CN108234188A - A kind of business platform resource scheduling processing method and device - Google Patents

Abstract

The invention discloses a service platform resource scheduling processing method and device. Wherein, the method includes: obtaining work index information of each resource in the service platform through a fault collection interface, and obtaining a resource fault set according to the work index information and preset health index information used to indicate whether a resource fails; Wherein, the resource failure set includes resources that have failed in the current cycle; a data flow table is obtained through a data switch, and the data flow table includes resources that were in a normal working state in the previous cycle; according to the resource failure set, the The above data flow table, as well as the preset scheduling rules, determine the resources that need to be isolated and/or recovered in the current period. The device is used to perform the above method. The service platform resource scheduling processing method and device provided by the present invention avoid using faulty resources in the resource scheduling process, thereby improving the efficiency of resource scheduling.

Description

A service platform resource scheduling processing method and device

technical field

The invention relates to the technical field of the Internet, in particular to a method and device for scheduling and processing resources of a business platform.

Background technique

In the Internet+ era, wireless communication service functions continue to be enhanced and improved, and industry SMS services are booming. For example, industry gateway-based industry SMS services provided to major customers such as banks, securities, and e-commerce mainly include verification code SMS, member notification SMS, Affiliate Marketing SMS etc. With the increasing scale and importance of industry SMS business, customers have put forward higher requirements for the arrival speed, arrival rate, channel stability and security of industry SMS.

At present, major network operators are exploring various methods, either to reduce the probability of failure, or to deal with sudden failure most effectively, in order to minimize the scope and degree of business impact. Existing resource scheduling methods mainly focus on the resources themselves, such as resource CPU, memory, network bandwidth, utilization rate, etc., and use these indicators to reach a certain threshold as the key basis, and perform resource scheduling according to a predetermined strategy to achieve resource mitigation. The purpose of using pressure. However, there are many reasons for failures that actually affect business stability and normal operation. In addition to the resource itself, it also involves application software operation, link status, and service indicators. In addition, if the resource in use is abnormal, the resource node fails, or a sudden failure occurs, it will also affect the scheduling of resources. In addition, if resources are over-scheduled or frequently scheduled, there may be too few resources in use to carry the current business and cause new failures, or frequent resource scheduling may lead to poor system stability.

Therefore, how to propose a method that can improve the efficiency of resource scheduling in the process of resource scheduling has become an important issue to be solved urgently in the industry.

Contents of the invention

Aiming at the defects in the prior art, the present invention provides a method and device for scheduling and processing service platform resources.

On the one hand, the present invention proposes a resource scheduling processing method for a service platform, including:

The controller obtains the work index information of each resource in the service platform, and obtains a resource fault set according to the work index information and the preset health index information used to indicate whether the resource fails; wherein, the resource fault set includes the current A resource that failed during the period;

The controller obtains a data flow table, and the data flow table includes resources in a normal working state in the previous cycle;

The controller determines resources that need to be isolated and/or recovered in the current period according to the resource failure set, the data flow table, and a preset scheduling rule.

In another aspect, the present invention provides a service platform resource scheduling processing device, including:

A processing unit, configured to obtain work index information of each resource in the service platform, and obtain a resource failure set according to the work index information and preset health index information used to indicate whether a resource fails; wherein, the resource failure The pool includes resources that failed during the current cycle;

An acquisition unit, configured to acquire a data flow table, the data flow table including resources in a normal working state in the previous cycle;

A processing unit, configured to determine resources that need to be isolated and/or recovered in the current period according to the resource failure set, the data flow table, and a preset scheduling rule.

The resource scheduling processing method and device of the service platform provided by the present invention can obtain the resource failure set by comparing the work indicator information and the health indicator information, and obtain the data flow table through the controller, and according to the resource failure set, data flow table and preset scheduling rules to determine the resources that need to be isolated and/or restored, avoiding the use of faulty resources in the resource scheduling process, thereby improving the efficiency of resource scheduling.

Description of drawings

In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings that need to be used in the description of the embodiments or the prior art. Obviously, the accompanying drawings in the following description These are some embodiments of the present invention. Those skilled in the art can also obtain other drawings based on these drawings without creative work.

FIG. 1 is a schematic diagram of an architecture of a resource scheduling system based on SDN technology according to an embodiment of the present invention;

FIG. 2 is a schematic flow diagram of a method for processing resource scheduling on a service platform according to an embodiment of the present invention;

FIG. 3 is a schematic flow diagram of a method for processing resource scheduling on a service platform according to another embodiment of the present invention;

FIG. 4 is a schematic flowchart of a method for processing resource scheduling of a service platform according to another embodiment of the present invention;

FIG. 5 is a schematic structural diagram of a service platform resource scheduling processing device according to an embodiment of the present invention;

6 is a schematic structural diagram of a service platform resource scheduling processing device according to another embodiment of the present invention;

7 is a schematic structural diagram of a service platform resource scheduling processing device according to another embodiment of the present invention;

FIG. 8 is a schematic diagram of a physical structure of an electronic device according to an embodiment of the present invention.

Detailed ways

In order to make the purpose, technical solutions and advantages of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly described below in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the present invention Examples, not all examples. Based on the embodiments of the present invention, all other embodiments obtained by persons of ordinary skill in the art without making creative efforts belong to the protection scope of the present invention.

FIG. 1 is a schematic diagram of a resource scheduling system architecture based on SDN technology according to an embodiment of the present invention, which is an application environment of the embodiment of the present invention. Among them, Software Defined Network (Software Defined Network, hereinafter referred to as SDN) is a new network innovation architecture and a way to realize network virtualization. Its core technology, OpenFlow, separates the control plane of network equipment from the data plane. In this way, the flexible control of network traffic is realized, and the network becomes more intelligent as a pipeline.

As shown in FIG. 1 , the service platform 105 is composed of multiple application servers, which provide resources required for processing services. The function of the fault collection interface 104 is to obtain information of the service platform 105 and upload the obtained information to the controller 102 . The data switch 103 adopts the OpenFlow protocol interface of the SDN technology, can upload the data flow table to the controller 102, and forward the data packets according to the instructions issued by the controller 102. The controller 102 can receive the information uploaded by the fault collection interface 104 , and can also send a control command to the data switch 103 to update the flow table information of the data switch 103 . The SDN Manager 101 is a manual console, which provides a management man-machine interface for the SDN controller 102, and manages various parameters of the controller 102, parameters of the OpenFlow protocol, parameters of the fault collection interface 104, etc. through the console 102.

Fig. 2 is a schematic flow chart of a method for scheduling and processing service platform resources according to an embodiment of the present invention. As shown in Fig. 2 , the method for scheduling and processing service platform resources provided by the present invention includes:

S201. The controller obtains the work index information of each resource in the service platform, and obtains a resource fault set according to the work index information and the preset health index information used to indicate whether a resource fails; wherein, the resource fault set Include resources that failed during the current cycle;

Specifically, the fault collection interface can obtain the work index information of each resource in the service platform, and upload the obtained work index information to the controller, and the fault collection interface can obtain the work index information periodically , that is, it is acquired every predetermined time, for example, every 30s. The predetermined time can be set according to the actual situation, which is not limited in this embodiment of the present invention. The work index information is preset from the perspective of business processing, including but not limited to: processor process or thread running status, processor and database connectivity, processor message processing success rate, processor message queue backlog, Proportion of processor exception error codes. The controller receives the work information index, and compares the work index information of each resource with preset health index information, and if the work index information does not meet the preset health index information condition, the controller determines the resource corresponding to the work index information as a faulty resource. All the faulty resources judged in the current cycle constitute a faulty resource set. The preset health index information corresponds to the work information index, and the conditions for satisfying the resource are set. The resource must meet all the conditions before it is judged as a resource in a normal working state by the controller. i.e. normal resources. For example, the process or thread of the processor is running normally, the connectivity between the processor and the database is normal, and the message processing success rate of the processor is not less than 60%. The work index information and the preset health index information are correspondingly set according to the actual working service platform, which is not limited in this embodiment of the present invention.

S202. The controller obtains a data flow table, and the data flow table includes resources in a normal working state in the previous cycle;

Specifically, the data switch uploads the stored data flow table to the controller, and the data flow table includes resources in a normal working state in the previous cycle, hereinafter referred to as normal resources, and the controller receives the data The data flow table uploaded by the switch.

S203. The controller determines resources that need to be isolated and/or recovered in the current period according to the resource failure set, the data flow table, and a preset scheduling rule.

Specifically, after the controller acquires the resource fault set and the data flow table, according to the resource fault set, the data flow table, and a preset scheduling rule, obtains that the current cycle needs isolation and/or recovery Resources. For example, the controller may determine a normal resource that belongs to the resource failure set and appears in the data flow table as a resource that needs to be isolated; it will not belong to the resource failure set and does not appear in all The normal resources in the above data flow table are determined as the resources that need to be restored. For the resources that need to be isolated and/or recovered in the current cycle, the isolation operation will be performed on the resources that need to be isolated in subsequent processing, that is, the resources that need to be isolated that exist in the data flow table will be deleted , performing a recovery operation on the resource that needs to be recovered, that is, adding the resource that does not exist in the data flow table and that needs to be recovered.

The service platform resource scheduling processing method provided by the present invention can obtain the resource fault set by comparing the work index information and the health index information, and obtain the data flow table through the controller, and according to the resource fault set, data flow table and preset The scheduling rules are set to determine the resources that need to be isolated and/or restored, avoiding the use of faulty resources in the resource scheduling process, thereby improving the efficiency of resource scheduling.

Fig. 3 is a schematic flowchart of a method for scheduling resource scheduling on a service platform according to another embodiment of the present invention. As shown in Fig. 3 , on the basis of the foregoing embodiments, further, the method for scheduling resource scheduling on a service platform provided by an embodiment of the present invention further includes :

S204. The controller sends a control instruction to the data switch according to the resource that needs to be isolated and/or restored, so that the data switch updates the data flow table, and according to the updated data flow table to Each resource in the service platform forwards data.

Specifically, after obtaining the resource that needs to be isolated and/or restored, the controller sends a control instruction to the data switch to update the data flow table of the data switch, and the resource that needs to be isolated delete from the data flow table, add resources to be restored to the data flow table, and the data switch forwards data to each resource in the service platform according to the updated data flow table.

The service platform resource scheduling processing method provided by the present invention can obtain the resource fault set by comparing the work index information and the health index information, and obtain the data flow table through the controller, and according to the resource fault set, data flow table and preset The scheduling rules are set to determine the resources that need to be isolated and/or restored, avoiding the use of faulty resources in the resource scheduling process, thereby improving the efficiency of resource scheduling. By updating the data flow table in the data switch, it is ensured that the resources used in the data flow table are all normal resources.

On the basis of the above embodiments, further, the dispatching rules include:

If it is judged that the number of isolated resources at the current moment reaches the scheduling critical value, the recovered isolated resources will be restored;

If it is determined that the number of isolated resources at the current moment has not reached the scheduling threshold, and the recovered isolated resources do not belong to frequently scheduled resources, recover the recovered isolated resources;

If it is determined that the number of isolated resources at the current moment has not reached the scheduling threshold, and the recovered isolated resources belong to frequently scheduled resources, then the recovered isolated resources are not recovered;

Wherein, according to the resource failure set and the data flow table, determine the recovered isolated resource that occurs in the current cycle, and the recovered isolated resource refers to a fault in the previous cycle but in the current cycle The resources that are in normal working state during the cycle; the scheduling critical value refers to the maximum number of resources that can be isolated in the service platform under the condition that the current business volume is guaranteed to be processed normally.

Specifically, the controller compares the number of isolated resources at the current moment with the scheduling critical value, and if the number of isolated resources at the current moment is greater than or equal to the scheduling critical value, then the controller The recovered isolated resources are recovered to ensure the processing of traffic. In the current cycle, the number of isolated resources at the current moment may change, and when the controller compares the number of isolated resources at the current moment with the scheduling threshold for the first time, the The number of isolated resources at the current moment is equal to the number of faulty resources in the previous cycle; in the current cycle, if a resource is restored, the corresponding number of isolated resources at the current moment is correspondingly reduced. If a certain resource has been isolated, the number of corresponding resources that have been isolated at the current moment increases accordingly.

The controller compares the number of isolated resources at the current moment with the scheduling threshold, and if the number of isolated resources at the current moment is less than the scheduling threshold, and the recovered isolated resources do not belong to frequent scheduling resource, then the controller restores the restored isolated resource.

The controller compares the number of isolated resources at the current moment with the scheduling threshold, and if the number of isolated resources at the current moment is less than the scheduling threshold, and the recovered isolated resources belong to frequently scheduled resource, the controller does not restore the restored isolated resource.

Wherein, the controller obtains all faulty resources in the current cycle according to the resource fault set, and can determine the normal resources in the current cycle in combination with all resources of the service platform; the controller according to the data flow table , obtain the normal resources in the last cycle, combined with all the resources of the service platform, can determine all the faulty resources in the last cycle; the controller according to the normal resources in the current cycle and the All faulty resources in the current cycle can be determined to determine the recovered isolated resources, and the recovered isolated resources are resources that failed in the previous cycle but are in a normal working state in the current cycle.

The scheduling critical value refers to the maximum number of resources that can be isolated in the service platform under the condition that the current business volume is guaranteed to be processed normally. For example, according to the business volume, the business platform has 5 resources in total. If 3 resources are needed to ensure the normal processing of the current business volume, that is, the current maximum number of resources that can be isolated is 2, and the scheduling critical value Δ=2. The determination of the scheduling critical value is related to the current business volume of the service platform, and it is adjusted adaptively with the change of the current business volume. The greater the current business volume, the more resources are needed, and the fewer resources that can be isolated should be. Therefore, the scheduling The critical value should also be smaller. The scheduling threshold can be determined according to the following formula Among them, N is the current business volume of the business platform, M is the maximum processing business volume of the business platform, S is the total number of resources of the business platform, and k is greater than 0 but less than or equal to 1, which is the utilization rate of the business platform, usually to ensure the normal operation of the business platform , k is less than 1, for example, k is 90%.

The service platform resource scheduling processing method provided by the present invention can obtain the resource fault set by comparing the work index information and the health index information, and obtain the data flow table through the controller, and according to the resource fault set, data flow table and preset The scheduling rules are set to determine the resources that need to be isolated and/or restored, avoiding the use of faulty resources in the resource scheduling process, thereby improving the efficiency of resource scheduling. By giving the restored isolated resource scheduling rule, it is convenient to judge whether to restore the restored isolated resource.

On the basis of the above-mentioned embodiments, further, the method further includes:

Obtain the number of recovery times of the recovered isolated resource within a unit time, if the number of times is greater than a threshold, determine that the recovered isolated resource belongs to a frequently scheduled resource; if the number of times is not greater than the threshold, then It is determined that the recovered isolated resource does not belong to a frequently scheduled resource.

Specifically, the frequently scheduled resource means that the number of times the resource is isolated or restored within a unit time reaches a threshold, and the unit time and the threshold are set according to actual conditions and are not limited in this embodiment of the present invention. The controller acquires the number of recovery times of the recovered isolated resource within a unit time, if the number of recovery times is greater than the threshold, it indicates that the recovered isolated resource belongs to the frequently scheduled resource, if the recovered If the number of recovery times is not greater than the threshold, it indicates that the recovered isolated resource belongs to the frequently scheduled resource. When the resource is frequently restored or isolated, it means that the resource has been judged as a faulty resource many times in a short period of time. Although the resource has been restored many times, if the resource is used, there is a great risk of failure, increasing Instability of resource scheduling Therefore, it is judged that the frequently scheduled resource will not be recovered.

For example, mark the time point when each resource is isolated or recovered, and time windows with different granularities can be selected according to the nature of the business. When a resource is isolated or recovered three or more times in the same time window, then It can be determined that this resource is a resource that is frequently scheduled. For services that require high network stability on the service platform, it is recommended to select a time window with a smaller granularity.

The service platform resource scheduling processing method provided by the present invention can obtain the resource fault set by comparing the work index information and the health index information, and obtain the data flow table through the controller, and according to the resource fault set, data flow table and preset The scheduling rules are set to determine the resources that need to be isolated and/or restored, avoiding the use of faulty resources in the resource scheduling process, thereby improving the efficiency of resource scheduling. By judging frequently scheduled resources, the efficiency of resource scheduling is further improved.

On the basis of the above embodiments, further, the dispatching rules include:

If it is determined that the number of isolated resources at the current moment reaches the scheduling critical value, then calculate the number of times that all faulty resources in the service platform are isolated within a preset time period, and prioritize all faulty resources according to the number of times sorting, recovering faulty resources with low priority, and isolating faulty resources with high priority; wherein, the lower the number of times, the lower the corresponding priority of the faulty resources;

If it is determined that the number of isolated resources at the current moment has not reached the scheduling critical value, and the faulty resource is a faulty resource that has not been isolated, then isolate the faulty resource; wherein, the faulty resource that is not isolated Faulty resources refer to resources that were not isolated in the previous cycle and failed in the current cycle;

Wherein, according to the resource fault set, the fault resources occurring in the current cycle are determined; the scheduling critical value refers to the maximum number of resources that can be isolated in the service platform under the condition of ensuring normal processing of the current traffic.

Specifically, the controller compares the number of isolated resources at the current moment with a scheduling critical value, and if the number of isolated resources at the current moment is greater than or equal to the scheduling critical value, then the faulty resource with a low priority Restoring is performed so that the number of isolated resources is less than the scheduling critical value, ensuring that the number of available resources in the current cycle can meet the normal processing of the current traffic. The low-priority faulty resources are obtained by the controller by calculating the number of times all faulty resources in the service platform are isolated within a preset time period, and prioritizing all faulty resources according to the number of times , the lower the number of times, the lower the priority corresponding to the faulty resource. It can be understood that, after the number of isolated resources is less than the scheduling threshold, unrecovered faulty resources with high priority may be isolated.

The controller compares the number of isolated resources at the current time with the scheduling critical value, and if the number of isolated resources at the current time is less than the scheduling critical value, then check whether the faulty resource is not Judgment of faulty resources that are isolated. The controller obtains the resources in the normal working state in the previous period according to the data flow table, and if the faulty resource appears in the data flow table, then the faulty resource is the resource that has not been isolated, Isolating the faulty resources that have not been isolated. The non-isolated faulty resource refers to a resource that was not isolated in the previous cycle and has a fault in the current cycle.

Wherein, the controller obtains the faulty resources occurring in the current cycle according to the resource fault set. The scheduling critical value has been explained in the above-mentioned embodiments, and will not be repeated here.

The service platform resource scheduling processing method provided by the present invention can obtain the resource fault set by comparing the work index information and the health index information, and obtain the data flow table through the controller, and according to the resource fault set, data flow table and preset The scheduling rules are set to determine the resources that need to be isolated and/or restored, avoiding the use of faulty resources in the resource scheduling process, thereby improving the efficiency of resource scheduling. By giving fault resource scheduling rules, it is convenient to judge whether to recover or isolate fault resources.

FIG. 4 is a schematic flowchart of a method for scheduling and processing service platform resources according to another embodiment of the present invention. The method for scheduling and processing service platform resources provided by the present invention will be illustrated below with reference to FIG. 4 . Assuming that the business platform has eight resources A, B, C, D, E, F, G, and H in total, and the fault set obtained by the controller in the current cycle is: A, C, and H; then the normal cycle of the current cycle The resources are: B, D, E, F, G; if the normal resources included in the flow table obtained by the controller are: B, C, E, F, G; then the resource that failed in the last cycle is: A , D, H.

As shown in FIG. 4 , if the controller determines whether the resource is a faulty resource in step 1, it first processes the normal resources in the current period, that is, processes B, D, E, F, and G. In step 2, it is judged whether the normal resource is a restored isolated resource, only D is a faulty resource in the last cycle, and it is a normal resource in the current cycle, that is, D is the restored isolated resource, for B, E, F, G will not do any processing. For D, proceed to step 3 to judge whether the number of isolated resources at the current moment has reached the scheduling critical value, if the number of resources required for current business processing is 4, and when the utilization rate of the business platform is 100%, the scheduling The critical value is 8-4=4; since the number of isolated resources at the current moment is equal to the number of faulty resources in the previous period, both are 3, which are smaller than the scheduling critical value, the controller proceeds to step 4, Judging whether D is a frequently scheduled resource, if D is determined to be the frequently scheduled resource, D will not be restored; if D is determined not to be the frequently scheduled resource, then start the general scheduling mechanism for D, The general scheduling mechanism refers to changing the status of resources in the previous cycle, that is, restoring isolated resources or isolating normal resources. When continuing step 3 for D to judge whether the scheduling critical value is reached, if the number of resources required for current business processing is 5, then the scheduling critical value is 3, which is equal to the number of currently isolated resources, then the controller Directly enable the general scheduling mechanism for D, that is, restore D.

After the controller finishes processing the normal resources B, D, E, F, and G, it then sequentially processes the faulty resources A, C, and H of the current cycle. If the number of resources required for current business processing is 4, the result of processing the normal resource D is not to restore D. For A, the controller proceeds to step 5, judging whether the number of isolated resources at the current moment reaches the scheduling critical value, since the number of currently isolated resources is 3 less than the scheduling critical value, then proceed to step A 6. Judgment of whether it is a faulty resource that has not been isolated. Since A was a faulty resource in the last cycle, A is not a faulty resource that has not been isolated, and A will not be restored; for C, the currently isolated resource The number is 3 and is less than the scheduling critical value, and then it is judged whether C is a faulty resource that has not been isolated in step 6. Since C was not isolated in the previous cycle and is a faulty resource in the current cycle, C is For the faulty resources that are not isolated, use a general scheduling mechanism for C to isolate C; for H, because C is isolated, the number of isolated resources at the current moment becomes 4 and is equal to the scheduling critical value, The controller activates the protection scheduling mechanism, that is, calculates the number of times that all faulty resources in the service platform are isolated within a preset time period, and prioritizes all the faulty resources according to the number of times. recover the faulty resource, and isolate the faulty resource with high priority; the controller prioritizes A, C, D, and H, and if A has the lowest priority, that is, it is isolated within the preset time interval The number of times is the least, and A is restored first; after A is restored, the number of currently isolated resources becomes 3, which is less than the scheduling critical value 4, which can meet the current business processing requirements. The controller controls C, D, H remains in isolation.

FIG. 5 is a schematic structural diagram of a service platform resource scheduling processing device according to an embodiment of the present invention. As shown in FIG. 5 , the service platform resource scheduling processing device provided by the present invention includes an acquisition unit 501, a receiving unit 502, and a processing unit 503, wherein:

The obtaining unit 501 is used to obtain the work index information of each resource in the service platform, and obtain a resource failure set according to the work index information and the preset health degree index information used to indicate whether the resource fails; wherein, the resource failure The collection includes resources that have failed in the current cycle; the receiving unit 502 is used to obtain a data flow table, and the data flow table includes resources that were in a normal working state in the previous cycle; the processing unit 503 is used to according to the resource failure set, The data flow table, as well as the preset scheduling rules, determine the resources that need to be isolated and/or restored in the current period.

Specifically, the fault collection interface can obtain the work index information of each resource in the service platform, and upload the obtained work index information to the acquisition unit 501, and the fault collection interface can obtain the work index information periodically That is, it is acquired every predetermined time, for example, every 30s. The predetermined time can be set according to the actual situation, which is not limited in this embodiment of the present invention. The work index information is preset from the perspective of business processing, including but not limited to: processor process or thread running status, processor and database connectivity, processor message processing success rate, processor message queue backlog, Proportion of processor exception error codes. The acquiring unit 501 receives the work information index, and compares the work index information of each resource with preset health index information, and if the work index information does not meet the preset health index information condition, then the acquiring unit 501 determines that the resource corresponding to the work index information is a faulty resource. All the faulty resources judged in the current cycle constitute a faulty resource set. The preset health index information corresponds to the work information index, and the condition that the resource is in a normal working state is set, and the resource must meet all the conditions before it is judged to be in a normal working state by the controller State resources, that is, normal resources. For example, the process or thread of the processor is running normally, the connectivity between the processor and the database is normal, and the message processing success rate of the processor is not less than 60%. The work index information and the preset health index information are correspondingly set according to the actual working service platform, which is not limited in this embodiment of the present invention.

The data switch uploads the stored data flow table to the receiving unit 502. The data flow table includes the resources in the normal working state in the previous cycle, hereinafter referred to as normal resources. The receiving unit 502 receives the data uploaded by the data switch. Data flow table.

After obtaining the resource failure set and the data flow table, the processing unit 503 obtains resources that need to be isolated and/or recovered in the current period according to the resource failure set, the data flow table, and a preset scheduling rule. For example, the controller may determine a normal resource that belongs to the resource failure set and appears in the data flow table as a resource that needs to be isolated; it will not belong to the resource failure set and does not appear in all The normal resources in the above data flow table are determined as the resources that need to be restored. For the resources that need to be isolated and/or recovered in the current cycle, the isolation operation will be performed on the resources that need to be isolated in subsequent processing, that is, the resources that need to be isolated that exist in the data flow table will be deleted , performing a recovery operation on the resource that needs to be recovered, that is, adding the resource that does not exist in the data flow table and that needs to be recovered.

The service platform resource scheduling processing device provided by the present invention can obtain the resource fault set by comparing the work index information and the health index information, and obtain the data flow table through the controller, and according to the resource fault set, data flow table and preset The scheduling rules are set to determine the resources that need to be isolated and/or restored, avoiding the use of faulty resources in the resource scheduling process, thereby improving the efficiency of resource scheduling.

Fig. 6 is a schematic structural diagram of a service platform resource scheduling processing device according to another embodiment of the present invention. As shown in Fig. 6, the service platform resource scheduling processing device provided by the present invention further includes:

The sending unit 504 is configured to send a control instruction to the data switch according to the resource that needs to be isolated and/or recovered, so that the data switch updates the data flow table, and sends a message to the data switch according to the updated data flow table. Each resource in the business platform forwards data.

Specifically, after obtaining the resource that needs to be isolated and/or restored, the controller sends a control instruction to the data switch to update the data flow table of the data switch, and the resource that needs to be isolated delete from the data flow table, add resources to be restored to the data flow table, and the data switch forwards data to each resource in the service platform according to the updated data flow table.

The service platform resource scheduling processing device provided by the present invention can obtain the resource fault set by comparing the work index information and the health index information, and obtain the data flow table through the controller, and according to the resource fault set, data flow table and preset The scheduling rules are set to determine the resources that need to be isolated and/or restored, avoiding the use of faulty resources in the resource scheduling process, thereby improving the efficiency of resource scheduling. By updating the data flow table in the data switch, it is ensured that the resources used in the data flow table are all normal resources.

On the basis of the above embodiments, further, the dispatching rules include:

If it is judged that the number of isolated resources at the current moment reaches the scheduling critical value, the recovered isolated resources will be restored;

If it is determined that the number of isolated resources at the current moment has not reached the scheduling threshold, and the recovered isolated resources do not belong to frequently scheduled resources, recover the recovered isolated resources;

If it is determined that the number of isolated resources at the current moment has not reached the scheduling threshold, and the recovered isolated resources belong to frequently scheduled resources, then the recovered isolated resources are not recovered;

Wherein, according to the resource failure set and the data flow table, the recovered isolated resource occurring in the current cycle is determined, and the recovered isolated resource refers to a fault in the previous cycle but in the The resources in the normal working state in the current cycle; the scheduling critical value refers to the maximum number of resources that can be isolated in the service platform under the condition that the current business volume is guaranteed to be processed normally.

Specifically, the controller compares the number of isolated resources at the current moment with the scheduling critical value, and if the number of isolated resources at the current moment is greater than or equal to the scheduling critical value, then the controller The recovered isolated resources are recovered to ensure the processing of traffic. In the current cycle, the number of isolated resources at the current moment may change, and when the controller compares the number of isolated resources at the current moment with the scheduling threshold for the first time, the The number of isolated resources at the current moment is equal to the number of faulty resources in the previous cycle; in the current cycle, if a resource is restored, the corresponding number of isolated resources at the current moment is correspondingly reduced. If a certain resource has been isolated, the number of corresponding resources that have been isolated at the current moment increases accordingly.

The controller compares the number of isolated resources at the current moment with the scheduling threshold, and if the number of isolated resources at the current moment is less than the scheduling threshold, and the recovered isolated resources do not belong to frequent scheduling resource, then the controller restores the restored isolated resource.

The controller compares the number of isolated resources at the current moment with the scheduling threshold, and if the number of isolated resources at the current moment is less than the scheduling threshold, and the recovered isolated resources belong to frequently scheduled resource, the controller does not restore the restored isolated resource.

Wherein, the controller obtains all faulty resources in the current cycle according to the resource fault set, and can determine the normal resources in the current cycle in combination with all resources of the service platform; the controller according to the data flow table , obtain the normal resources in the last cycle, combined with all the resources of the service platform, can determine all the faulty resources in the last cycle; the controller according to the normal resources in the current cycle and the All faulty resources in the current cycle can be determined to determine the recovered isolated resources, and the recovered isolated resources are resources that failed in the previous cycle but are in a normal working state in the current cycle.

The scheduling critical value refers to the maximum number of resources that can be isolated in the service platform under the condition that the current business volume is guaranteed to be processed normally. For example, according to the business volume, the business platform has 5 resources in total. If 3 resources are needed to ensure the normal processing of the current business volume, that is, the current maximum number of resources that can be isolated is 2, and the scheduling critical value Δ=2. The determination of the scheduling critical value is related to the current business volume of the service platform, and it is adjusted adaptively with the change of the current business volume. The greater the current business volume, the more resources are needed, and the fewer resources that can be isolated should be. Therefore, the scheduling The critical value should also be smaller. The scheduling threshold can be determined according to the following formula Among them, N is the current business volume of the business platform, M is the maximum processing business volume of the business platform, S is the total number of resources of the business platform, and k is greater than 0 but less than or equal to 1, which is the utilization rate of the business platform, usually to ensure the normal operation of the business platform , k is less than 1, for example, k is 90%.

The service platform resource scheduling processing device provided by the present invention can obtain the resource fault set by comparing the work index information and the health index information, and obtain the data flow table through the controller, and according to the resource fault set, data flow table and preset The scheduling rules are set to determine the resources that need to be isolated and/or restored, avoiding the use of faulty resources in the resource scheduling process, thereby improving the efficiency of resource scheduling. By giving the restored isolated resource scheduling rule, it is convenient to judge whether to restore the restored isolated resource.

Fig. 7 is a schematic structural diagram of a service platform resource scheduling processing device according to another embodiment of the present invention. As shown in Fig. 7, the service platform resource scheduling processing device provided by the present invention further includes:

The judging unit 505 is configured to obtain the number of times the restored isolated resource is restored within a unit time, and if the number is greater than a threshold, then determine that the restored isolated resource belongs to a frequently scheduled resource; if the number is not greater than The threshold value, then it is determined that the recovered isolated resources do not belong to frequently scheduled resources.

Specifically, the frequently scheduled resource means that the number of times the resource is isolated or restored within a unit time reaches a threshold, and the unit time and the threshold are set according to actual conditions and are not limited in this embodiment of the present invention. The judging unit 505 acquires the number of recovery times of the recovered isolated resource within a unit time, if the number of recovery times is greater than the threshold, it indicates that the recovered isolated resource belongs to the frequently scheduled resource, if the The number of recovery times is not greater than the threshold, indicating that the recovered isolated resource belongs to the frequently scheduled resource. When the resource is frequently restored or isolated, it means that the resource has been judged as a faulty resource many times in a short period of time. Although the resource has been restored many times, if the resource is used, there is a great risk of failure, increasing Instability of resource scheduling, therefore, resources that are judged to be frequently scheduled will not be recovered.

The service platform resource scheduling processing device provided by the present invention can obtain the resource fault set by comparing the work index information and the health index information, and obtain the data flow table through the controller, and according to the resource fault set, data flow table and preset The scheduling rules are set to determine the resources that need to be isolated and/or restored, avoiding the use of faulty resources in the resource scheduling process, thereby improving the efficiency of resource scheduling. By judging frequently scheduled resources, the efficiency of resource scheduling is further improved.

On the basis of the above embodiments, further, the dispatching rules include:

If it is determined that the number of isolated resources at the current moment reaches the scheduling critical value, then calculate the number of times each resource in the service platform is isolated within a preset time period, and prioritize all the faulty resources according to the number of times , recovering faulty resources with low priority, and isolating faulty resources with high priority; wherein, the lower the number of times, the lower the corresponding priority of the faulty resource;

If it is determined that the number of isolated resources at the current moment has not reached the scheduling critical value, and the faulty resource is a faulty resource that has not been isolated, then isolate the faulty resource; wherein, the faulty resource that is not isolated Faulty resources refer to resources that were not isolated in the previous cycle and failed in the current cycle;

Wherein, according to the resource fault set, the faulty resources occurring in the current period are determined; the scheduling critical value refers to the maximum number of resources that can be isolated in the service platform under the condition that the normal processing of the current business volume is ensured.

Specifically, the controller compares the number of isolated resources at the current moment with a scheduling critical value, and if the number of isolated resources at the current moment is greater than or equal to the scheduling critical value, then the faulty resource with a low priority Restoring is performed so that the number of isolated resources is less than the scheduling critical value, ensuring that the number of available resources in the current cycle can meet the normal processing of the current traffic. The low-priority faulty resources are obtained by the controller by calculating the number of times all faulty resources in the service platform are isolated within a preset time period, and prioritizing all faulty resources according to the number of times , the lower the number of times, the lower the priority corresponding to the faulty resource. It can be understood that, after the number of isolated resources is less than the scheduling threshold, unrecovered faulty resources with high priority may be isolated.

The controller compares the number of isolated resources at the current time with the scheduling critical value, and if the number of isolated resources at the current time is less than the scheduling critical value, then check whether the faulty resource is not Judgment of faulty resources that are isolated. The controller obtains the resources in the normal working state in the previous period according to the data flow table, and if the faulty resource appears in the data flow table, then the faulty resource is the resource that has not been isolated, , isolating the faulty resource that has not been isolated. The non-isolated faulty resource refers to a resource that was not isolated in the previous cycle and has a fault in the current cycle.

Wherein, the controller obtains the faulty resources occurring in the current period according to the resource fault set; the scheduling critical value has been explained in the above-mentioned embodiments, and will not be repeated here.

The service platform resource scheduling processing method provided by the present invention can obtain the resource fault set by comparing the work index information and the health index information, and obtain the data flow table through the controller, and according to the resource fault set, data flow table and preset The scheduling rules are set to determine the resources that need to be isolated and/or restored, avoiding the use of faulty resources in the resource scheduling process, thereby improving the efficiency of resource scheduling. By giving fault resource scheduling rules, it is convenient to judge whether to recover or isolate fault resources.

The embodiments of the apparatus provided by the present invention can be specifically used to execute the processing procedures of the above-mentioned method embodiments, and the functions thereof will not be repeated here, and reference can be made to the detailed description of the above-mentioned method embodiments.

FIG. 8 is a schematic diagram of the physical structure of an electronic device according to an embodiment of the present invention. As shown in FIG. 8, the electronic device provided by the present invention includes:

Processor (processor) 801, memory (memory) 802 and communication bus 803;

in,

The processor 801 and the memory 802 complete mutual communication through the communication bus 803;

The processor 801 is used to call the program instructions in the memory 802 to execute the methods provided by the above method embodiments, for example, including: obtaining work index information of each resource in the business platform, and according to the work index information and The preset health index information used to represent whether a resource fails, and obtain a resource failure set; wherein, the resource failure set includes resources that have failed in the current cycle; obtain a data flow table, and the data flow table includes the above Resources that are in a normal working state within a period; resources that need to be isolated and/or recovered in the current period are determined according to the resource failure set, the data flow table, and a preset scheduling rule.

An embodiment of the present invention provides a computer program product, the computer program product includes a computer program stored on a non-transitory computer-readable storage medium, the computer program includes program instructions, and when the program instructions are executed by a computer, The computer can execute the methods provided by the above method embodiments, for example, including: obtaining work index information of each resource in the service platform, and according to the work index information and preset health index information used to indicate whether a resource has failed, Obtain a resource failure set; wherein, the resource failure set includes resources that have failed in the current cycle; obtain a data flow table, and the data flow table includes resources that were in a normal working state in the previous cycle; according to the resource failure set , the data flow table, and a preset scheduling rule to determine resources that need to be isolated and/or restored in the current period.

An embodiment of the present invention provides a non-transitory computer-readable storage medium, the non-transitory computer-readable storage medium stores computer instructions, and the computer instructions cause the computer to execute the methods provided in the above method embodiments, for example Obtain the work index information of each resource in the service platform, and obtain a resource fault set according to the work index information and the preset health index information used to indicate whether the resource fails; wherein, the resource fault set includes Resources that have failed; obtain a data flow table, which includes resources that were in normal working status in the previous cycle; determine the resource failure set, the data flow table, and preset scheduling rules resources that need to be quarantined and/or restored for the current cycle.

Those of ordinary skill in the art can understand that all or part of the steps for realizing the above-mentioned method embodiments can be completed by hardware related to program instructions, and the aforementioned program can be stored in a computer-readable storage medium. When the program is executed, the It includes the steps of the above method embodiments; and the aforementioned storage medium includes: ROM, RAM, magnetic disk or optical disk and other various media that can store program codes.

Finally, it should be noted that: the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that: it can still be Modifications are made to the technical solutions described in the foregoing embodiments, or equivalent replacements are made to some of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the various embodiments of the present invention.

Claims (10)

1. A service platform resource scheduling processing method, characterized in that, comprising: The controller obtains the work index information of each resource in the service platform, and obtains a resource fault set according to the work index information and the preset health index information used to indicate whether the resource fails; wherein, the resource fault set includes the current A resource that failed during the period; The controller obtains a data flow table, and the data flow table includes resources in a normal working state in the previous cycle; The controller determines resources that need to be isolated and/or recovered in the current period according to the resource failure set, the data flow table, and a preset scheduling rule. 2. The method according to claim 1, characterized in that the method further comprises: The controller issues a control instruction to the data switch according to the resource that needs to be isolated and/or recovered, so that the data switch updates the data flow table, and sends the data to the data flow table according to the updated data flow table. Each resource in the service platform forwards data. 3. The method according to claim 1 or 2, wherein the dispatching rule comprises: If it is judged that the number of isolated resources at the current moment reaches the scheduling critical value, the recovered isolated resources will be recovered; If it is determined that the number of isolated resources at the current moment has not reached the scheduling threshold, and the recovered isolated resources do not belong to frequently scheduled resources, recover the recovered isolated resources; If it is determined that the number of isolated resources at the current moment has not reached the scheduling threshold, and the recovered isolated resources belong to frequently scheduled resources, then the recovered isolated resources are not recovered; Wherein, according to the resource failure set and the data flow table, the recovered isolated resource occurring in the current cycle is determined, and the recovered isolated resource refers to a fault in the previous cycle but in the The resources in the normal working state in the current cycle; the scheduling critical value refers to the maximum number of resources that can be isolated in the service platform under the condition that the current business volume is guaranteed to be processed normally. 4. method according to claim 3, is characterized in that, described method also comprises: Obtain the number of recovery times of the recovered isolated resource within a unit time, if the number of times is greater than a threshold, determine that the recovered isolated resource belongs to a frequently scheduled resource; if the number of times is not greater than the threshold, then It is determined that the recovered isolated resource does not belong to a frequently scheduled resource. 5. The method according to claim 1 or 2, wherein the dispatching rule comprises: If it is determined that the number of isolated resources at the current moment reaches the scheduling critical value, then calculate the number of times that all faulty resources in the service platform are isolated within a preset time period, and prioritize all faulty resources according to the number of times sorting, recovering faulty resources with low priority, and isolating faulty resources with high priority; wherein, the lower the number of times, the lower the corresponding priority of the faulty resources; If it is determined that the number of isolated resources at the current moment has not reached the scheduling critical value, and the faulty resource is a faulty resource that has not been isolated, then isolate the faulty resource; wherein, the faulty resource that is not isolated Faulty resources refer to resources that were not isolated in the previous cycle and failed in the current cycle; Wherein, according to the resource fault set, determine the fault resource that occurs in the current cycle; the scheduling threshold refers to the maximum isolable resource in the service platform under the condition that the current business volume is guaranteed to be processed normally number. 6. A service platform resource scheduling processing device, characterized in that it comprises: The obtaining unit is used to obtain the work index information of each resource in the service platform, and obtain a resource fault set according to the work index information and the preset health index information used to indicate whether the resource is faulty; wherein, the resource fault The pool includes resources that failed during the current cycle; a receiving unit, configured to obtain a data flow table, the data flow table including resources in a normal working state in the previous cycle; A processing unit, configured to determine resources that need to be isolated and/or recovered in the current period according to the resource failure set, the data flow table, and a preset scheduling rule. 7. The device according to claim 6, further comprising: The sending unit is configured to send a control instruction to the data switch according to the resource that needs to be isolated and/or restored, so that the data switch updates the data flow table, and sends the data to the data flow table according to the updated data flow table. Each resource in the business platform forwards data. 8. The device according to claim 6 or 7, wherein the scheduling rule comprises: If it is judged that the number of isolated resources at the current moment reaches the scheduling critical value, the recovered isolated resources will be restored; If it is determined that the number of isolated resources at the current moment has not reached the scheduling threshold, and the recovered isolated resources do not belong to frequently scheduled resources, recover the recovered isolated resources; If it is determined that the number of isolated resources at the current moment has not reached the scheduling threshold, and the recovered isolated resources belong to frequently scheduled resources, then the recovered isolated resources are not recovered; Wherein, according to the resource failure set and the data flow table, the recovered isolated resource occurring in the current cycle is determined, and the recovered isolated resource refers to a fault in the previous cycle but in the The resources in the normal working state in the current cycle; the scheduling critical value refers to the maximum number of resources that can be isolated in the service platform under the condition that the current business volume is guaranteed to be processed normally. 9. The device according to claim 8, further comprising: A judging unit, configured to obtain the number of times the restored isolated resource is restored within a unit time, and if the number is greater than a threshold, determine that the restored isolated resource belongs to a frequently scheduled resource; if the number is not greater than The threshold value, then it is determined that the recovered isolated resources do not belong to frequently scheduled resources. 10. The method according to claim 6 or 7, wherein the dispatching rule comprises: If it is determined that the number of isolated resources at the current moment reaches the scheduling critical value, then calculate the number of times that all faulty resources in the service platform are isolated within a preset time period, and prioritize all faulty resources according to the number of times sorting, recovering faulty resources with low priority, and isolating faulty resources with high priority; wherein, the lower the number of times, the lower the corresponding priority of the faulty resources; If it is determined that the number of isolated resources at the current moment has not reached the scheduling critical value, and the faulty resource is a faulty resource that has not been isolated, then isolate the faulty resource; wherein, the faulty resource that is not isolated Faulty resources refer to resources that were not isolated in the previous cycle and failed in the current cycle; Wherein, according to the resource fault set, the fault resources occurring in the current cycle are determined; the scheduling critical value refers to the maximum number of resources that can be isolated in the service platform under the condition of ensuring normal processing of the current traffic.