based on the parameter transfer chain, and the parameter transfer rule template includes a source mode, a target mode, a conversion rule, and a constraint condition. 5.根据权利要求1所述的方法,其特征在于,根据所述参数传递规则将上游任务实例的输出参数信息传递给下游任务实例的输入参数信息,基于所述下游任务实例的输入参数信息完整性判断下游任务实例是否满足执行条件,当满足执行条件时,向任务调度系统发送任务实例调度指令,触发下游任务实例的动态调度执行包括:5. The method according to claim 1 is characterized in that the output parameter information of the upstream task instance is passed to the input parameter information of the downstream task instance according to the parameter transfer rule, and whether the downstream task instance meets the execution condition is judged based on the integrity of the input parameter information of the downstream task instance. When the execution condition is met, a task instance scheduling instruction is sent to the task scheduling system, and the dynamic scheduling execution of the downstream task instance is triggered, which includes: 获取上游任务实例的输出参数集合,计算所述输出参数集合中各参数的准确度和时效性指标,基于所述准确度和所述时效性指标确定参数质量评估值,根据预设权重系数将所述参数质量评估值转换为参数可用性指标;Obtain an output parameter set of an upstream task instance, calculate the accuracy and timeliness index of each parameter in the output parameter set, determine a parameter quality evaluation value based on the accuracy and timeliness index, and convert the parameter quality evaluation value into a parameter availability index according to a preset weight coefficient; 统计下游任务实例所需的输入参数集合,计算所述上游任务实例的输出参数集合与所述下游任务实例所需的输入参数集合的参数完整性度量值,所述参数完整性度量值为实际接收的参数数量与任务所需参数数量的比值,基于所述参数完整性度量值和所述参数可用性指标计算条件满足度;Counting the input parameter set required by the downstream task instance, calculating the parameter integrity metric value of the output parameter set of the upstream task instance and the input parameter set required by the downstream task instance, the parameter integrity metric value being the ratio of the number of parameters actually received to the number of parameters required for the task, and calculating the condition satisfaction based on the parameter integrity metric value and the parameter availability index; 获取下游任务实例的资源消耗信息和任务紧急程度信息,基于所述资源消耗信息和所述任务紧急程度信息计算任务基础优先级,获取当前时间点的负载均衡因子,所述负载均衡因子随系统当前负载增大而减小,将所述任务基础优先级与所述负载均衡因子相乘得到任务动态优先级;Obtain resource consumption information and task urgency information of downstream task instances, calculate the basic task priority based on the resource consumption information and the task urgency information, obtain the load balancing factor at the current time point, the load balancing factor decreases as the current load of the system increases, and multiply the basic task priority by the load balancing factor to obtain the dynamic task priority; 基于所述任务动态优先级对可用计算资源进行匹配度评估,所述匹配度评估包含资源特征维度的匹配程度计算,生成包含任务标识、资源标识、优先级信息、执行约束和参数需求的调度指令,对所述调度指令进行资源可用性和参数完备性验证。Based on the dynamic priority of the task, a matching degree evaluation is performed on the available computing resources, and the matching degree evaluation includes a matching degree calculation of the resource feature dimension. A scheduling instruction including a task identifier, a resource identifier, priority information, execution constraints and parameter requirements is generated, and the scheduling instruction is verified for resource availability and parameter completeness. 6.基于数据流的运行时上下文传参及任务实例动态调度系统,用于实现如权利要求1-5中任一项所述的方法,其特征在于,包括:6. A runtime context parameter transmission and task instance dynamic scheduling system based on data flow, used to implement the method as claimed in any one of claims 1 to 5, characterized in that it includes: 第一单元,用于接收任务调度请求,所述任务调度请求包含任务标识信息和数据流配置信息,基于所述数据流配置信息构建数据流依赖图,所述数据流依赖图包含多个任务节点及任务节点之间的数据流向关系,针对所述数据流依赖图中的每个任务节点,获取对应的任务实例运行时上下文信息,所述任务实例运行时上下文信息包含任务实例标识、任务实例状态信息、任务实例输入参数信息和任务实例输出参数信息;A first unit is used to receive a task scheduling request, wherein the task scheduling request includes task identification information and data flow configuration information, construct a data flow dependency graph based on the data flow configuration information, wherein the data flow dependency graph includes multiple task nodes and data flow direction relationships between task nodes, and for each task node in the data flow dependency graph, obtain corresponding task instance runtime context information, wherein the task instance runtime context information includes task instance identification, task instance status information, task instance input parameter information, and task instance output parameter information; 第二单元,用于基于所述数据流依赖图中任务节点之间的数据流向关系,分析所述任务实例运行时上下文信息,建立任务实例输入参数信息与任务实例输出参数信息之间的映射关系,根据所述映射关系生成参数传递规则,所述参数传递规则定义了上游任务实例的输出参数信息向下游任务实例的输入参数信息的传递方式;The second unit is used to analyze the runtime context information of the task instance based on the data flow relationship between the task nodes in the data flow dependency graph, establish a mapping relationship between the input parameter information of the task instance and the output parameter information of the task instance, and generate a parameter transfer rule according to the mapping relationship, wherein the parameter transfer rule defines the transfer method of the output parameter information of the upstream task instance to the input parameter information of the downstream task instance; 第三单元,用于监控所述数据流依赖图中各任务节点的任务实例状态信息,当检测到上游任务实例的状态信息发生变化时,根据所述参数传递规则将上游任务实例的输出参数信息传递给下游任务实例的输入参数信息,基于所述下游任务实例的输入参数信息完整性判断下游任务实例是否满足执行条件,当满足执行条件时,向任务调度系统发送任务实例调度指令,触发下游任务实例的动态调度执行。The third unit is used to monitor the task instance status information of each task node in the data flow dependency graph. When it is detected that the status information of the upstream task instance changes, the output parameter information of the upstream task instance is passed to the input parameter information of the downstream task instance according to the parameter passing rule. Based on the integrity of the input parameter information of the downstream task instance, it is judged whether the downstream task instance meets the execution condition. When the execution condition is met, the task instance scheduling instruction is sent to the task scheduling system to trigger the dynamic scheduling execution of the downstream task instance. 7.一种电子设备,其特征在于,包括:7. An electronic device, comprising: 处理器;processor; 用于存储处理器可执行指令的存储器;a memory for storing processor-executable instructions; 其中,所述处理器被配置为调用所述存储器存储的指令,以执行权利要求1至5中任意一项所述的方法。The processor is configured to call the instructions stored in the memory to execute the method described in any one of claims 1 to 5. 8.一种计算机可读存储介质,其上存储有计算机程序指令,其特征在于,所述计算机程序指令被处理器执行时实现权利要求1至5中任意一项所述的方法。8. A computer-readable storage medium having computer program instructions stored thereon, wherein the computer program instructions, when executed by a processor, implement the method according to any one of claims 1 to 5.
CN202510140290.8A 2025-02-08 2025-02-08 Data flow-based running context parameter transmission and task instance dynamic scheduling method Active CN119578850B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202510140290.8A CN119578850B (en) 2025-02-08 2025-02-08 Data flow-based running context parameter transmission and task instance dynamic scheduling method

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202510140290.8A CN119578850B (en) 2025-02-08 2025-02-08 Data flow-based running context parameter transmission and task instance dynamic scheduling method

Publications (2)

Publication Number Publication Date
CN119578850A CN119578850A (en) 2025-03-07
CN119578850B true CN119578850B (en) 2025-04-08

Family

ID=94813037

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202510140290.8A Active CN119578850B (en) 2025-02-08 2025-02-08 Data flow-based running context parameter transmission and task instance dynamic scheduling method

Country Status (1)

Country Link
CN (1) CN119578850B (en)

Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN119248458A (en) * 2024-12-04 2025-01-03 厦门两万里文化传媒有限公司 AI workflow automation management method based on creative scenarios
CN119356879A (en) * 2024-12-23 2025-01-24 廊坊市讯云数据科技有限公司 A distributed real-time data stream processing system and method

Family Cites Families (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20230123322A1 (en) * 2021-04-16 2023-04-20 Strong Force Vcn Portfolio 2019, Llc Predictive Model Data Stream Prioritization

Patent Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN119248458A (en) * 2024-12-04 2025-01-03 厦门两万里文化传媒有限公司 AI workflow automation management method based on creative scenarios
CN119356879A (en) * 2024-12-23 2025-01-24 廊坊市讯云数据科技有限公司 A distributed real-time data stream processing system and method

Also Published As

Publication number Publication date
CN119578850A (en) 2025-03-07

Similar Documents

Publication Publication Date Title
US9235440B2 (en) Managing job execution
US20220027744A1 (en) Resource data modeling, forecasting and simulation
Zhang et al. Combined fault tolerance and scheduling techniques for workflow applications on computational grids
CN118939438A (en) Intelligent scheduling method and system for heterogeneous equipment system
CN119228319A (en) Visual project management method and system based on task tracking
US8832839B2 (en) Assessing system performance impact of security attacks
EP2472397A1 (en) Load distribution scheduling method in data processing system
US8180716B2 (en) Method and device for forecasting computational needs of an application
CN115202847A (en) Task scheduling method and device
CN112163154A (en) Data processing method, device, equipment and storage medium
CN112948092A (en) Batch job scheduling method and device, electronic equipment and storage medium
CN117742795A (en) Instruction scheduling method and device based on time-dependent priority, computer equipment and medium
CN119576590A (en) An AI-based intelligent scheduling system and method for computing resources
CN120276831A (en) Log analysis task scheduling system based on neural network
CN119578850B (en) Data flow-based running context parameter transmission and task instance dynamic scheduling method
CN120178818A (en) A multi-factory collaborative production management method, device, equipment and storage medium
CN118034887A (en) Big data platform task management method and system
CN117149383A (en) Data processing method, device, terminal equipment and storage medium
US20160004982A1 (en) Method and system for estimating the progress and completion of a project based on a bayesian network
CN112860763B (en) Real-time streaming data processing method and device, computer equipment and storage medium
CN116048709A (en) High-expansibility real-time and off-line data processing system
CN120416256B (en) Operation management method and system for server cluster
US12169737B1 (en) Task orchestration system in a trustless computing environment
CN120338744B (en) Conference activity execution task decomposition arrangement management method based on AI technology
US20250217537A1 (en) System and method for sigma-based and glideslope multi-path process modeling

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant