TWI676374B - Server and method for assigning job for microservices thereof - Google Patents

Abstract

The invention provides a server and a method for arranging work for a microservice. The method includes: obtaining a service request; selecting a first microservice from a plurality of microservices, wherein the first microservice includes a plurality of processing programs; assigning a service request to a first processing program of the foregoing processing programs; The microservice applies a first global lock and applies a first region lock to a first processing program. The first global lock time of the first global lock is the same as the first region lock time of the first regional lock. The second micro service stopped requesting to process the service request because of the first global lock, and the second processing program in the foregoing processing program stopped requesting to process the service request because of the first zone lock.

Description

Server and method for arranging work for microservices

The invention relates to a server and a method for arranging work for microservices, and in particular, to a method and a server for arranging work for microservices based on global locks and area locks of microservices.

Microservices is a software architecture style. It is based on small building blocks that focus on a single responsibility and function. It uses a modular approach to compose complex large-scale applications. Each functional area Blocks communicate with each other using a language-independent / language agnostic API set.

However, after the cloud system begins to introduce the microservice architecture and design in a stateless manner, it may encounter problems with microservice expansion and availability between homogeneous microservices. And, since microservices also include two availability modes, Active / Active and Active / Standby, how to make the overall processing performance, availability, and service quality of microservices through proper design Being able to improve is indeed an important issue of concern to those skilled in the art.

In view of this, the present invention provides a method for arranging work for a microservice, which can be used to improve the availability and performance of a microservice system.

The invention provides a method for arranging work for a microservice, including: obtaining a service request; selecting a first microservice from a plurality of microservices, wherein the first microservice includes a plurality of processing programs; and assigning a service request to the aforementioned processing A first processing program in the program; applying a first global lock to the first microservice, and applying a first region lock to the first processing program, wherein a first global lock time of the first global lock is the same as the first A first area lock time of the area lock, a second micro service in the foregoing micro service stops requesting to process a service request due to the first global lock, and a second processing program in the foregoing processing program corresponds to the first area lock And stop asking to process service requests.

Based on the above, the server and method for arranging work of the microservice provided by the present invention can lock the service request to the first microservice through global lock after selecting the first microservice for processing the service request, and through the area Lock to lock this service request to the first handler in the first microservice. In this way, other microservices and processing programs can be prevented from spending extra time in repeatedly sending requests for processing service requests, thereby improving the working efficiency of the microservice system architecture.

In order to make the above features and advantages of the present invention more comprehensible, embodiments are hereinafter described in detail with reference to the accompanying drawings.

Please refer to FIG. 1 and FIG. 2, where FIG. 1 is a functional block diagram of a server according to an embodiment of the present invention, and FIG. 2 is a microservice system architecture diagram according to FIG. 1. In FIG. 1, the server 100 includes a storage circuit 102 and a processor 104. The storage circuit 102 is, for example, a memory, a hard disk, or any other component that can be used to store data, and can be used to record multiple codes or modules. The processor 104 is coupled to the storage circuit 102 and may be a general-purpose processor, a special-purpose processor, a conventional processor, a digital signal processor, a plurality of microprocessors, and one or more integrated digital signal processors. Microprocessor core, controller, microcontroller, Application Specific Integrated Circuit (ASIC), Field Programmable Gate Array (FPGA), any other kind of integrated circuit, State machines, processors based on Advanced Reduced Instruction Set (ARM) machines, and the like.

The server 100 of this embodiment may be used to construct a microservice system architecture 200 as shown in FIG. 2, which includes a request queue 210, a microservice management terminal 220, and a microservice 231, 232, 233,..., 23n. In one embodiment, the processor 104 may establish multiple virtual machines (VMs) on the server 100, and use one of them as the microservice management end 220, and the remaining VMs as the microservices 231 ~ 23n. However, the present invention is not limited to this.

As shown in FIG. 2, when receiving a service request RQ from a user, the processor 104 may store the service request RQ in the request queue 210 accordingly. In an embodiment, the service request RQ is, for example, a request to build a new VM, but it is not limited thereto. The request queue 210 may be a stacked structure that can be used to serially store different service requests from one or more users.

In an embodiment, the processor 104 may allocate the service request to the microservices 231 to 23n through the microservice management terminal 220, so that the processing programs in the microservices 231 to 23n assist in completing the work corresponding to the service request. The following will be supplemented by FIG. 3 for further explanation.

Please refer to FIG. 3, which is a flowchart of a method for arranging work for a microservice according to FIG. 2. The method of this embodiment may be executed by the architecture shown in FIG. 2. The details of the steps of the method are described below with reference to the elements of FIG. 2.

First, the micro service management terminal 220 can obtain the service request RQ1. It should be understood that although the microservice management terminal 220 in FIG. 2 obtains the service request RQ1 from the request queue 210, in other embodiments, the microservice management terminal 220 may also obtain the service directly from the user and request RQ1, but The present invention may not be limited to this.

In step S220, the microservice management terminal 220 may select a first microservice from a plurality of microservices 231 to 23n, where the first microservice is, for example, an idle microservice, that is, a microservice to which no work is assigned. In different embodiments, the manner in which the microservice management terminal 220 selects the first microservice may vary according to the mode in which the microservice system architecture operates, as shown below.

Please refer to FIG. 4A, which is a schematic diagram of selecting a first microservice in an active / active mode according to an embodiment of the present invention. In FIG. 4A, the microservice system architecture 200 is assumed to operate in an active / active mode, and the microservice management terminal 220 is used to manage six microservices 231-236. In the active / active mode, the microservice management end 200 can divide the microservices 231 to 236 into multiple groups GP1, GP2, and GP3. In this embodiment, the microservice management end 220 may first sort the microservices 231-236 according to the global availability score of each microservice 231-236, and then sort the sorted microservices 231-236 into a preset number. Groups. The relevant computer system for the global availability scores of each microservice 231 ~ 236 will be detailed in the following pages.

For ease of illustration, in FIG. 4A, it is assumed that the global availability scores of microservices 231 to 236 are from high to low, and the preset number is 3, so the microservices included in groups GP1, GP2, and GP3 are shown in Figure 4A. It is shown, but it is not intended to limit the possible embodiments of the present invention. Based on the above assumptions, the global availability scores of microservices 231 and 232 in group GP1 are higher than the global availability scores of microservices 233 and 234 in group GP2, and the global availability of microservices 233 and 234 in group GP2 The availability scores are higher than the global availability scores of microservices 235 and 236 in group GP3. Correspondingly, the micro service management terminal 220 can set the priorities of the groups GP1 to GP3 as high, medium, and low respectively, and select the first micro service for processing the service request RQ1 according to the priorities of the groups GP1 to GP3. Right to pick.

Since the microservice 231 is located in the group GP1 and has the highest global availability score, the microservice management end 220 can select the microservice 231 as the microservice (ie, the first microservice) for processing the service request RQ1.

Please refer to FIG. 4B, which is a schematic diagram of selecting a first microservice in an active / standby mode according to an embodiment of the present invention. The difference from FIG. 4A is that the microservice system architecture 200 in FIG. 4B is assumed to operate in an active / standby mode. In the active / standby mode, the microservice management terminal 200 can first sort the microservices 231-236 according to the global availability scores of each microservice 231-236, and then sort the microservices 231-236, and then sort the sorted microservices. 231 ~ 236 are divided into a master group GP1 'and a slave group GP2'. In this embodiment, it is assumed that microservices 231 and 232 are allocated to the main group GP1 ', and microservices 233 to 236 are allocated to the slave group GP2'. In this case, when the micro service management terminal 200 wants to select the first micro service for processing the service request RQ1, the micro service management terminal 200 will only perform the round robin principle in the main group GP1 '. Pick. In addition, even if the microservices 231 and 232 in the main group GP1 'have been used to process other service requests, the microservice management terminal 220 will still not select the microservices 234 ~ 236 in the slave group GP2' to process services. Request RQ1.

In FIG. 4B, since the microservice 231 is located in the main group GP1 ', the microservice management terminal 220 can select the microservice 231 as the microservice (ie, the first microservice) for processing the service request RQ1.

Please refer to FIG. 3 again. After step S320 is performed to select the first microservice, the microservice management terminal 220 may assign the service request RQ1 to the first processing program in the first microservice.

Moreover, in step S330, a first global lock may be applied to the first microservice, and a first region lock may be applied to the first processing program. The first global lock time of the first global lock is the same as that of the first region lock. A zone lock time. In addition, the second microservice in microservices 231 to 236 will stop requesting to process the service request RQ1 due to the first global lock, and the second processing program in the first microservice will also stop requesting due to the first regional lock. Process service request RQ1.

In one embodiment, the micro service management terminal 220 may record the related information of the first global lock and the first global lock time in a global lock data table maintained by the micro service management terminal 220. Similarly, the first microservice may also record information about the first region lock and the first region lock time in the region lock cache table maintained by the first microservice. In order to make the above concepts clearer, a further description is given below with reference to FIG. 5.

Please refer to FIG. 5, which is a schematic diagram of allocating work for microservices according to an embodiment of the present invention. In this embodiment, it is assumed that the microservice 231 is selected for processing the service request RQ1. Correspondingly, the microservice management end 220 may apply a global lock GL1 to the microservice 231 for processing the service request RQ1, and record the correspondence between the microservice 231 and the service request RQ1, and the global lock time GT1 of the global lock GL1 in the micro The global lock data table 220a maintained by the service management terminal 220.

After the microservice 231 learns that it is assigned to process the service request RQ1, the microservice 231 may assign a handler 231a (eg, a thread) among a plurality of handlers included in it to process the service request RQ1. Correspondingly, the microservice 231 may apply the area lock LL1 to the processing program 231a, and record the correspondence between the processing program 231a and the service request RQ1, and the area lock time LT1 of the area lock LL1 in the area lock information table 231b maintained by the microservice 231. in.

In addition, it is also assumed that the microservice 232 is selected to process the service request RQ2. Correspondingly, the microservice management terminal 220 may apply the global lock GL2 to the microservice 232 for processing the service request RQ2, and record the correspondence between the microservice 232 and the service request RQ2, and the global lock time GT2 of the global lock GL2 in the micro The global lock data table 220a maintained by the service management terminal 220.

After the microservice 232 learns that it is assigned to process the service request RQ2, the microservice 232 may assign a handler 232a (eg, a thread) of a plurality of handlers included in it to process the service request RQ2. Accordingly, the microservice 232 can apply the area lock LL2 to the processing program 232a, and record the correspondence between the processing program 232a and the service request RQ2, and the area lock time LT2 of the area lock LL2 in the area lock data table 232b maintained by the microservice 232. in.

Based on the content recorded in the global lock data table 220a, other microservices managed by the microservice management terminal 220 (for example, microservices 233 to 236) can learn that service requests RQ1 and RQ2 have been locked to microservices 231 and 232, respectively. The request for processing the service request RQ1 or RQ2 will not be repeatedly submitted to the microservice management end 220. Thereby, the working efficiency of the microservice system architecture 200 can be improved.

At the same time, based on the content recorded in the area lock data table 231b, other processing programs in the microservice 231 can know that the service request RQ1 has been locked to the processing program 231a, so the request for processing the service request RQ1 will not be repeated. Similarly, based on the content recorded in the area lock data table 232b, other processing programs in the microservice 232 can know that the service request RQ2 has been locked to the processing program 232a, so the request for processing the service request RQ2 will not be repeated. This can improve the working efficiency of microservices 231 and 232.

It can be known from the above that the server and the method for arranging work of the microservice provided by the present invention can select the first microservice (which needs to be a microservice that is not locked by the global lock) for processing the service request, and then use the global lock The service request is locked to the first microservice, and the service request is locked to the first handler in the first microservice through an area lock. In this way, other microservices and processing programs can be prevented from spending extra time in repeatedly sending requests for processing service requests, thereby improving the working efficiency of the microservice system architecture.

In one embodiment, the processor 104 may use docker technology to implement the microservice system architecture 200, and use spring boot to develop and provide web services. The microservice management terminal 220 can implement the work distribution for microservices 231 ~ 23n based on Eureka, and the request queue 210 can be implemented by microservices 231 ~ 23n combined with Redis, and is realized asynchronously by a schedule thread. Work processing. In addition, the microservice management end 220 may implement a Semaphore lock mechanism by Redisson to implement the first global lock applied to the first microservice. and. The first microservice may use a concurrent hash map and a valid time stamp to implement the first region lock. The global lock data table maintained by the microservice management terminal 220 can be implemented by setting up mongodb.

In the embodiment of FIG. 5, if the microservice 231 cannot complete the service request RQ1 at the global lock time GL1 or the regional lock time LL1, the microservice management terminal 220 may reduce the global service score of the microservice 231 and send the outstanding service request RQ1 Queued to request queue 210. In this way, the service request RQ1 can still have the opportunity to be processed by other microservices in the future, and then complete the work required by the user.

In addition, in the embodiment of FIG. 5, when the microservice management terminal 220 ranks the outstanding service request RQ1 into the request queue 210, the microservice management terminal 220 may also record the processing progress of the microservice 231 for the service request RQ1. After that, when the service request RQ1 is again assigned to another microservice for processing, the another microservice can continue to process the service request RQ1 according to the processing progress of the service request RQ1. This can speed up the processing of the service request RQ1, and then complete the work required by the user with better efficiency.

In the embodiment of FIG. 5, if the microservice 231 completes the service request RQ1 before the area lock time LT1 expires, the microservice 231 may release the area lock LL1 and report the processing result of the service request RQ1 to the microservice management end 220. Correspondingly, the microservice management terminal 220 may release the global lock GL1 according to the processing result of the service request RQ1 returned by the microservice 231. Thereby, the micro service 231 can be selected again for processing other service requests, so as to better utilize the resources of the micro service system architecture 200.

As mentioned in the previous embodiment, each of the microservices 231 to 236 has a global availability score. The computer system of the global availability scores of each of the microservices 231 to 236 will be described below with reference to FIGS. 6A and 6B.

Please refer to FIG. 6A and FIG. 6B, wherein FIG. 6A is a schematic diagram of adjusting a regional availability score according to an embodiment of the present invention, and FIG. 6B is a schematic diagram of adjusting a global availability score according to FIG. 6A. For convenience of explanation, only microservices 231 and 232 of FIG. 5 are taken as examples in this embodiment, but the method of the present invention can also be applied to other microservices (for example, microservices 233 to 236 of FIG. 5).

In FIG. 6A, the microservice management terminal 220 may send an inquiry signal QS to the microservices 231 and 232. In this embodiment, the microservice management end 220 may add a sending timestamp to the query signal QS (which is, for example, a heartbeat signal implemented by an HTTP / HTTPS service request signal), to indicate where the query signal QS is. The point in time is sent.

After the microservice 231 receives the query signal QS, it may return another HTTP / HTTPS service request signal, which may carry a return time stamp. In addition, when the microservice 231 returns the above-mentioned another HTTP / HTTPS service request signal, a processing time (for example, 10ms shown in FIG. 6A) for the inquiry signal QS may be calculated, where the processing time may be returned by, for example, The time stamp is calculated by subtracting the transmission time stamp, but it is not limited thereto. Similarly, the microservice 232 can also calculate the processing time of the microservice 232 for the query signal QS (for example, 66ms shown in FIG. 6A) based on the above teaching.

In different embodiments, the microservices 231 and 232 may update the regional availability scores according to their respective processing times. Taking FIG. 6A as an example, since the processing time (ie, 10 ms) of the micro service 231 does not exceed a preset time (for example, 20 ms), the micro service 231 can increase the regional availability score of the micro service 231 (for example, add 10 points). The area availability score of the increased microservice 231 is, for example, 56 points, but it is not limited thereto. In an embodiment, the 56 points illustrated in FIG. 6A may be points that the microservice 231 accumulates after receiving a plurality of query signals according to a plurality of processing times corresponding to the query signals.

On the other hand, since the processing time (ie, 66 ms) of the micro service 232 exceeds a preset time (eg, 20 ms), the micro service 232 may reduce the regional availability score (eg, minus 5 points) of the micro service 232. The area availability score of the reduced microservice 232 is, for example, 80 points, but it is not limited thereto. In an embodiment, the 80 points illustrated in FIG. 6A may be points that the microservice 232 accumulates after receiving multiple query signals according to multiple processing times corresponding to the query signals.

In other embodiments, the designer may also use other mechanisms to increase or decrease the regional availability scores of the microservices 231 and 231 according to requirements, and is not limited to the mechanism described in the above teaching.

After that, the microservice management terminal 220 may request the microservice 231 to return the difference between its regional availability score and the regional preset score, and adjust the global availability score of the microservice 231 accordingly, which will be described in detail below with reference to FIG. 6B.

In FIG. 6B, the microservice management terminal 220 may record the global availability score of the microservice 231 in the global availability score data table 220b maintained by the microservice management terminal 220. As shown in FIG. 6B, the microservice management terminal 220 may request the microservice 231 to return the difference between the regional availability score (ie, 56 points) and the regional preset score (ie, 300 points) (ie, -244 points). . Next, assuming that the original global availability score of the microservice 231 is 700 points shown in FIG. 6B, the microservice management terminal 220 can use the sum of the global availability score and the difference value as the updated global availability score (ie, 700 points-244 points = 456 points). Correspondingly, the micro service 231 may supplement its own regional availability score from 56 points to the regional preset score (ie, 300 points), which is used as a subsequent adjustment of the amount of the regional availability score according to the above processing time. With this, the microservice management terminal 220 can more easily control the global availability score of the microservice 231.

In an embodiment, after the microservice 231 returns its regional availability score to 300 points, it can be used as a basis for adding or subtracting the global availability score to other query signals subsequently sent by the microservice management terminal 220. This can achieve the effect of decentralized management.

In another embodiment, if the microservice management terminal 220 does not receive the above-mentioned difference value returned by the microservice 231 within a preset time, it means that the performance of the microservice 231 may be extremely poor, so the microservice management terminal 220 may directly subtract a global score of the microservice 231 from a preset score (for example, 600 points) to characterize the low working efficiency of the microservice 231.

In addition, the embodiment of the present invention further proposes a mechanism for timely eliminating poorly performing microservices according to the performance of the microservices, which will be further described below with reference to FIGS. 7A and 7B.

Please refer to FIG. 7A, which illustrates a microservice elimination mechanism in an active / active mode according to an embodiment of the present invention. In this embodiment, it is assumed that the global availability scores of the microservices 231 to 233 recorded by the global availability score data table 220b maintained by the microservice management terminal 220 are 456 points, 1100 points, and 30 points, respectively. Following the teachings of the previous embodiment, since the global availability score of the microservice has a negative correlation with the processing time of the microservice processing the query signal QS, a higher global availability score represents a shorter processing time, that is, a higher work efficiency.

In FIG. 7A, it is assumed that the microservice management terminal 220 can distinguish the microservices 232 with a global availability score greater than 1000 points into the group GP1, and the microservices 231 with the global availability score between 1000 and 300 points into the group GP2. And the micro service 233 with a global availability score of less than 300 points is divided into the group GP3. Since higher global availability scores indicate higher work efficiency, the groups GP1 ~ GP3 can also be regarded as corresponding to high performance, medium performance, and low performance, respectively.

Then, the microservice management terminal 220 can determine whether the microservice 233 classified into the low-performance group GP3 has been classified into the group GP3 for a preset period. If it is, it means that the microservice 233 has been in low working efficiency for a long time, so the microservice management end 220 can decide to eliminate the microservice 233; if not, it means that the microservice 233 has not been in low working efficiency for a long time, so the microservice management end 220 It may be determined that microservices 233 do not need to be eliminated.

In one embodiment, if the microservice management terminal 220 determines that the microservice 233 needs to be eliminated, the microservice management terminal 220 may delete the microservice 233 and add a new microservice to replace the microservice 233.

Please refer to FIG. 7B, which illustrates a microservice elimination mechanism in an active / standby mode according to an embodiment of the present invention. 7A differs from FIG. 7A in that the embodiment of FIG. 7A can classify microservices 232 with higher working performance into the main group GP1 ′, and classify microservices 231 and 233 with poor working performance into the slave group GP2 '.

Then, the micro service management terminal 220 may determine whether the micro service 231 (or 233) classified into the slave group GP2 'having low performance has been classified into the slave group GP2' for a preset period. If it is, it means that the microservice 231 (or 233) has been in a low working efficiency for a long time, so the microservice management end 220 can determine that the microservice 231 (or 233) is eliminated; if not, it means that the microservice 231 (or 233) has not been Due to low working efficiency for a long time, the microservice management terminal 220 may determine that the microservice 231 (or 233) does not need to be eliminated.

In an embodiment, if the microservice management terminal 220 determines that the microservice 231 (or 233) needs to be eliminated, the microservice management terminal 220 may delete the microservice 231 (or 233) and add a new microservice to replace it. Microservices 231 (or 233).

In summary, the servo and the method for arranging work for the microservices provided by the present invention can determine the first microservice for processing service requests according to the global availability score of each microservice, and establish corresponding global locks and area locks. To make the built microservices system architecture highly available. In addition, the global lock and area lock also have corresponding global lock time and area lock time, and when one of them expires, the global lock and area lock will be invalidated at the same time, so that the microservice management can re-request for the service. Arrange other microservices for processing. In this way, when microservices are horizontally expanded, related logical services can maintain high availability, and can be applied to two high availability modes such as active / active and active / standby.

In addition, the present invention can also determine the global availability score of the microservice based on the processing time of the microservice for the interrogation signal, as a basis for assigning a service request. In addition, the above-mentioned global availability score can also be used as a basis for whether to immediately eliminate this microservice, so that the overall microservice can maintain high processing performance.

Although the present invention has been disclosed as above with the examples, it is not intended to limit the present invention. Any person with ordinary knowledge in the technical field can make some modifications and retouching without departing from the spirit and scope of the present invention. The protection scope of the present invention shall be determined by the scope of the attached patent application.

100‧‧‧Server

102‧‧‧Storage Circuit

104‧‧‧Processor

200‧‧‧microservice system architecture

210‧‧‧ Request queue

220‧‧‧Micro-service management terminal

220a‧‧‧Global Lock Data Sheet

220b‧‧‧Global Availability Score Data Sheet

231, 232, 233, ..., 23n‧‧‧ microservices

231a, 232a‧‧‧Process

231b, 232b‧‧‧ Zone Lock Data Sheet

GL1, GL2‧‧‧Global lock

GT1, GT2‧‧‧Global lock time

GP1, GP2, GP3‧‧‧ group

GP1’‧‧‧Main group

GP2’‧‧‧ servant group

LL1, LL2‧‧‧area lock

LT1, LT2‧‧‧ Zone lock time

RQ, RQ1, RQ2‧‧‧ Service Request

S310 ~ S340‧‧‧step

QS‧‧‧Inquiry signal

FIG. 1 is a functional block diagram of a server according to an embodiment of the present invention. FIG. 2 is a microservice system architecture diagram according to FIG. 1. FIG. 3 is a flowchart of a method for arranging work for a microservice according to FIG. 2. FIG. 4A is a schematic diagram of selecting a first microservice in an active / active mode according to an embodiment of the present invention. FIG. 4B is a schematic diagram of selecting a first microservice in an active / standby mode according to an embodiment of the present invention. FIG. 5 is a schematic diagram illustrating a microservice allocation task according to an embodiment of the present invention. FIG. 6A is a schematic diagram of adjusting an area availability score according to an embodiment of the present invention. FIG. 6B is a schematic diagram of adjusting the global availability score according to FIG. 6A. FIG. 7A illustrates a microservice elimination mechanism in an active / active mode according to an embodiment of the present invention. FIG. 7B illustrates a microservice elimination mechanism in an active / standby mode according to an embodiment of the present invention.

Claims (13)

A method for arranging work for a microservice includes: obtaining a service request; selecting a first microservice from a plurality of microservices, wherein the first microservice includes a plurality of processing programs; and assigning the service request to the processings A first processing program in the program; applying a first global lock to the first microservice and applying a first region lock to the first processing program, wherein a first global lock time of the first global lock is the same At a first area lock time of the first area lock, a second micro service of the micro services stops requesting to process the service request due to the first global lock, and a second process of the processing programs The program stops requesting to process the service request due to the first region lock, and if the first microservice cannot complete the service request within the first global lock time or the first region lock time, the method further includes reducing the A first global availability score for the first microservice, and queue outstanding service requests into a request queue. The method according to item 1 of the scope of patent application, further comprising: recording a processing progress of the service request outstanding; and assigning the service request in the request queue to one of the microservices, In order to make one of the microservices continue to process the service request according to the processing progress. The method according to item 1 of the scope of patent application, further comprising: sending an inquiry signal to a third micro service of the micro services, so that the third micro service calculates a processing time for the inquiry signal, and according to The processing time updates a region availability score of the third microservice; the third microservice is required to return a difference between the region availability score and a region preset score, and adjust the third microservice's A global availability score; judging whether to replace the third microservice based on the global availability score of the third microservice; if so, delete the third microservice and add a new microservice to the microservices; and if not , Retain the third microservice in the microservices. The method according to item 3 of the scope of patent application, wherein: when the processing time for processing the interrogation signal by the third microservice exceeds a preset time, the third microservice reduces the area availability score; and when the When the processing time for the three microservices to process the interrogation signal does not exceed the preset time, the third microservice increases the area availability score. The method according to item 4 of the scope of patent application, wherein the third microservice is required to return the difference between the regional availability score and the regional preset score, and adjust the global availability score of the third microservice accordingly. The step includes: using a sum of the global availability score and the difference as the updated global availability score. The method according to item 4 of the scope of patent application, wherein if the difference value returned by the third microservice is not received within a preset time, the method further includes the global availability score of the third microservice Subtract a preset score. The method described in claim 3, wherein the microservices each have a global availability score, and the step of determining whether to replace the third microservice based on the global availability score of the third microservice includes: The global availability score of the microservices ranks the microservices in descending order; the sorted microservices are divided into multiple groups, where the groups include at least a first group and a second group, and are The global availability score of each microservice classified to the second group is lower than the global availability score of each microservice classified to the first group; judging whether the third microservice has been classified Class to the second group for a preset period; if yes, determine to replace the third microservice; and if not, determine not to replace the third microservice. The method according to item 7 of the scope of patent application, wherein if the microservices operate based on an active / standby mode, the first group is a main group and the second group is a slave group. The method of claim 1, wherein if the first microservice completes the service request before the first region lock time expires, the first microservice releases the first region lock and returns the service A requested processing result. The method according to item 6 of the scope of patent application, further comprising releasing the first global lock in response to the processing result returned by the first microservice. The method of claim 1, wherein if the microservices operate based on an active / active mode, the first global availability score of the first microservice is the highest among the microservices. The method according to item 1 of the scope of patent application, wherein each microservice has a global availability score, and if the microservices operate based on an active / standby mode, the microservices are based on the global availability of each microservice The score is divided into a master group and a slave group, wherein the global availability score of each microservice in the master group is higher than the global availability score of each microservice in the slave group, and from The step of selecting the first microservice among the microservices includes: selecting the first microservice only from the microservices located in the main group. A server includes: a storage circuit storing a plurality of modules; and a processor coupled to the storage circuit and accessing the modules to perform the following steps: obtaining a service request; selecting from a plurality of microservices A first microservice, wherein the first microservice includes a plurality of processing programs; assigning the service request to a first processing program among the processing programs; applying a first global lock to the first microservice, and A first region lock is applied to the first processing program, wherein a first global lock time of the first global lock is the same as a first region lock time of the first regional lock, and a second micro of the micro services The service stops requesting to process the service request due to the first global lock, and a second processing program among the processing programs stops requesting to process the service request due to the first regional lock. If the first microservice The service request cannot be completed within the first global lock time or the first regional lock time, the processor is further configured to reduce a first global availability score of the first microservice, and The completion of the service request into a request queue.