Granted Patent B2

US 12,639,648 · App. 18/968,095 · Granted May 26, 2026

Inventors: Zhensheng Yin (Beijing, CN); Xionghua Hu (Beijing, CN); Guitao Ding (Beijing, CN); Shanshan Wang (Beijing, CN); Huarui Zhang (Beijing, CN)

Assignee: Beijing Volcano Engine Technology Co., Ltd.

G06Q10/06315G06F11/32G06F11/327G06Q10/06316

Abstract

Claims (74)

1 . A method for baseline monitoring and alarming, the method comprising:

obtaining, by a server, business operation-related data of all task instances on a target baseline from a baseline guarantee task instance;

generating, by the server, a baseline monitoring link graph by adding the task instances and the corresponding business operation-related data to a buffer queue and traversing the buffer queue;

traversing, by the server, the baseline monitoring link graph from a baseline margin water level, wherein the baseline margin water level is a set of task instances corresponding to nodes with an in-degree of 0 in a baseline monitoring link, and determining, by the server, a predicted completion time of a task instance according to a predicted start time of the task instance, an upstream dependency state of the task instance, and a historical running duration of the task instance in the baseline monitoring link graph, comprising:

in a case that the task instance is traversed for a plurality of times, determining a plurality of initial predicted completion times of the task instance according to the predicted start time of the task instance, the upstream dependency state of the task instance, and the historical running duration of the task instance in the baseline monitoring link graph; and

comparing time magnitudes of the individual initial predicted completion times, and using an initial predicted completion time with a largest time as the predicted completion time of the task instance;

determining, by the server, a margin value of the target baseline according to a commitment completion time and the predicted completion time set for the task instance; and

determining, by the server and according to a comparison result between the margin value of the target baseline and an early warning margin or a comparison result between the margin value of the target baseline and a preset threshold, whether to trigger alarm information for the target baseline.

2 . The method according to claim 1 , wherein the generating, by the server, a baseline monitoring link graph by adding the task instances and the corresponding business operation-related data to a buffer queue and traversing the buffer queue comprises:

obtaining an upstream task instance on which the task instance depends;

obtaining other downstream task instances, other than the task instance, associated with the upstream task instance;

determining, according to a business time corresponding to an execution of a same task by the other downstream task instances, whether to add the upstream task instance to the baseline monitoring link graph; and

in response to the upstream task instance needing to be added to the baseline monitoring link graph, adding the upstream task instance to the baseline monitoring link graph, and recording a dependency relationship between the upstream task instance and the task instance.

3 . The method according to claim 1 , wherein after the generating, by the server, a baseline monitoring link graph by adding the task instances and the corresponding business operation-related data to a buffer queue and traversing the buffer queue, the method further comprises:

in a case that a dependency relationship between the task instances contained in the baseline monitoring link graph changes, determining a newly added upstream dependent task instance corresponding to a changed task instance, and adding the newly added upstream dependent task instance and an upstream dependent task instance of the newly added upstream dependent task instance to the baseline monitoring link graph together to obtain an updated baseline monitoring link graph; and/or

in a case that a dependency relationship between the task instances contained in the baseline monitoring link graph changes, determining an upstream dependent task instance removed corresponding to a changed task instance, and removing the removed upstream dependent task instance and an upstream dependent task instance of the removed upstream dependent task instance from the baseline monitoring link graph together to obtain an updated baseline monitoring link graph.

4 . The method according to claim 1 , wherein before the traversing, by the server, the baseline monitoring link graph from a baseline margin water level, the method further comprises:

obtaining a running change state of the task instance contained in the baseline monitoring link graph; and

in a case that a reference task instance is run and completed, removing the reference task instance from the baseline monitoring link graph, and updating the baseline margin water level according to the baseline monitoring link graph from which the reference task instance is removed.

5 . The method according to claim 1 , wherein before the traversing, by the server, the baseline monitoring link graph from a baseline margin water level, the method further comprises:

obtaining a polling time for traversing the baseline monitoring link graph; and

in response to a current moment reaching the polling time and the current moment is less than a baseline margin safety time threshold, no longer traversing the baseline monitoring link graph, wherein the baseline margin safety time threshold is a minimum value of safety running time thresholds of all the task instances on the target baseline, and the safety running time threshold of the task instance is a minimum value of predicted start times of all downstream task instances of the task instance.

6 . The method according to claim 1 , wherein after the determining, by the server, a predicted completion time of a task instance, the method further comprises:

in response to the predicted completion time being less than a baseline margin safety time threshold, removing the task instance from the baseline monitoring link graph, and no longer traversing the task instance, wherein the baseline margin safety time threshold is a minimum value of safety running time thresholds of all the task instances on the target baseline, and the safety running time threshold of the task instance is a minimum value of predicted start times of all downstream task instances of the task instance.

7 . The method according to claim 1 , wherein the determining, by the server and according to a comparison result between the margin value of the target baseline and an early warning margin or a comparison result between the margin value of the target baseline and a preset threshold, whether to trigger alarm information for the target baseline comprises:

in a case that the margin value of the target baseline is less than the preset threshold, or the margin value of the target baseline is less than the early warning margin, triggering an alarm; and

in a case that the margin value of the target baseline is greater than or equal to the preset threshold, or the margin value of the target baseline is greater than or equal to the early warning margin, not triggering an alarm.

8 . A computer device, comprising:

a memory and a processor, the memory and the processor being communicatively connected to each other, the memory storing computer instructions, and the processor executing the computer instructions to execute a method comprising:

obtaining business operation-related data of all task instances on a target baseline from a baseline guarantee task instance;

generating a baseline monitoring link graph by adding the task instances and the corresponding business operation-related data to a buffer queue and traversing the buffer queue;

traversing the baseline monitoring link graph from a baseline margin water level, wherein the baseline margin water level is a set of task instances corresponding to nodes with an in-degree of 0 in a baseline monitoring link, and determining a predicted completion time of a task instance according to a predicted start time of the task instance, an upstream dependency state of the task instance, and a historical running duration of the task instance in the baseline monitoring link graph, comprising:

in a case that the task instance is traversed for a plurality of times, determining a plurality of initial predicted completion times of the task instance according to the predicted start time of the task instance, the upstream dependency state of the task instance, and the historical running duration of the task instance in the baseline monitoring link graph; and

comparing time magnitudes of the individual initial predicted completion times, and using an initial predicted completion time with a largest time as the predicted completion time of the task instance;

determining a margin value of the target baseline according to a commitment completion time and the predicted completion time set for the task instance; and

determining, according to a comparison result between the margin value of the target baseline and an early warning margin or a comparison result between the margin value of the target baseline and a preset threshold, whether to trigger alarm information for the target baseline.

9 . The computer device according to claim 8 , wherein the generating a baseline monitoring link graph by adding to the task instances and the corresponding business operation-related data to a buffer queue and traversing the buffer queue comprises:

obtaining an upstream task instance on which the task instance depends;

obtaining other downstream task instances, other than the task instance, associated with the upstream task instance;

determining, according to a business time corresponding to an execution of a same task by the other downstream task instances, whether to add the upstream task instance to the baseline monitoring link graph; and

in response to the upstream task instance needing to be added to the baseline monitoring link graph, adding the upstream task instance to the baseline monitoring link graph, and recording a dependency relationship between the upstream task instance and the task instance.

10 . The computer device according to claim 8 , wherein after the generating a baseline monitoring link graph by adding the task instances and the corresponding business operation-related data to a buffer queue and traversing the buffer queue, the method further comprises:

in a case that a dependency relationship between the task instances contained in the baseline monitoring link graph changes, determining a newly added upstream dependent task instance corresponding to a changed task instance, and adding the newly added upstream dependent task instance and an upstream dependent task instance of the newly added upstream dependent task instance to the baseline monitoring link graph together to obtain an updated baseline monitoring link graph; and/or

in a case that a dependency relationship between the task instances contained in the baseline monitoring link graph changes, determining an upstream dependent task instance removed corresponding to a changed task instance, and removing the removed upstream dependent task instance and an upstream dependent task instance of the removed upstream dependent task instance from the baseline monitoring link graph together to obtain an updated baseline monitoring link graph.

11 . The computer device according to claim 8 , wherein before the traversing the baseline monitoring link graph from a baseline margin water level, the method further comprises:

obtaining a running change state of the task instance contained in the baseline monitoring link graph; and

in a case that a reference task instance is run and completed, removing the reference task instance from the baseline monitoring link graph, and updating the baseline margin water level according to the baseline monitoring link graph from which the reference task instance is removed.

12 . The computer device according to claim 8 , wherein before the traversing the baseline monitoring link graph from a baseline margin water level, the method further comprises:

obtaining a polling time for traversing the baseline monitoring link graph; and

in response to a current moment reaching the polling time and the current moment is less than a baseline margin safety time threshold, no longer traversing the baseline monitoring link graph, wherein the baseline margin safety time threshold is a minimum value of safety running time thresholds of all the task instances on the target baseline, and the safety running time threshold of the task instance is a minimum value of predicted start times of all downstream task instances of the task instance.

13 . The computer device according to claim 8 , wherein after the determining a predicted completion time of a task instance, the method further comprises:

in response to the predicted completion time being less than a baseline margin safety time threshold, removing the task instance from the baseline monitoring link graph, and no longer traversing the task instance, wherein the baseline margin safety time threshold is a minimum value of safety running time thresholds of all the task instances on the target baseline, and the safety running time threshold of the task instance is a minimum value of predicted start times of all downstream task instances of the task instance.

14 . The computer device according to claim 8 , wherein the determining, according to a comparison result between the margin value of the target baseline and an early warning margin or a comparison result between the margin value of the target baseline and a preset threshold, whether to trigger alarm information for the target baseline comprises:

in a case that the margin value of the target baseline is less than the preset threshold, or the margin value of the target baseline is less than the early warning margin, triggering an alarm; and

in a case that the margin value of the target baseline is greater than or equal to the preset threshold, or the margin value of the target baseline is greater than or equal to the early warning margin, not triggering an alarm.

15 . A non-transitory computer-readable storage medium, having computer instructions stored thereon for causing a computer to execute a method comprising:

obtaining business operation-related data of all task instances on a target baseline from a baseline guarantee task instance;

generating a baseline monitoring link graph by adding the task instances and the corresponding business operation-related data to a buffer queue and traversing the buffer queue;

traversing the baseline monitoring link graph from a baseline margin water level, wherein the baseline margin water level is a set of task instances corresponding to nodes with an in-degree of 0 in a baseline monitoring link, and determining a predicted completion time of a task instance according to a predicted start time of the task instance, an upstream dependency state of the task instance, and a historical running duration of the task instance in the baseline monitoring link graph, comprising:

in a case that the task instance is traversed for a plurality of times, determining a plurality of initial predicted completion times of the task instance according to the predicted start time of the task instance, the upstream dependency state of the task instance, and the historical running duration of the task instance in the baseline monitoring link graph; and

comparing time magnitudes of the individual initial predicted completion times, and using an initial predicted completion time with a largest time as the predicted completion time of the task instance;

determining a margin value of the target baseline according to a commitment completion time and the predicted completion time set for the task instance; and

determining, according to a comparison result between the margin value of the target baseline and an early warning margin or a comparison result between the margin value of the target baseline and a preset threshold, whether to trigger alarm information for the target baseline.

16 . The non-transitory computer-readable storage medium according to claim 15 , wherein the generating a baseline monitoring link graph by adding the task instances and the corresponding business operation-related data to a buffer queue and traversing the buffer queue comprises:

obtaining an upstream task instance on which the task instance depends;

obtaining other downstream task instances, other than the task instance, associated with the upstream task instance;

determining, according to a business time corresponding to an execution of a same task by the other downstream task instances, whether to add the upstream task instance to the baseline monitoring link graph; and

in response to the upstream task instance needing to be added to the baseline monitoring link graph, adding the upstream task instance to the baseline monitoring link graph, and recording a dependency relationship between the upstream task instance and the task instance.

17 . The non-transitory computer-readable storage medium according to claim 15 , wherein after the generating a baseline monitoring link graph by adding the task instances and the corresponding business operation-related data to a buffer queue and traversing the buffer queue, the method further comprises:

in a case that a dependency relationship between the task instances contained in the baseline monitoring link graph changes, determining a newly added upstream dependent task instance corresponding to a changed task instance, and adding the newly added upstream dependent task instance and an upstream dependent task instance of the newly added upstream dependent task instance to the baseline monitoring link graph together to obtain an updated baseline monitoring link graph; and/or

in a case that a dependency relationship between the task instances contained in the baseline monitoring link graph changes, determining an upstream dependent task instance removed corresponding to a changed task instance, and removing the removed upstream dependent task instance and an upstream dependent task instance of the removed upstream dependent task instance from the baseline monitoring link graph together to obtain an updated baseline monitoring link graph.

18 . The non-transitory computer-readable storage medium according to claim 15 , wherein before the traversing the baseline monitoring link graph from a baseline margin water level, the method further comprises:

obtaining a running change state of the task instance contained in the baseline monitoring link graph; and

in a case that a reference task instance is run and completed, removing the reference task instance from the baseline monitoring link graph, and updating the baseline margin water level according to the baseline monitoring link graph from which the reference task instance is removed.

Assignments (3)

Priority Claims (1)

Continuity (1)

References Cited (16)