CN115421371A - Computing power balance method of controller group, storage medium and redundancy system - Google Patents

Computing power balance method of controller group, storage medium and redundancy system Download PDF

Info

Publication number
CN115421371A
CN115421371A CN202211034221.1A CN202211034221A CN115421371A CN 115421371 A CN115421371 A CN 115421371A CN 202211034221 A CN202211034221 A CN 202211034221A CN 115421371 A CN115421371 A CN 115421371A
Authority
CN
China
Prior art keywords
controller
controllers
group
force
statistical process
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.)
Pending
Application number
CN202211034221.1A
Other languages
Chinese (zh)
Inventor
陈楚洪
牟桂贤
李丽强
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.)
Gree Electric Appliances Inc of Zhuhai
Original Assignee
Gree Electric Appliances Inc of Zhuhai
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 Gree Electric Appliances Inc of Zhuhai filed Critical Gree Electric Appliances Inc of Zhuhai
Priority to CN202211034221.1A priority Critical patent/CN115421371A/en
Publication of CN115421371A publication Critical patent/CN115421371A/en
Pending legal-status Critical Current

Links

Images

Classifications

    • G—PHYSICS
    • G05—CONTROLLING; REGULATING
    • G05B—CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
    • G05B9/00—Safety arrangements
    • G05B9/02—Safety arrangements electric
    • G05B9/03—Safety arrangements electric with multiple-channel loop, i.e. redundant control systems
    • Y—GENERAL 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
    • Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02D—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE
    • Y02D10/00—Energy efficient computing, e.g. low power processors, power management or thermal management

Landscapes

  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Engineering & Computer Science (AREA)
  • Automation & Control Theory (AREA)
  • Feedback Control In General (AREA)

Abstract

The invention discloses a computing power balancing method, a storage medium and a redundancy system of a controller group. The calculation force balance method of the controller group comprises a main controller and a controller, and comprises the following steps: distributing a corresponding statistical process for each controller, wherein the priority of the statistical process is lower than that of other processes; counting the initial calling times of the counting process in unit time when each controller only runs the counting process; when each controller simultaneously runs a statistical process and other processes, counting the current calling times of the statistical process in each unit time; calculating the computational load of each controller according to the initial times and the current calling times; and if the calculated force load between the controllers is unbalanced, performing calculated force balance adjustment on the corresponding controllers. The invention enables the calculation force between the controllers to be circulated according to the calculation force, and is beneficial to accurately balancing the calculation force before the controllers.

Description

Computing power balance method of controller group, storage medium and redundancy system
Technical Field
The invention relates to the technical field of redundant systems, in particular to a calculation force balancing method for a controller group.
Background
In a redundant system, the number of tasks performed by the primary controller is much greater than that performed by the secondary controller, such as: data acquisition, control issuing, data synchronization and the like, and the controller only needs to synchronize the data of the main controller, which is equivalent to a low power consumption state, and the working states of the controller and the controller have larger difference.
Under the environment of large engineering, the operating pressure of the main controller is increased sharply, the efficiency of task execution is also greatly reduced, and therefore when the main controller is in high load, the period of data acquisition of the main controller is correspondingly prolonged, and the service life of the main controller is not friendly.
And the controller still saves a low power consumption state, the utilization of hardware resources is small, and the performance of the whole system is low.
In addition, how to define the load of the controllers in the prior art has no simple and feasible method, and how to quantitatively balance the calculation force among the controller groups also has no reliable and feasible method.
Therefore, it is an urgent technical problem to be solved in the art to provide a method for computing power balance of a controller group.
Disclosure of Invention
The invention provides a calculation force balancing method, a storage medium and a redundancy system of a controller group, aiming at solving the technical problem that the load of a controller cannot be accurately defined when the calculation force of the controller group is balanced in the prior art.
The invention provides a calculation force balancing method of a controller group, wherein the controller group comprises a main controller and a controller, and the calculation force balancing method comprises the following steps:
distributing a corresponding statistical process for each controller, wherein the priority of the statistical process is lower than that of other processes;
counting the initial calling times of the counting process in unit time when each controller only runs the counting process;
when each controller simultaneously runs a statistical process and other processes, counting the current calling times of the statistical process in each unit time;
calculating the computational load of each controller according to the initial times and the current calling times;
and if the calculated force load between the controllers is unbalanced, performing calculated force balance adjustment on the corresponding controllers.
Further, whether the computational load between the controllers is balanced is judged by the following steps:
calculating the difference absolute value of the force calculation load between the two controllers;
if the absolute values of the difference values of the two controllers in unit time of a preset number exceed a preset range, judging that the computational load between the two controllers is unbalanced; otherwise, a computational load balance between the two controllers is determined.
Further, the calculation force balance adjustment of the two controllers is specifically a calculation force balance adjustment of equipment control tasks of the two controllers.
And further, receiving equipment control tasks of the main controller and the slave controller through the cache of the main controller.
Further, the force calculation balance adjustment of the equipment control tasks of the two controllers comprises the following steps:
calculating the buffer limit of the main controller according to the current calculation load of the two controllers;
and if the cache of the main controller receives that the equipment control tasks of the two controllers exceed the cache limit, forwarding the corresponding equipment control tasks to the controller.
Further, calculating the buffer limit of the main controller according to the calculation load of the current two controllers
Figure BDA0003818289270000021
Performing calculation, the buffer Limit of Buffer for the buffer limit of the host controller General assembly Is the maximum buffer of the main controller, status Master and slave Status for the calculated force load of the main controller Prepare for The calculated force load for the controller.
Further, by the formula
Figure BDA0003818289270000022
Calculating the computational load of each controller, wherein the status is the computational load of the corresponding controller, the maxcnt is the initial calling times, and the cnt is the current calling times.
Further, the computational force balance method further comprises the following steps:
in the initialization stage of each controller, generating the same task list for each controller, wherein the attributes of the task list comprise: the method comprises the steps of acquiring a unique identification number of equipment to be subjected to periodic data acquisition, the number of data point positions to be subjected to periodic data acquisition by the equipment and a unique identification number of a controller for controlling the periodic data acquisition of the equipment;
and synchronously modifying the task lists of all the controllers to ensure that the absolute value of the difference value between the number of the data point positions corresponding to each controller and the average number of the data point positions of the controllers is within a preset difference value range.
Further, in the power-on stage of the controller, only the statistical process is operated, and the initial calling times of the statistical process by the controller are counted.
Further, the priority of the statistical process is lower than the default priority, and the priorities of the other processes are greater than or equal to the default priority.
Furthermore, a corresponding control object is distributed to each controller, the control object is controlled to be opened and closed once when the counting process of the controller runs every time, and the calling times of the counting process are counted by counting the opening and closing times of the control object.
Further, the control object comprises any one of at least one led lamp, a display screen and a webpage.
The computer-readable storage medium provided by the present invention is used for running a computer program, and the computer program executes the computing power balancing method of the controller group according to the above technical solution when running.
The redundancy system provided by the invention comprises a main controller, a controller and at least one device controlled by the main controller and the controller, and the computing power of the controller is balanced by adopting the computing power balancing method between the main controller and the controller in the technical scheme.
The invention defines the calculation rule of the controller load by setting the statistical process, so that the load between the controllers can be visually displayed in a quantitative form, and meanwhile, a favorable basis is provided for accurately balancing the calculation force between the controllers in the follow-up process. The invention solves the problems of long data acquisition period and extremely low control task execution efficiency under the working condition of high load of the main controller, and also solves the problems that the controller is in a standby state for a long time when a redundancy system normally operates and the utilization of hardware resources is low. When the number of the controllers of the controller group is two, the invention effectively balances the working state between the two controllers, reduces the occupation of hardware resources such as a CPU (central processing unit) and a memory of the main controller, improves the operation efficiency of the main controller, effectively utilizes the hardware resources of the controller and improves the upper limit of the operation of a redundant system.
Drawings
The invention is described in detail below with reference to examples and figures, in which:
FIG. 1 is a principal flow diagram of the present invention.
FIG. 2 is a diagram of a redundant system architecture according to an embodiment of the present invention.
FIG. 3 is a flow chart illustrating the statistics of the number of calls made by the controller to the statistics process according to one embodiment of the present invention.
FIG. 4 is a flow chart of a method for balancing the forces between two controllers according to an embodiment of the present invention.
FIG. 5 is a flow chart illustrating control of the cache limit of the host controller according to one embodiment of the present invention.
Detailed Description
In order to make the technical problems, technical solutions and advantageous effects to be solved by the present invention more clearly apparent, the present invention is further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the invention and do not limit the invention.
Thus, a feature indicated in this specification will serve to explain one of the features of one embodiment of the invention, and does not imply that every embodiment of the invention must have the stated feature. Further, it should be noted that this specification describes many features. Although some features may be combined to show a possible system design, these features may also be used in other combinations not explicitly described. Thus, the combinations illustrated are not intended to be limiting unless otherwise specified.
The computing power balance method of the controller group comprises two controllers, wherein one controller is a main controller, and the other controller is a sub controller. The selection of the master controller may be based on a number of rules, such as selecting the controller that starts the fastest as the master controller, and one skilled in the art may select the most suitable master controller based on other rules. Based on the established rules, the defined master controller after each boot is not necessarily the same as the last defined master controller.
As shown in fig. 1, the present invention allocates a corresponding statistical process to each controller, that is, each controller has a statistical process of its own, and all statistical processes have lower priority than other processes running in the same controller. In one embodiment, the priority of the statistical process may be set lower than the default priority, and the priorities of the other processes may be set equal to or greater than the default priority.
Firstly, counting the initial calling times of the statistical process in unit time when each controller only runs the statistical process. And then when the controllers simultaneously run the statistical process and other processes, counting the current calling times of the statistical process of each controller in each unit time, calculating the computational load of each controller according to the initial times and the current calling times, and if the computational load between the controllers is unbalanced, performing computational balance adjustment on the corresponding controllers.
In one embodiment, since one of the plurality of controllers is the main controller, the operations of counting various call times of the counting process of each controller, calculating the computing load of each controller, balancing and adjusting the computing power among the controllers, and the like, can be realized by the management process of the main controller.
In one embodiment, only the statistical process may be run by each controller during the power-up phase, thereby obtaining the initial number of calls to the statistical process by the controller.
After the power is on for a period of time, each controller can normally run, namely the controllers run a statistical process and other processes at the same time, and the other processes of the invention refer to processes except the statistical process and comprise other task processes, management processes and the like. The priority of the statistical process is lower than that of other processes, so that the controller can preferentially execute other processes, the current calling times of the statistical process in each unit time are counted, and the calculation load of each controller can be obtained through comparison calculation of the initial times and the current calling times.
In one embodiment, by formula
Figure BDA0003818289270000041
Calculating the computational load of each controller, wherein status is the computational load of the corresponding controller, maxcnt is the initial calling times of the corresponding controller, and cnt is the current calling times of the controller.
In an embodiment, in order to implement the invocation of the statistical process by the controller, a corresponding control object may be allocated to each controller, that is, the statistical process of each controller has a control object dedicated to the statistical process, the statistical process of the controller controls the control object corresponding to the statistical process to be opened and closed once when running each time, and the invocation frequency of the statistical process is counted by counting the opening and closing frequency of the control object.
The control object of the present invention includes, but is not limited to, any one of at least one led lamp, a display screen, and a web page. As shown in fig. 3, taking an led lamp as an example, when the statistical process runs, the led lamp corresponding to the statistical process is controlled to flash once, that is, turned on and turned off once, so that it can be intuitively observed that the statistical process is called once at this time. The LED lamp controlled by the controller is subjected to flicker control by using a statistical process, and the number of times of controlling the LED lamp by the statistical process in a time period T (second-level time, for example, 1 second) is recorded as maxcnt, namely the number of times of initial calling of the controller. And then starting a management process and other task processes, wherein the priority of the other processes is higher than the scheduling priority of the statistical process, and the control times of the LED lamp of the statistical process in the time period T are different from the times when only the statistical process operates and are recorded as cnt, namely the current calling times of each time period. The running state of the program of the controller, namely the calculation power born by the controller can be obtained through the calculation formula.
Taking a web page as an example, when the statistical process runs, the corresponding web page is opened and closed once, so that the statistical process can be captured at the computer program end and called once.
As shown in fig. 4 and 5, based on the above embodiment, when the number of controllers is 2, one of the controllers is the master controller, and the other controller is the slave controller, the device control tasks of the two controllers are received through the cache of the master controller, that is, the device control tasks of the slave controller are derived from the allocation of the master controller. At this time, the force balance adjustment for the plant control tasks of the two controllers is mainly performed as follows.
Calculating the buffer limit of the main controller according to the current calculated force loads of the two controllers, and in one embodiment, calculating the buffer limit of the main controller according to the current calculated force loads of the two controllers is specifically calculated by a formula
Figure BDA0003818289270000051
Performing calculation, buffer Limit of Buffer for the buffer limit of the host controller General assembly Is the maximum buffer of the main controller, status Master and slave Status for the calculated force load of the main controller Prepare for The calculated force load for the controller. The load of the master controller is approximately balanced with the load of the slave controller by dynamically setting the buffer limit of the master controller.
For example, there are two controllers in total, one of which is a main controller and the other is a slave controller, when the redundant system operates normally, the management process of the main controller collects the status value of the statistical process of the main controller and the status value of the statistical process of the slave controller respectively according to the time period T, and simultaneously performs difference calculation to obtain the value of the parameter sub:
sub＝|status master and slave -status From |
The value of sub needs to be kept within a preset range. If the sub values exceed the preset range in a plurality of continuous time periods T, the redundancy system performs a round of task strategy adjustment: the controller with the higher status value transfers a portion of the running task to the one-side controller with the lower status value.
In the above embodiment, the calculation force balance adjustment performed on the two controllers is specifically performed on the device control tasks of the two controllers. When the controller performs calculation balance adjustment, tasks to be adjusted can be divided into equipment periodic data acquisition tasks and equipment control tasks. The load regulated by the above embodiments mainly refers to the plant control task, and the regulation controller expends the computational effort on the plant control task.
The calculation power balance method of the invention also comprises the step of adjusting the calculation power spent on the periodical data acquisition task of the equipment by the controller, and the adjustment of the periodical data acquisition task of the equipment is described below.
In the initialization stage of each controller, the same task list is generated for each controller, the attributes of the task list are mainly 3 columns, and the method comprises the following steps: the unique identification number of the device (such as a device number) to be subjected to the periodic data acquisition, the number of data points to be subjected to the periodic data acquisition by the device, and the unique identification number of the controller (such as a controller name) to control the periodic data acquisition by the device. A specific embodiment of the task list is shown in the following table, and the task lists of all the controllers are modified synchronously, so that the absolute value of the difference between the number of data point locations corresponding to each controller and the average number of data point locations of the controllers is within a preset difference range.
When the controller executes a periodic data acquisition task, a task list is inquired, and the controller is determined to acquire data of the equipment according to the equipment number and the executor.
Device numbering Number of points of the device Executive person
Device 1 n Main controller
Device 2 m Controller
…… …… ……
TABLE 1 one embodiment of a task List
When task scheduling adjustment is carried out in an initialization stage, the main controller counts the equipment and the point location number thereof which are respectively responsible for acquisition by the main controller and the sub-controller, and the total number of the acquisition point locations of all the controllers in the controller group is close to the same as much as possible by synchronously modifying the 'executor' attribute of the task list of the main controller and the sub-controller, thereby achieving the effect of dynamic balance of acquisition tasks.
The invention can ensure that the redundant system adopting the controller group can intuitively and accurately reflect the load born by each controller, thereby dynamically balancing the calculated force among the controllers according to the calculated force load of each controller, avoiding overhigh load among the controllers in the redundant system and overlow utilization rate of hardware resources, ensuring that all the controllers can bear all the loads in a balanced manner, and prolonging the service life of each controller.
The invention also protects a redundant system, which comprises two controllers which are respectively a main controller and a controller, and at least one device controlled by the main controller and the controller, wherein the redundant system adopts the calculation force balancing method of the controller group in the technical scheme to balance the calculation force between the main controller and the controller.
Fig. 2 shows a specific hardware architecture diagram of a redundant system of the present invention, in this embodiment, two controllers are provided, one is a master controller, and the other is a slave controller (also referred to as slave controller), the master controller and the slave controller are connected via redundant communication lines, the upstream of each of the two controllers is connected to an upper computer via a router, the downstream of each of the two controllers is connected to at least one field device via a router, and the two controllers control the field devices together according to control instructions of the upper computer, i.e., the master controller and the slave controller perform good division cooperation according to the method for balancing computational power of the controller group of the present invention, so as to control the field devices.
The invention also protects a computer-readable storage medium for running a computer program which, when running, executes the method for computing force balancing of a controller group according to the above-mentioned solution.
The above description is only for the purpose of illustrating the preferred embodiments of the present invention and is not to be construed as limiting the invention, and any modifications, equivalents and improvements made within the spirit and principle of the present invention are intended to be included within the scope of the present invention.

Claims (14)

1. A method of computing force balance of a controller group, the controller group comprising a master controller and a slave controller, the method comprising:
distributing a corresponding statistical process for each controller, wherein the priority of the statistical process is lower than that of other processes;
counting the initial calling times of the counting process in unit time when each controller only runs the counting process;
when each controller simultaneously runs a statistical process and other processes, counting the current calling times of the statistical process in each unit time;
calculating the computational load of each controller according to the initial times and the current calling times;
and if the calculated force load between the controllers is unbalanced, performing calculated force balance adjustment on the corresponding controllers.
2. The method for computing force balance of a group of controllers as claimed in claim 1, wherein whether the computing force loads between the controllers are balanced is determined by:
calculating the difference absolute value of the force calculation load between the two controllers;
if the absolute values of the difference values of the two controllers in unit time of a preset number exceed a preset range, judging that the calculated force load between the two controllers is unbalanced; otherwise, a computational load balance between the two controllers is determined.
3. The method for computing force balance of a controller group according to claim 2, wherein the computing force balance adjustment of the two controllers is specifically the computing force balance adjustment of the equipment control tasks of the two controllers.
4. The method of force balancing of a group of controllers according to claim 3, wherein the device control tasks of the master controller and the slave controller are received via the cache of the master controller.
5. The method for computational power balance of a group of controllers as claimed in claim 4, wherein the computational power balance adjustment of the plant control tasks of two controllers comprises the steps of:
calculating the buffer limit of the main controller according to the current calculation load of the two controllers;
and if the cache of the main controller receives that the equipment control tasks of the two controllers exceed the cache limit, forwarding the corresponding equipment control tasks to the controller.
6. The method of claim 5, wherein the calculation of the compute power balance of the controller group based on the compute power loads of the two controllers is performed by a formula
Figure FDA0003818289260000011
Calculating the buffer Limit of Buffer for the buffer limit of the host controller General assembly Status is the maximum buffer of the main controller Main For the calculated force load of the main controller, status Is provided with The calculated force load for the controller.
7. Method for the computational force balancing of a group of controllers according to claim 1, characterised by the fact that it is obtained by means of a formula
Figure FDA0003818289260000021
Calculating the computational load of each controller, wherein the status is the computational load of the corresponding controller, the maxcnt is the initial calling times, and the cnt is the current calling times.
8. The method of computational force balancing of a group of controllers according to claim 1, further comprising:
in the initialization stage of each controller, generating the same task list for each controller, wherein the attributes of the task list comprise: the method comprises the steps of acquiring a unique identification number of equipment to be subjected to periodic data acquisition, the number of data point positions to be subjected to periodic data acquisition by the equipment and a unique identification number of a controller for controlling the periodic data acquisition of the equipment;
and synchronously modifying the task lists of all the controllers to ensure that the absolute value of the difference value between the number of the data point positions corresponding to each controller and the average number of the data point positions of the controllers is within a preset difference value range.
9. The method for force balance of a controller group according to claim 1, wherein only the statistical process is run during the power-on phase of the controller, and the initial number of calls of the controller to the statistical process is counted.
10. The method for computationally balancing a group of controllers as claimed in claim 1, wherein said statistical process has a lower priority than a default priority and said other processes have a priority greater than or equal to the default priority.
11. The method for force balance of controller group according to claim 1, wherein each controller is assigned with a corresponding control object, the statistical process of the controller controls the control object to open and close once every time it runs, and the number of times of calling of the statistical process is counted by counting the number of times of opening and closing of the control object.
12. The method for force balancing of a controller group according to claim 11, wherein the control object comprises any one of at least one led lamp, a display screen, and a web page.
13. A computer-readable storage medium for running a computer program, characterized in that the computer program when running performs the method of computational force balancing of a group of controllers according to any one of claims 1 to 11.
14. A redundant system comprising a main controller and a sub-controller, and at least one device controlled by said main controller and sub-controller, characterized in that the computational power between said main controller and sub-controller is balanced using the computational power balancing method of the group of sub-controllers according to any one of claims 1 to 12.
CN202211034221.1A 2022-08-26 2022-08-26 Computing power balance method of controller group, storage medium and redundancy system Pending CN115421371A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202211034221.1A CN115421371A (en) 2022-08-26 2022-08-26 Computing power balance method of controller group, storage medium and redundancy system

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202211034221.1A CN115421371A (en) 2022-08-26 2022-08-26 Computing power balance method of controller group, storage medium and redundancy system

Publications (1)

Publication Number Publication Date
CN115421371A true CN115421371A (en) 2022-12-02

Family

ID=84200697

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202211034221.1A Pending CN115421371A (en) 2022-08-26 2022-08-26 Computing power balance method of controller group, storage medium and redundancy system

Country Status (1)

Country Link
CN (1) CN115421371A (en)

Similar Documents

Publication Publication Date Title
CN107066332B (en) Distributed system and scheduling method and scheduling device thereof
CN106557369B (en) Multithreading management method and system
US10423217B1 (en) Dynamic power capping of multi-server nodes in a chassis based on real-time resource utilization
US9612875B2 (en) Operational-task-oriented system and method for dynamically adjusting operational environment
CN103179048A (en) Method and system for changing main machine quality of service (QoS) strategies of cloud data center
CN113032102B (en) Resource rescheduling method, device, equipment and medium
CN112711479A (en) Load balancing system, method and device of server cluster and storage medium
WO2024060788A1 (en) Intelligent-computing-oriented adaptive adjustment system and method for pipeline-parallel training
CN112486642B (en) Resource scheduling method, device, electronic equipment and computer readable storage medium
CN106569897A (en) Shared bus polling method and device based on collaborative multi-task scheduling mechanism
CN108446180A (en) A kind of data center dynamic method for scheduling task based on Data Migration
CN114968601B (en) Scheduling method and scheduling system for AI training jobs with resources reserved in proportion
CN109739634A (en) A kind of atomic task execution method and device
CN113688001A (en) Dynamic balancing method and device for server hard disk power consumption, terminal and storage medium
CN114816699A (en) Data center job scheduling method and system based on temperature prediction
EP3916554A1 (en) A throughput-optimized, quality-of-service aware power capping system
CN110794949A (en) Power consumption reduction method and system for automatically allocating computing resources based on component temperature
CN113961353A (en) Task processing method and distributed system for AI task
CN117453360A (en) Resource scheduling method and device for computing task
CN115421371A (en) Computing power balance method of controller group, storage medium and redundancy system
JP4409568B2 (en) Band control program and multiprocessor system
CN115248734B (en) Private cloud multi-tenant resource quota self-adaptive adjustment method and device
CN116643858A (en) Service priority pod-based rescheduling method, device, equipment and medium
CN112084089B (en) Method, device and equipment for determining upper limit of power consumption of data center node and storage medium
CN115202810A (en) Kubernetes working node distribution method and system

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