WO2014194869A1

WO2014194869A1 - Request processing method, device and system

Info

Publication number: WO2014194869A1
Application number: PCT/CN2014/079489
Authority: WO
Inventors: 马久跃; 包云岗; 隋秀峰; 任睿
Original assignee: 华为技术有限公司
Priority date: 2013-06-08
Filing date: 2014-06-09
Publication date: 2014-12-11
Also published as: CN104243405B; CN104243405A

Abstract

Disclosed is a request processing method, which is used for dynamically adjusting the execution priority and resource allocation of a request according to the processing time of the request. The method in the embodiments of the present invention comprises: receiving a remote procedure call (RPC) request sent by a service allocation server, and adding the RPC request to a service request queue; and according to the response time restraint information about each RPC request in the service request queue, setting the execution priority of each of the RPC requests, and/or allocating an execution resource of each of the RPC requests.

Description

The present invention claims the Chinese patent application filed on June 8, 2013, the Chinese Patent Application No. 201310228246. X, the Chinese patent application entitled "A Request Processing Method, Apparatus and System" Priority is hereby incorporated by reference in its entirety. Technical field

The present invention relates to the field of communications, and in particular, to a request processing method, apparatus, and system.

Background technique

Internet applications such as e-mail, search, online shopping, social networking, online video, web maps, etc., have become part of people's lives. These applications often serve hundreds of millions of users, meaning that Internet applications have become a social public service, and data centers that support Internet applications with massive users have become the core infrastructure of society.

The number of active users and user visits are the main factors affecting Internet company revenue. Fast service response time is the key to satisfying users and retaining users. Internet companies generally use free services to attract users, so response time is a key metric for measuring Quality of Service (QoS). Because Internet applications need to provide services to hundreds of millions of users at the same time, for performance and scalability reasons, most of them are implemented in a distributed manner, and an application is decomposed into many services deployed on multiple service processing servers, so one User requests are ultimately assigned to multiple different business processing servers for processing. Sequential/Dependent Mode A typical service aggregation model in which the output of the previous stage service is the input to the next stage of service, and the services of the adjacent two stages have dependencies.

Affected by various factors such as request characteristics, network, and background operations in the data center, the response time of each service phase will fluctuate constantly. At the same time, Internet companies usually use data center operations in pursuit of low-cost targets. The resource sharing method improves resource utilization. The more serious of this sharing is the delay of the response time of the previous stage service in the service processing server, which may be further amplified in the next stage of service processing, in sequential/dependent services. In aggregation mode, each level The response time delays are superimposed step by step, so that there is a large delay in the response time requested by the end user. This response time delay can severely impact the quality of service for many delay-sensitive applications such as search, online shopping, and more. In the prior art, before deploying a shared application, a technician pre-analyzes and tests a large number of existing applications, and selects an application with minimal interference to perform hybrid deployment, thereby reducing the influence of mutual interference between applications on response time. Avoid delays in response time as much as possible. However, this prior art requires pre-analysis and testing of a large number of existing applications before deployment, which is difficult to operate, and, due to the variety and uncertainty of existing applications, the application is analyzed before deployment. It can only reduce the mutual interference between some applications, and there are still large interferences between many applications, and the effect of avoiding response time delay is not good. Summary of the invention

An embodiment of the present invention provides a request processing method for dynamically adjusting a request execution priority and resource allocation according to a requested processing time. The request processing method provided by the first aspect of the embodiments of the present invention includes:

The service processing server receives a remote procedure call RPC request sent by the service distribution server,

The RPC request includes: response time constraint information; the response time constraint information is used to mark a processing constraint time of the user request corresponding to the RPC request and a processing time has occurred; the service processing server adds the RPC request to the service request a queue; the service processing server sets an execution priority of the respective RPC requests according to response time constraint information of each RPC request in the service request queue, and/or allocates execution resources of the respective RPC requests. In a first possible implementation of the first aspect, if the service processing server is capable of simultaneously processing all RPC requests in the service request queue, the service processing server is configured according to each RPC request in the service request queue. Response time constraint information is allocated to the execution resources of the respective RPC requests; if the service processing server can only process any one of the service request queues separately

The RPC requests, the service processing server sets the execution priority of each RPC request according to the response time constraint information of each RPC request in the service request queue; If the service processing server cannot process all the RPC requests in the service request queue at the same time, or does not only process any one of the RPC requests in the service request queue, the service processing server according to the service request The response time constraint information of each RPC request in the queue sets the execution priority of each RPC request, and allocates the execution resources of the respective RPC requests. With reference to the first aspect or the first implementation method of the first aspect, in the second possible implementation method of the first aspect, when the service processing server sets the response time constraint information according to each RPC request in the service request queue When the execution priority of each RPC request is described, the execution priority of each RPC request is set according to the response time constraint information of each RPC request in the service request queue, including:

Obtaining a constraint remaining time of the processing constraint time of the respective RPC request according to the response time constraint information; setting an execution priority of each RPC request according to a predefined rule, where the predefined rule includes: Less, the higher the execution priority of the setting. With reference to the first aspect or the first implementation method of the first aspect, in a third possible implementation method of the first aspect, when the service processing server constrains the information distribution according to the response time of each RPC request in the service request queue When the execution resource of each RPC request is described, the allocating the execution resource of each RPC request according to the response time constraint information of each RPC request in the service request queue includes: acquiring the RPC request according to the response time constraint information Processing the remaining time of the constraint time; allocating the execution resources of the respective RPC requests according to the predefined rules, where the predefined rules include: the fewer the remaining time of the constraints, the more execution resources are allocated. In conjunction with the second implementation method of the second aspect or the first aspect, in the fourth possible implementation method of the first aspect, the predefined rule further includes: if the processing constraint time remaining time If the prediction completion time of the RPC request is less than, the execution priority and the allocation execution resource are set to be set for the RPC request. With reference to the second implementation method of the second aspect or the third implementation method of the first aspect, in the fifth possible implementation method of the first aspect, the predefined rule further includes: if there are more than two RPC requests An RPC request with a small ratio is set to a higher execution priority, and/or allocates more execution resources. In combination with the first aspect or any one of the first to third aspects of the first aspect, the first aspect After the remote process call RPC request sent by the receiving service distribution server, the method includes: selecting, according to the response time constraint information and the emergency threshold, a processing policy of the RPC request, where the processing policy includes And the first processing policy and the second processing policy, if the constraint remaining time of the response time constraint information is greater than the emergency threshold, selecting to execute the first processing policy, if the constraint time remaining information of the response time constraint information If the value is less than or equal to the emergency threshold, the second processing policy is selected to be executed, and the processing complexity of the first processing policy is greater than the second processing policy. With reference to the first aspect, or any one of the first to third aspects of the first aspect, in the seventh possible implementation method of the first aspect, the responding time constraint information setting according to each RPC request in the service request queue After the execution priority of each RPC request, and/or the execution resources of the respective RPC requests, the following:

The service processing server processes the RPC request with the highest priority, records the processing time of the RPC request, and updates the response time constraint information of the RPC request according to the processing time; the service processing server sends the service to the service The allocation server feeds back the response time constraint information of the RPC request that has been processed.

The request processing method provided by the second aspect of the embodiments of the present invention includes:

The service distribution server receives the user request; the service distribution server allocates response time constraint information for the user request, the response time constraint information is used to mark the processing constraint time of the user request and the processing time that has occurred; the service allocation The server generates a remote procedure call RPC request according to the user request; the service distribution server sends an RPC request to the service processing server, where the RPC request carries the response time constraint information; and causes the service processing server to respond according to the response time The constraint information sets an execution priority for the RPC request, and/or allocates an execution resource.

In the first possible implementation method of the second aspect, the allocating the response time constraint information of the user request, including: according to the historical value of the response time constraint information and the hardware information of the service processing server The user requests to allocate response time constraint information.

In conjunction with the first possible implementation of the second aspect, in the second possible implementation method of the second aspect, if the hardware performance of the current service processing server is relative to the recording of the response time constraint When the historical value of the information is changed, the historical value according to the response time constraint information and the hardware information of the service processing server are used to allocate the response time constraint information to the user request, including: acquiring the first processing speed and the second a ratio of the processing speed, multiplying the historical value of the response time constraint information by the ratio, to obtain response time constraint information currently allocated for the user request; the first processing speed is a hardware performance of the current service processing server a processing speed of the RPC request, the second processing speed is a processing speed of the hardware processing performance of the service processing server to the RPC request when the historical value of the response time constraint information is recorded; if the current service processing server The hardware performance is not changed when the historical value of the response time constraint information is recorded, and the historical value according to the response time constraint information and the hardware information of the service processing server allocate response time constraint information to the user request. , including: history of the response time constraint information As the current allocation request response time constraint for the user information.

With reference to the second aspect, or any one of the first to the second aspects of the second aspect, in the third possible implementation method of the second aspect, the generating, by the user request, the remote procedure call RPC request comprises: extracting the completion Determining a service parameter required by the user request; determining, according to the preset service logic, a step that needs to be performed to complete the user request, determining, according to the step that needs to be performed, a service processing server that needs to be invoked; The server allocates the corresponding service parameter, and generates an RPC request corresponding to the service processing server.

With reference to the second aspect, or any one of the first to the second aspects of the second aspect, in the fourth possible implementation method of the second aspect, after the sending the RPC request to the service processing server, the method includes: receiving the service Processing the response time constraint information fed back by the server; updating the response time constraint information of the corresponding RPC request by using the response time constraint information.

With the fourth implementation method of the second aspect, in the fifth possible implementation method of the second aspect, after the receiving the response time constraint information fed back by the service processing server, the method includes: And an emergency allocation threshold to select an allocation processing policy corresponding to the user request, where the allocation processing policy includes: a first allocation processing policy and a second allocation processing policy, if the constraint remaining time of the response time constraint information is greater than the emergency To handle the threshold, choose to execute And the first allocation processing policy, if the constraint remaining time of the response time constraint information is less than or equal to the emergency processing threshold, selecting to perform the second allocation processing policy, the processing complexity of the first allocation processing policy Greater than the second allocation processing policy. The service processing server provided by the third aspect of the embodiments of the present invention includes:

a request receiving unit, configured to receive a remote procedure call RPC request sent by the service allocation server, where the RPC request includes: response time constraint information; the response time constraint information is used to mark a processing constraint of the user request corresponding to the RPC request a storage unit, configured to add the RPC request to the service request queue, and a setting unit, configured to set execution of the respective RPC request according to response time constraint information of each RPC request in the service request queue Priority, and/or allocation of execution resources for the respective RPC requests. A service processing server according to a fourth aspect of the present invention includes: a user request receiving unit, configured to receive a user request; an information distribution unit, configured to allocate response time constraint information for the user request, where the response time constraint information is used And a request processing unit, configured to generate a remote procedure call RPC request according to the user request, and a request sending unit, configured to send an RPC request to the service processing server, where The RPC request carries the response time constraint information; causing the service processing server to set an execution priority for the RPC request according to the response time constraint information, and/or allocate an execution resource. The service processing server provided by the fifth aspect of the embodiments of the present invention includes:

a service distribution server and a service processing server; the service distribution server is configured to receive a user request; and allocate response time constraint information for the user request, where the response time constraint information is used to mark a processing constraint time of the user request and has occurred Processing time; generating a remote procedure call RPC request according to the user request; sending an RPC request to the service processing server; the RPC request carrying the response time constraint information; the service processing server is configured to receive a remotely sent by the service distribution server The process invokes an RPC request, adding the RPC request to the service request queue; setting an execution priority of the respective RPC request according to response time constraint information of each RPC request in the service request queue, and/or allocating the respective RPC request Execution resources. It can be seen from the above technical solution that the embodiment of the present invention has the following advantages: In the embodiment of the present invention, the remote procedure call (RPC) request received by the service processing server carries the response time constraint information, and the response The time constraint information is used to mark the processing constraint time of the user request corresponding to the RPC request and the processing time that has occurred, so that the service processing server can allocate information according to the response time of each RPC request in the service request queue before processing the RPC request. The execution priority and/or execution resources of the respective RPC requests are such that the time-critical RPC request can be prioritized. DRAWINGS

1 is a schematic flow chart of a request processing method according to an embodiment of the present invention;

2 is another schematic flowchart of a request processing method according to an embodiment of the present invention;

3 is another schematic flowchart of a request processing method according to an embodiment of the present invention;

4 is another schematic flowchart of a request processing method according to an embodiment of the present invention;

FIG. 5 is a schematic structural diagram of a service processing server according to an embodiment of the present invention; FIG.

6 is a schematic structural diagram of a service distribution server according to an embodiment of the present invention; FIG. 7 is a schematic structural diagram of a request processing system according to an embodiment of the present invention;

FIG. 8 is a schematic structural diagram of a computer device according to an embodiment of the present invention. detailed description

An embodiment of the present invention provides a request processing method for dynamically adjusting a request execution priority and resource allocation according to a requested processing time.

Referring to FIG. 1, an embodiment of a request processing method in an embodiment of the present invention includes:

101. The service processing server receives an RPC request sent by the service distribution server.

The service processing server receives an RPC request sent by the service allocation server, where the RPC request includes: response time constraint information; the response time constraint information is used to mark a processing constraint time and a processing time of the user request corresponding to the RPC request .

Exemplarily, after the response time constraint information is obtained, the response time constraint information may be stored in a thread local storage (Thread Loca l Storage, TLS) of the service processing server. In an actual application, the service distribution server receives the user request sent by the user, and allocates a response time constraint information for the user request, where the response time constraint information includes a processing constraint time corresponding to the application or service requested by the user. For limiting the maximum response time of the user request during the entire processing process (that is, the user request needs to be processed within the processing constraint time), and an already processed processing time with an initial state of zero for counting the user. The processing time that is requested to be generated at the processing stage of each business processing server.

After the service distribution server allocates the response time constraint information to the user request, the service distribution server generates an RPC request according to the user request, and sends the RPC request to the target service processing server (that is, the service processing server corresponding to the RPC requesting the requested application or service) Sending the RPC request, the RPC request carries response time constraint information corresponding to the user request, and the RPC request is used to request service processing from the target service processing server. In an actual application, completing a user request may need to be processed in multiple stages. The service distribution server generates an RPC request for each stage, and each RPC request is sent to the corresponding service distribution server for processing. Each phase of the RPC request carries the current response time constraint information. Each time an RPC request is completed, the response time constraint information is updated once. The current response time constraint information also reflects the processing constraint time of the user request, and before. The total processing time generated at each stage. Alternatively, since the difference between the processing constraint time and the processing time that has occurred is equal to the remainder of the processing constraint time, the obvious substitution between the parameters should not be construed as limiting the invention.

1 02, the service processing server adds the RPC request to the service request queue; the service processing server adds the RPC request to the service request queue, where the service request queue is used to store different RPC requests. In practical applications, in order to reduce the cost of operation and improve resource utilization, resources in a service processing server may be shared by multiple service allocation servers. Therefore, the service processing server may need to process multiple RPC requests at the same time. And these RPC requests will be added to the service request queue for allocation processing.

1 03. The business processing server allocates execution priority and/or execution resources for the RPC request. The service processing server according to the response time constraint of each RPC request in the service request queue The information sets an execution priority of the respective RPC requests, and/or allocates execution resources of the respective RPC requests.

Optionally, allocating execution resources of the respective RPC requests may include: adjusting a thread's central processing unit (Cent ra l Proces s ing Uni t , CPU ) scheduling priority, and adjusting thread input/output (I / O) priority Level and so on. In the embodiment of the present invention, the RPC request received by the service processing server carries the response time constraint information, where the response time constraint information is used to mark the processing constraint time and the processing time of the user request corresponding to the RPC request, so that The service processing server may allocate the execution priority and/or the execution resource of each RPC request according to the response time constraint information of each RPC request in the service request queue before processing the RPC request, so that the time-critical RPC request can be preferentially processed. .

In an actual application, the service processing server may have various policies according to the regulation of the response time constraint information. Referring to FIG. 2, another embodiment of the request processing method in the embodiment of the present invention includes:

201. The service processing server receives an RPC request sent by the service distribution server, and the service processing server receives a remote procedure call RPC request sent by the service distribution server, where

The RPC request includes: response time constraint information; the response time constraint information is used to mark a processing constraint time of the user request corresponding to the RPC request and a processing time that has occurred.

202. The service processing server selects a processing policy of the RPC request according to the response time constraint information and the emergency threshold. The service processing server selects a processing policy of the RPC request according to the response time constraint information and the emergency threshold. If the constraint remaining time of the response time constraint information is greater than the emergency threshold, the first processing policy is selected to be executed, and if the constraint remaining time of the response time constraint information is less than or equal to the emergency threshold, then the execution is selected. The second processing policy includes: a first processing policy and a second processing policy, where the processing complexity of the first processing policy is greater than the second processing policy. The emergency threshold is a judgment threshold of a processing policy for selecting a RPC request by the service processing server, and may be a time threshold. Exemplarily, in an actual application, the first processing policy may include: recommending content that may be of interest to the user by means of data mining analysis according to the request of the user; and the second processing policy may specifically include recommending the current concern directly to the user. High content; because the processing complexity of the first processing strategy is greater than the second processing strategy, the processing server can select the processing complexity when the processing time is urgent (specifically, the constraint remaining time is less than or equal to the emergency threshold) A smaller second processing strategy to save processing time. It can be understood that the first processing policy and the second processing policy represent only two types of policies with different processing complexity, and do not specifically refer to any two strategies, and the first processing strategy may indicate that the processing complexity is the same. Or two or more similar processing strategies, and the second processing strategy may also represent two or more processing strategies with the same or similar processing complexity.

Further, the service processing server may also set multiple levels of emergency thresholds to correspond to processing strategies of multiple complexity types, which are not specifically limited herein.

203. The service processing server adds the RPC request to the service request queue. The service processing server adds the RPC request to the service request queue, where the service request queue is used to store different RPC requests. In practical applications, in order to reduce the cost of operation and improve resource utilization, resources in a service processing server may be shared by multiple service allocation servers. Therefore, the service processing server may need to process multiple RPC requests at the same time. And these RPC requests will be added to the service request queue for allocation processing. It can be understood that, in an actual application, in step 202 and step 203, there is no strict sequential relationship, that is, "add the RPC request to the service request queue", and then execute "according to the response time constraint." Information and emergency thresholds select the processing strategy for the RPC request."

204. The service processing server allocates an execution priority and/or an execution resource to the RPC request. The service processing server sets an execution priority of the each RPC request according to the response time constraint information of each RPC request in the service request queue, and/or Allocating execution resources of the respective RPC requests.

Optionally, assigning execution priorities and/or executing resources can be done through the operating system or super management The program (Hyperv is or ) adjusts the resource usage of different request processing threads, such as adjusting the CPU's CPU scheduling priority, adjusting the thread's I / O priority, etc.; also can support the request processing thread through the hardware that supports the priority The execution priority is adjusted, for example, the hardware resources used by each processing thread are allocated on the CPU's cache memory, memory controller, or system bus according to the priority determined by the response time constraint information.

In an actual application, the service processing server determines the processing mode of the RPC request according to the processing capability of the device and the number and type of RPC requests in the service request queue before processing the RPC request; if the service processing server can simultaneously process All the RPC requests in the service request queue, the service processing server allocates the execution resources of the respective RPC requests according to the response time constraint information of each RPC request in the service request queue; if the service processing server can only The service processing server sets the execution priority of each RPC request according to the response time constraint information of each RPC request in the service request queue, if the RPC request is processed by the service request queue separately; The processing server cannot process all the RPC requests in the service request queue at the same time, and can not only process any one of the RPC requests in the service request queue, but the service processing server requests each RPC request in the queue according to the service request. Response time The bundle information sets an execution priority of the respective RPC requests and allocates execution resources of the respective RPC requests.

Further, when the service processing server sets the execution priority of each RPC request according to the response time constraint information of each RPC request in the service request queue, the service processing server may obtain the RPC according to the response time constraint information. The processing processing constraint time constraint remaining time; the execution priority of the respective RPC request is set according to the predefined rule, the predefined rule includes: the less the constraint remaining time, the higher the execution priority is set. When the service processing server allocates the execution resources of the respective RPC requests according to the response time constraint information of each RPC request in the service request queue, the service processing server may acquire the processing constraints of the respective RPC requests according to the response time constraint information. The remaining time of the time; the execution resources of the respective RPC requests are allocated according to a predefined rule, and the predefined rules include: the fewer the remaining time of the constraint, the more execution resources are allocated. Further, after the service processing server obtains the remaining time of the constraint of each RPC request, if the remaining time of the constraint of two or more RPC requests is found to be equal, the remaining constraint is obtained. The ratio of the time to the processing constraint time sets a higher execution priority for the RPC request with a smaller ratio, and/or allocates more execution resources. Further, after the service processing server obtains the constraint remaining time of the processing constraint time, the remaining time of the constraint may be first determined, and if the remaining time of the processing constraint time is less than the predicted completion time of the RPC request, then the device abandons Setting an execution priority and an allocation execution resource for the RPC request to save processing resources of the service processing server. Optionally, the service processing server may select to send a request failure response to the service allocation server, and the service distribution server provides feedback to the user. The processing of the user request failed, causing the service request to be resent. In an actual application, the service processing server may know the minimum processing time of the user request at each processing stage (which may be notified by the service allocation server), and the current service processing server adds the minimum processing time of each remaining processing stage of the user request. , the prediction completion time can be obtained.

205. The service processing server feeds back, to the service allocation server, response time constraint information of the RPC request that is processed. After completing the setting of the execution priority and/or the allocation of the execution resources, the service processing server processes the highest priority RPC request, records the processing time of the RPC request, and updates the RPC request according to the processing time. The response time constraint information (that is, the processing time added to the processing time in the response time constraint information is obtained, and the updated processing time has been obtained). The service processing server feeds back the response time constraint information (i.e., the updated response time constraint information) of the RPC request that has been processed to the service distribution server, so that the service distribution server updates the response time constraint information requested by the corresponding user. The request processing method in the embodiment of the present invention is described above from the perspective of the service processing server. The request processing method in the embodiment of the present invention is described from the service distribution server side. Referring to FIG. 3, in the embodiment of the present invention, Another embodiment of the request processing method includes:

301. The service distribution server receives the user request. The service distribution server receives the user request, and the user requests the service requirement or the application requirement that the user submits to the network side, and the service requirement or the application requirement may include: searching, online shopping, and the like.

302. The service allocation server allocates response time constraint information to the user request. The service distribution server allocates response time constraint information for the user request, and the response time constraint information is used to mark the processing constraint time of the user request and the processing time that has occurred.

Exemplarily, the response time constraint information can be stored in the TLS of the service distribution server. The response time constraint information includes a processing constraint time corresponding to the application or service requested by the user, and is used to define a maximum response time of the user request during the entire processing process (ie, the user request needs to be within the processing constraint time) Processing is completed), there is also an processing time that has an initial state of zero, which is used to count the processing time that the user requests a total of processing time in each processing server. Optionally, the response time constraint information may be set by the administrator according to the service or application type requested by the different user. After the initialization setting is completed, the service distribution server may automatically allocate according to the historical value of the response time constraint information requested by the corresponding user. .

303. The service distribution server generates an RPC request according to the user request. After the service distribution server allocates the response time constraint information to the user request, the service distribution server generates an RPC request according to the user request, where the RPC request carries the request with the user. Corresponding response time constraint information, the RPC request is used to request service processing from the target service processing server. In an actual application, a user request may be processed in multiple stages. The service distribution server generates an RPC request for each stage, and each RPC request is sent to the corresponding service distribution server for processing. Each phase of the RPC request carries the current response time constraint information. Each time an RPC request is completed, the response time constraint information is updated once. The current response time constraint information also reflects the processing constraint time of the user request, and before. The total processing time generated at each stage.

304. The service distribution server sends an RPC request to the service processing server. The service distribution server sends an RPC request to the service processing server, where the RPC request carries the response time constraint information; causing the service processing server to set an execution priority for the RPC request according to the response time constraint information, and/or Assign execution resources. In the embodiment of the present invention, the RPC request sent by the service distribution server to the service processing server carries the response time constraint information, where the response time constraint information is used to mark the processing constraint time of the user request and the processing time that has occurred, so that the service The processing server can process the RPC request before it can The execution priority and/or execution resources of the respective RPC requests are allocated according to the response time constraint information of each RPC request in the service request queue, so that the time-critical RPC request can be preferentially processed.

The following is a detailed description of the method for performing the request processing on the service distribution server. Referring to FIG. 4, another embodiment of the request processing method in the embodiment of the present invention includes:

401. The service distribution server receives a user request.

The service distribution server receives the user request, and the user requests the service demand or the application requirement that the user makes to the network side, and the service requirement or the application requirement may include: searching, online shopping, and the like.

402. The service distribution server allocates response time constraint information to the user request. The service distribution server allocates response time constraint information for the user request, where the response time constraint information is used to mark the processing constraint time of the user request and has occurred. Processing time

The response time constraint information includes a processing constraint time corresponding to the application or service requested by the user, and is used to define a maximum response time of the user request during the entire processing process (ie, the user request needs to be within the processing constraint time) Processing is completed), there is also an processing time that has an initial state of zero, which is used to count the processing time that the user requests a total of processing time in each processing server. Optionally, the response time constraint information may be set by the administrator according to the service or application type requested by the different user. After the initialization setting is completed, the service distribution server may automatically allocate according to the historical value of the response time constraint information requested by the corresponding user. . Optionally, the service allocation server may allocate response time constraint information to the user request according to the historical value of the response time constraint information and the hardware information of the service processing server; specifically, if the current service processing server has hardware performance And comparing the historical value of the response time constraint information, obtaining a ratio of the first processing speed to the second processing speed, multiplying the historical value of the response time constraint information by the ratio, to obtain a current Responsive time constraint information assigned to the user request; the first processing speed is a processing speed of a hardware performance of a current service processing server to the RPC request, and the second processing speed is a recording response time constraint information The historical value of the business processing server hardware performance processing speed of the RPC request; if the current business processing server hardware performance, relative to the historical value of the recording response time constraint information When a change occurs, the historical value of the response time constraint information is taken as the response time constraint information currently allocated for the user request.

403. The service distribution server generates an RPC request according to the user request. After the service distribution server allocates the response time constraint information to the user request, the service distribution server generates an RPC request according to the user request, where the RPC request carries the request with the user. Corresponding response time constraint information, the RPC request is used to request service processing from the target service processing server. In an actual application, a user request may be processed in multiple stages. The service distribution server generates an RPC request for each stage, and each RPC request is sent to the corresponding service distribution server for processing. Each phase of the RPC request carries the current response time constraint information. Each time an RPC request is completed, the response time constraint information is updated once. The current response time constraint information also reflects the processing constraint time of the user request, and before. The total processing time generated at each stage. Exemplarily, the RPC request generating method may be: extracting a service parameter (such as a request type, a user identifier, a product identifier, and the like) required to complete the user request; determining, according to preset business logic, that the user request needs to be performed to complete the user request. And determining, according to the step that needs to be performed, the service processing server that needs to be invoked; respectively, assigning the corresponding service parameter to the service processing server that needs to be invoked, and generating an RPC request corresponding to the service processing server.

404. The service distribution server sends an RPC request to the service processing server. The service distribution server sends an RPC request to the service processing server, where the RPC request carries the response time constraint information, so that the service processing server determines the information according to the response time constraint. An execution priority is set for the RPC request, and/or an execution resource is allocated.

405. The service distribution server receives the response time constraint information fed back by the service processing server. The service distribution server receives the response time constraint information fed back by the service processing server, and updates the response time constraint information of the corresponding RPC request by using the response time constraint information. 406. The service distribution server selects an allocation processing policy requested by the corresponding user according to the updated response time constraint information.

The service distribution server selects the use according to the response time constraint information and the emergency allocation threshold The user requests a corresponding allocation processing policy, where the allocation processing policy includes: a first allocation processing policy and a second allocation processing policy, if the constraint remaining time of the response time constraint information is greater than the emergency processing threshold, selecting an execution location The first allocation processing policy, if the constraint remaining time of the response time constraint information is less than or equal to the emergency processing threshold, selecting to perform the second allocation processing policy, the processing complexity of the first allocation processing policy Greater than the second allocation processing policy. The emergency allocation value is a judgment threshold of the processing policy that the service allocation server selects the user request, and may be a time threshold. Exemplarily, in a practical application, the first allocation processing policy may specifically include a service processing procedure necessary for completing a user request, and optionally data mining, statistics, and analysis process content for improving service quality; The second allocation processing policy may specifically include only the business process flow content necessary for completing the user request; since the processing complexity of the first allocation processing policy is greater than the second allocation processing policy, when the processing time is urgent (specifically, by constraining the remaining time) When the value is less than or equal to the emergency threshold, the service allocation server may choose to process the second allocation processing strategy with less complexity to save processing time. It can be understood that the first allocation processing policy and the second allocation processing strategy represent only two types of policies with different processing complexity, and do not specifically refer to any two strategies, and the first allocation processing strategy may indicate processing. For two or more processing strategies with the same or similar complexity, the second allocation processing strategy may also represent two or more processing strategies with the same or similar processing complexity.

Further, the service distribution server may also set a multi-level emergency allocation threshold to correspond to a plurality of complexity type allocation processing policies, which are not specifically limited herein. The following describes an embodiment of the service processing server of the present invention for performing the foregoing request processing method. For the logical structure, refer to FIG. 5. An embodiment of the service processing server in the embodiment of the present invention includes:

The request receiving unit 501 is configured to receive a remote procedure call RPC request sent by the service allocation server, where the RPC request includes: response time constraint information; and the response time constraint information is used to mark a process of processing the user request corresponding to the RPC request. Restricting time and processing time has occurred; the storage unit 502 is configured to add the RPC request to the service request queue; and the setting unit 503 is configured to: according to the response time of each RPC request in the service request queue The bundle information sets an execution priority of the respective RPC requests, and/or allocates execution resources of the respective RPC requests.

Further, the setting unit 503 in the embodiment of the present invention is configured to: if the service processing server can simultaneously process all RPC requests in the service request queue, the service processing server is configured according to each of the service request queues The response time constraint information requested by the RPC allocates execution resources of the respective RPC requests;

If the service processing server can only process any one of the RPC requests in the service request queue, the service processing server sets the RPC request according to the response time constraint information of each RPC request in the service request queue. Execution priority; if the service processing server cannot process all RPC requests in the service request queue at the same time, nor can it only process any RPC request in the service request queue separately, the service processing server is based on The response time constraint information of each RPC request in the service request queue sets an execution priority of each RPC request, and allocates execution resources of the respective RPC requests.

Further, the setting unit 503 in the embodiment of the present invention may include: a resource allocation module 5031, configured to acquire, according to the response time constraint information, a constraint remaining time of a processing constraint time of each RPC request; according to a predefined rule Setting the execution priority of each of the RPC requests, where the predefined rule includes: the less the remaining time of the constraint, the higher the execution priority is set; the priority setting module 5032 is configured to use the response time constraint information according to the response time constraint information. Acquiring the remaining time of the processing constraint time of the respective RPC request; allocating the execution resources of the respective RPC requests according to the predefined rule, where the predefined rule includes: the less the remaining time of the constraint, the further the allocated execution resources The service processing server in the embodiment of the present invention may further include: a processing policy selecting unit 504, configured to select a processing policy of the RPC request according to the response time constraint information and an emergency threshold, where the processing policy includes: a processing strategy and a second processing strategy, if the response time constraint information is about Remaining time value is greater than the width of emergency, is selected to perform the first processing strategy, if the remaining time of the response time of the constraining information is equal to or less than the emergency The threshold value is selected to execute the second processing policy, where the processing complexity of the first processing policy is greater than the second processing policy.

Further, the service processing server in the embodiment of the present invention further includes: a request processing unit 505, configured to process the RPC request with the highest priority, record the processing time of the RPC request, and update the processing according to the processing time. Response time constraint information for RPC requests;

The information feedback unit 506 is configured to feed back, to the service distribution server, response time constraint information of the RPC request that completes the processing. The specific interaction process of each unit of the service allocation server in the embodiment of the present invention is as follows: The request receiving unit 501 receives an RPC request sent by the service allocation server, where the RPC request includes: response time constraint information; the response time constraint information is used for marking The processing constraint time of the user request corresponding to the RPC request and the processing time that has occurred. Exemplarily, after the response time constraint information is obtained, the response time constraint information may be stored in a thread local storage of the service processing server.

After receiving the RPC request, the storage unit 502 adds the RPC request to the service request queue, where the service request queue is used to store different RPC requests. In practical applications, in order to reduce the cost of operation and improve resource utilization, resources in a service processing server may be shared by multiple service allocation servers. Therefore, the service processing server is likely to need to process multiple RPC requests at the same time. And these RPC requests will be added to the service request queue for allocation processing. Optionally, after the response time constraint information is obtained, the processing policy selection unit may be

The 504 selects a processing policy of the RPC request according to the response time constraint information and the emergency threshold. If the constraint remaining time of the response time constraint information is greater than the emergency threshold, the first processing policy is selected to be executed. And the second processing policy is selected to be executed, where the processing policy includes: a first processing policy and a second processing policy, where the first processing policy is performed, where the remaining processing time of the response time constraint information is less than or equal to the emergency threshold value. The processing complexity of the policy is greater than the second processing strategy. The emergency threshold is a judgment threshold of a processing policy for selecting a RPC request by the service processing server, and may be a time threshold. Exemplarily, in an actual application, the first processing policy may include: recommending content that may be of interest to the user by means of data mining analysis according to the request of the user; and the second processing policy may specifically include recommending the current concern directly to the user. High content; because the processing complexity of the first processing strategy is greater than the second processing strategy, the processing server can select the processing complexity when the processing time is urgent (specifically, the constraint remaining time is less than or equal to the emergency threshold) A smaller second processing strategy to save processing time. After the RPC request is added to the service request queue, the setting unit 503 sets the execution priority of the respective RPC requests according to the response time constraint information of each RPC request in the service request queue, and/or allocates the respective RPC requests. Execute resources. Specifically, before processing the RPC request, the RPC request processing manner is determined according to the processing capability of the device and the number and type of the RPC request in the service request queue; if the service processing server can simultaneously process the service request queue All the RPC requests, the service processing server allocates the execution resources of the respective RPC requests according to the response time constraint information of each RPC request in the service request queue; if the service processing server can only process the service separately If the RPC request is requested in the queue, the service processing server sets the execution priority of each RPC request according to the response time constraint information of each RPC request in the service request queue; if the service processing server cannot simultaneously Processing all the RPC requests in the service request queue, and not only individually processing any one of the RPC requests in the service request queue, and the service processing server according to the response time constraint information of each RPC request in the service request queue Set the execution priority of each RPC request And allocating the resources of each execution of the RPC request. Further, the resource allocation module 5031 of the setting unit 503 may acquire the constraint remaining time of the processing constraint time of each RPC request according to the response time constraint information; and set the execution priority of each RPC request according to the predefined rule, The predefined rule includes: the less the remaining time of the constraint, the higher the execution priority is set; and the remaining time of the processing constraint time of the respective RPC request is obtained by the priority setting module 5032 according to the response time constraint information. And allocating execution resources of the respective RPC requests according to a predefined rule, where the predefined rules include: the fewer the remaining time of the constraint, the more execution resources are allocated. After completing the setting of the execution priority and/or the allocation of the execution resource, the request processing unit 505 processes the RPC request with the highest priority, records the processing time of the RPC request, and according to the The processing time updates the response time constraint information of the RPC request (that is, the processing time plus the processing time that has occurred in the response time constraint information, and the updated processing time has been obtained). The information feedback unit 506 feeds back the response time constraint information (ie, the updated response time constraint information) of the RPC request that has been processed to the service distribution server, so that the service distribution server updates the response time constraint information requested by the corresponding user. An embodiment of the service distribution server of the present invention for performing the above-mentioned request processing method is described below. For the logical structure, please refer to FIG. 6. An embodiment of the service distribution server in the embodiment of the present invention includes:

a user request receiving unit 601, configured to receive a user request; an information allocating unit 602, configured to allocate response time constraint information for the user request, where the response time constraint information is used to mark a processing constraint time requested by the user and has occurred a processing time generating unit 603, configured to generate a remote procedure call RPC request according to the user request, and a request sending unit 604, configured to send an RPC request to the service processing server, where the RPC request carries the response time constraint information; And causing the service processing server to set an execution priority for the RPC request according to the response time constraint information, and/or allocate an execution resource.

Further, the information distribution unit 602 in the embodiment of the present invention is specifically configured to: allocate response time constraint information to the user request according to the historical value of the response time constraint information and the hardware information of the service processing server.

Further, the information distribution unit 602 in the embodiment of the present invention includes: a first allocation module 6021, configured to change, if the hardware performance of the current service processing server is relative to the historical value of the response time constraint information, Obtaining a ratio of the first processing speed to the second processing speed, multiplying the historical value of the response time constraint information by the ratio, and obtaining response time constraint information currently allocated for the user request; the first processing speed Processing speed of the RPC request for the hardware performance of the current service processing server, the second processing speed is processing of the RPC request by the hardware performance of the service processing server when recording the historical value of the response time constraint information Speed

a second allocation module 6022, configured to: if the current service processing server hardware performance, relative to When the historical value of the response time constraint information is recorded, the historical value of the response time constraint information is used as the response time constraint information currently allocated for the user request.

Further, the request generating unit 603 in the embodiment of the present invention includes: a parameter extraction module 6031, configured to extract a service parameter required to complete the user request; and a logic analysis module 6032, configured to determine a completion according to the preset service logic. Determining, by the user, the step that needs to be performed, determining, according to the step that needs to be performed, the service processing server that needs to be invoked; the request generating module 6033, configured to respectively allocate the corresponding service parameter to the service processing server that needs to be invoked, and generate and The service processing server corresponds to an RPC request.

Further, the service processing server in the embodiment of the present invention further includes: an information receiving unit 605, configured to receive response time constraint information fed back by the service processing server;

The information updating unit 606 is configured to update the response time constraint information of the corresponding RPC request by using the response time constraint information. The allocation policy selection unit 607 is configured to select an allocation processing policy corresponding to the user request according to the response time constraint information and the emergency allocation threshold, where the allocation processing policy includes: a first allocation processing policy and a second allocation processing policy, And if the constraint remaining time of the response time constraint information is greater than the emergency processing threshold, the first allocation processing policy is selected to be executed, if the constraint remaining time of the response time constraint information is less than or equal to the emergency processing threshold And executing the second allocation processing policy, where the processing complexity of the first allocation processing policy is greater than the second allocation processing policy. The specific interaction process of each unit of the service distribution server in the embodiment of the present invention is as follows: The user request receiving unit 601 receives a user request, and the user request is a service requirement or an application requirement that the user proposes to the network side, and the service requirement or application requirement may include : Search, shop online. After receiving the user request, the information allocating unit 602 requests the user to request response time constraint information for marking the processing constraint time requested by the user and the processing time that has occurred. Optionally, if the hardware performance of the current service processing server is changed relative to the historical value of the response time constraint information, the first allocation module 6021 obtains the first processing speed and a ratio of the second processing speed, multiplying the historical value of the response time constraint information by the ratio, to obtain response time constraint information currently allocated for the user request; the first processing speed is the current service processing server The processing speed of the RPC request by the hardware performance, the second processing speed is a processing speed of the hardware processing performance of the service processing server to the RPC request when the historical value of the response time constraint information is recorded; if the current service processing The hardware performance of the server is not changed when the historical value of the response time constraint information is recorded, and the second allocation module 6022 uses the historical value of the response time constraint information as the response time constraint currently allocated for the user request. information.

After the service allocation server allocates the response time constraint information to the user request, the request generating unit 603 generates an RPC request according to the user request, where the RPC request carries response time constraint information corresponding to the user request, and the RPC request is used to The target business processing server requests business processing. Specifically, the parameter extraction module 6031 extracts the service parameters (such as the request type, the user identifier, the product identifier, and the like) required to complete the user request; the logic analysis module 6032 determines, according to the preset service logic, that the user request needs to be executed. Steps: determining, according to the step that needs to be performed, a service processing server that needs to be invoked; the request generation module 6033 respectively assigns the corresponding service parameter to the service processing server that needs to be invoked, and generates an RPC request corresponding to the service processing server. . After generating the RPC request, the request sending unit 604 sends an RPC request to the service processing server, where the RPC request carries the response time constraint information; and causes the service processing server to set the RPC request according to the response time constraint information. Execution priorities, and/or allocation of execution resources. The information receiving unit 605 receives the response time constraint information fed back by the service processing server, and the trigger information updating unit 606 updates the response time constraint information of the corresponding RPC request using the response time constraint information. After the response time constraint information is updated, the allocation policy selection unit 607 selects an allocation processing policy corresponding to the user request according to the response time constraint information and the emergency allocation threshold, and the allocation processing policy includes: a first allocation processing policy and And a second allocation processing policy, if the constraint remaining time of the response time constraint information is greater than the emergency processing threshold, selecting to execute the first allocation processing policy, if the constraint remaining time of the response time constraint information is less than or equal to And performing the second allocation processing policy, where the processing complexity of the first allocation processing policy is greater than the second allocation processing policy. The emergency allocation value is a judgment threshold of the processing policy that the service allocation server selects the user request, and may be a time threshold. Exemplarily, in a practical application, the first allocation processing policy may specifically include a service processing procedure necessary for completing a user request, and optionally data mining, statistics, and analysis process content for improving service quality; The second allocation processing policy may specifically include only the business process flow content necessary for completing the user request; since the processing complexity of the first allocation processing policy is greater than the second allocation processing policy, when the processing time is urgent (specifically, by constraining the remaining time) When the value is less than or equal to the emergency threshold, the service allocation server may choose to process the second allocation processing strategy with less complexity to save processing time. The following describes an embodiment of the request processing system of the present invention for performing the foregoing request processing method. For the logical structure, please refer to FIG. 7. An embodiment of the request processing system in the embodiment of the present invention includes:

a service distribution server 701 and a service processing server 702; the service distribution server 701 is configured to receive a user request; to allocate response time constraint information for the user request, the response time constraint information is used to mark a processing constraint time of the user request And processing time has occurred; generating a remote procedure call RPC request according to the user request; sending an RPC request to the service processing server; the RPC request carrying the response time constraint information; the service processing server 702 is configured to receive a service allocation a remote procedure call RPC request sent by the server, adding the RPC request to the service request queue; setting an execution priority of each RPC request according to response time constraint information of each RPC request in the service request queue, and/or an allocation Execution resources for each RPC request. For the specific steps performed by the service distribution server 701 and the service processing server 702 in the embodiment of the present invention, reference may be made to the foregoing method embodiments, and details are not described herein again. The embodiment of the present invention further provides a computer storage medium, wherein the computer storage medium can store a program, and the program includes some or all of the steps of the request processing method described in the foregoing method embodiments.

Referring to FIG. 8, the embodiment of the present invention further provides a service processing server, which may specifically include: Receiver 801, transmitter 802, memory 803 and processor 804 (the number of processors in the service processing server may be one or more, and one processor in FIG. 8 is exemplified). In some embodiments of the present invention, The receiver 801, the transmitter 802, the memory 803, and the processor 804 may be connected by a bus or other means, wherein the bus connection is taken as an example in FIG.

The memory 803 may be configured to store the following content: the RPC request includes: response time constraint information; the response time constraint information is used to mark a processing constraint time of the user request corresponding to the RPC request and the processing has occurred Pre-defined rule: the less the remaining time of the constraint, the higher the execution priority is set; the less the remaining time of the constraint, the more the execution resources are allocated; if the remaining time of the processing constraint time is less than Determining the execution priority of the RPC request, and then setting the execution priority and the allocation execution resource for the RPC request; if the remaining time of the constraint of the two or more RPC requests is equal, acquiring the constraint remaining time and the processing constraint The ratio of time sets a higher execution priority for the RPC request with a smaller ratio, and/or allocates more execution resources.

And the specific content of the first processing strategy and the second processing strategy. The receiver 801 is configured to receive a remote procedure call RPC request sent by the receiving service distribution server. The sender 802 is configured to feed back to the service distribution server response time constraint information of the RPC request that is processed.

The processor 804 is configured to: after the receiver 801 receives the remote procedure call RPC request sent by the service allocation server, add the RPC request to the service request queue; according to the response time constraint of each RPC request in the service request queue The information sets an execution priority of the respective RPC requests, and/or allocates execution resources of the respective RPC requests. Specifically, the execution priority setting may be performed according to a predefined rule, and/or the allocation of resources may be performed.

If the service processing server is capable of processing all the RPC requests in the service request queue at the same time, the service processing server allocates the execution resources of the respective RPC requests according to the response time constraint information of each RPC request in the service request queue. ;

If the service processing server can only process any one of the service request queues separately The RPC requests, the service processing server sets the execution priority of each RPC request according to the response time constraint information of each RPC request in the service request queue;

If the service processing server cannot process all the RPC requests in the service request queue at the same time, or if only one RPC request in the service request queue can be separately processed, the service processing server according to the service request The response time constraint information of each RPC request in the queue sets an execution priority of the respective RPC requests, and allocates execution resources of the respective RPC requests. Determining, according to the response time constraint information and the emergency threshold, a processing policy of the RPC request, where the processing policy includes: a first processing policy and a second processing policy, if the constraint remaining time of the response time constraint information is greater than the And selecting the first processing policy, if the constraint remaining time of the response time constraint information is less than or equal to the emergency threshold, selecting to execute the second processing policy, the first processing policy The processing complexity is greater than the second processing strategy. Referring to FIG. 8 , an embodiment of the present invention further provides a service distribution server, which may specifically include: a receiver 801, a transmitter 802, a memory 803, and a processor 804. The number of processors in the service distribution server may be one. Or a plurality of processors in FIG. 8 as an example. In some embodiments of the present invention, the receiver 801, the transmitter 802, the memory 803, and the processor 804 may be connected by a bus or other means, wherein, in FIG. Take the bus connection as an example. The memory 803 of the service distribution server can be used to store the following content:

The RPC request includes: response time constraint information; the response time constraint information is used to mark a processing constraint time of the user request corresponding to the RPC request and a processing time that has occurred; the historical value of the response time constraint information and the The hardware information of the service processing server. The receiver 801 of the service distribution server is configured to receive a user request and receive response time constraint information fed back by the service processing server.

The transmitter 802 of the service distribution server is configured to send an RPC request to the service processing server. The processor 804 of the service distribution server is configured to perform the following steps: after receiving the receiver 801 user request; assigning response time constraint information to the user request, and generating a remote procedure call RPC request based on the user request.

Further, when the response time constraint information is allocated for the user request, if the current service department The hardware performance of the server is changed when the historical value of the response time constraint information is recorded, and the ratio of the first processing speed to the second processing speed is obtained, and the historical value of the response time constraint information is multiplied by The ratio is obtained, and the response time constraint information currently allocated for the user request is obtained; the first processing speed is a processing speed of the hardware processing performance of the current service processing server to the RPC request, and the second processing speed is a recording office. Determining the processing speed of the RPC request by the hardware performance of the service processing server when responding to the historical value of the time constraint information; if the hardware performance of the current service processing server does not occur with respect to recording the historical value of the response time constraint information If the change is made, the historical value of the response time constraint information is taken as the response time constraint information currently allocated for the user request. Further, after receiving the response time constraint information fed back by the service processing server, selecting an allocation processing policy corresponding to the user request according to the response time constraint information and the emergency allocation threshold, the allocation processing policy includes: a processing policy and a second allocation processing policy, if the constraint remaining time of the response time constraint information is greater than the emergency processing threshold, selecting to execute the first allocation processing policy, if the constraint time remaining information of the response time constraint information If the threshold is less than or equal to the emergency processing threshold, the second allocation processing policy is selected to be executed, and the processing complexity of the first allocation processing policy is greater than the second allocation processing policy.

In the several embodiments provided herein, it should be understood that the disclosed apparatus and method can be implemented in other ways. For example, the device embodiments described above are merely illustrative. For example, the division of the unit is only a logical function division. In actual implementation, there may be another division manner, for example, multiple units or components may be combined or Can be integrated into another system, or some features can be ignored, or not executed. In addition, the mutual coupling or direct coupling or communication connection shown or discussed may be an indirect coupling or communication connection through some interface, device or unit, and may be in an electrical, mechanical or other form.

The components displayed by the unit may or may not be physical units, that is, may be located in one place, or may be distributed to multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of the embodiment. In addition, each functional unit in each embodiment of the present invention may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. Medium. The above integrated unit can be implemented in the form of hardware or in the form of a software functional unit. The integrated unit, if implemented in the form of a software functional unit and sold or used as a standalone product, may be stored in a computer readable storage medium. Based on such understanding, the technical solution of the present invention may contribute to the prior art or all or part of the technical solution may be embodied in the form of a software product stored in a storage medium. A number of instructions are included to cause a computer device (which may be a personal computer, server, or network device, etc.) to perform all or part of the steps of the methods described in various embodiments of the present invention. The foregoing storage medium includes: a USB flash drive, a removable hard disk, a read-only memory (ROM), a random access memory (RAM, Random Acces s Memory), a magnetic disk or an optical disk, and the like, which can store program codes. medium.

The above is only a specific embodiment of the present invention, but the scope of the present invention is not limited thereto, and any person skilled in the art can easily think of changes or substitutions within the technical scope of the present invention. It should be covered by the scope of the present invention. Therefore, the scope of the invention should be determined by the scope of the claims.

Claims

Rights request

A request processing method, comprising:

The service processing server receives a remote procedure call RPC request sent by the service allocation server, where the RPC request includes: response time constraint information; the response time constraint information is used to mark a processing constraint time of the user request corresponding to the RPC request, and Processing time occurred;

The service processing server adds the RPC request to a service request queue;

The service processing server sets an execution priority of the respective RPC requests according to response time constraint information of each RPC request in the service request queue, and/or allocates execution resources of the respective RPC requests.

2. The method of claim 1 wherein

If the service processing server can only process any one of the RPC requests in the service request queue, the service processing server sets the RPC request according to the response time constraint information of each RPC request in the service request queue. Execution priority

If the service processing server cannot process all the RPC requests in the service request queue at the same time, or does not only process any one of the RPC requests in the service request queue, the service processing server according to the service request The response time constraint information of each RPC request in the queue sets the execution priority of each RPC request, and allocates the execution resources of the respective RPC requests.

The method according to claim 1 or 2, wherein, when the service processing server sets the execution priority of each RPC request according to the response time constraint information of each RPC request in the service request queue, Setting the execution priority of each RPC request according to the response time constraint information of each RPC request in the service request queue, including:

Obtaining an constraint remaining time of the processing constraint time of the respective RPC request according to the response time constraint information;

The execution priority of the respective RPC request is set according to a predefined rule, and the predefined rule includes: the less the remaining time of the constraint, the higher the execution priority of the setting.

The method according to claim 1 or 2, wherein the service processing server allocates the respective RPCs according to response time constraint information of each RPC request in the service request queue. The execution resources of the respective RPC requests are allocated according to the response time constraint information of each RPC request in the service request queue, including:

Obtaining a remaining time of the processing constraint time of the respective RPC request according to the response time constraint information;

The execution resources of the respective RPC requests are allocated according to a predefined rule, and the predefined rules include: the fewer the remaining time of the constraint, the more execution resources are allocated.

The method according to claim 3 or 4, wherein the predefined rule further comprises: abandoning the RPC if the remaining time of the processing constraint time is less than a predicted completion time of the RPC request Request to set execution priority and allocate execution resources.

The method according to claim 3 or 4, wherein the predefined rule further comprises: if the remaining time of the constraint of the two or more RPC requests is equal, acquiring the remaining time of the constraint and the processing The ratio of constraint times sets a higher execution priority for the RPC request with a smaller ratio and/or allocates more execution resources.

The method according to any one of claims 1 to 4, wherein after the remote procedure call RPC request sent by the receiving service distribution server, the method comprises:

Determining, according to the response time constraint information and the emergency threshold, a processing policy of the RPC request, where the processing policy includes: a first processing policy and a second processing policy, if the constraint remaining time of the response time constraint information is greater than the And selecting the first processing policy, if the constraint remaining time of the response time constraint information is less than or equal to the emergency threshold, selecting to execute the second processing policy, the first processing policy The processing complexity is greater than the second processing strategy.

The method according to any one of claims 1 to 4, wherein the setting of the execution priority of each RPC request according to response time constraint information of each RPC request in the service request queue, and/or allocation After the execution resources of the respective RPC requests, the method includes:

The service processing server processes the highest priority RPC request, records the processing time of the RPC request, and updates the response time constraint information of the RPC request according to the processing time;

The service processing server feeds back to the service distribution server feedback response time constraint information of the RPC request that is processed.

9. A request processing method, comprising:

The service distribution server receives the user request; The service distribution server allocates response time constraint information to the user request, where the response time constraint information is used to mark a processing constraint time requested by the user and a processing time that has occurred;

The service distribution server generates a remote procedure call RPC request according to the user request; the service distribution server sends an RPC request to the service processing server, where the RPC request carries the response time constraint information; The response time constraint information sets an execution priority for the RPC request, and/or allocates an execution resource.

The method according to claim 9, wherein the allocating response time constraint information of the user request comprises:

And assigning response time constraint information to the user request according to the historical value of the response time constraint information and the hardware information of the service processing server.

11. The method of claim 10, wherein

If the hardware performance of the current service processing server changes with respect to the historical value of the response time constraint information, the historical value according to the response time constraint information and the hardware information of the service processing server are The user requests to allocate response time constraint information, including:

Obtaining a ratio of the first processing speed to the second processing speed, multiplying the historical value of the response time constraint information by the ratio, and obtaining response time constraint information currently allocated for the user request; the first processing speed is The processing speed of the hardware processing performance of the current service processing server to the RPC request, and the second processing speed is the processing speed of the hardware processing performance of the service processing server to the RPC request when the historical value of the response time constraint information is recorded. ;

If the hardware performance of the current service processing server does not change with respect to the historical value of the response time constraint information, the historical value according to the response time constraint information and the hardware information of the service processing server are The user requesting the allocation of the response time constraint information includes: using the historical value of the response time constraint information as the response time constraint information currently allocated for the user request.

The method according to any one of claims 9 to 11, wherein the generating a remote procedure call RPC request according to a user request comprises:

Extracting business parameters required to complete the user request;

Determining, according to preset business logic, a step required to complete the user request, and determining a service processing server to be invoked according to the step to be performed;

Allocating corresponding service parameters to the service processing server that needs to be invoked, respectively, generating and The service processing server corresponds to an RPC request.

The method according to any one of claims 9 to 11, wherein after the sending the RPC request to the service processing server, the method includes:

Receiving response time constraint information fed back by the service processing server;

The response time constraint information of the corresponding RPC request is updated using the response time constraint information.

The method according to claim 13, wherein, after receiving the response time constraint information fed back by the service processing server, the method includes:

And selecting, according to the response time constraint information and the emergency allocation threshold, an allocation processing policy corresponding to the user request, where the allocation processing policy includes: a first allocation processing policy and a second allocation processing policy, if the response time constraint information is If the constraint remaining time is greater than the emergency processing threshold, the first allocation processing policy is selected to be executed, and if the constraint remaining time of the response time constraint information is less than or equal to the emergency processing threshold, then the second allocation is selected to be performed. The processing strategy, the processing complexity of the first allocation processing policy is greater than the second allocation processing policy.

15. A service processing server, comprising:

a request receiving unit, configured to receive a remote procedure call RPC request sent by the service allocation server, where the RPC request includes: response time constraint information; the response time constraint information is used to mark a processing constraint of the user request corresponding to the RPC request Time and processing time has occurred;

a storage unit, configured to add the RPC request to a service request queue;

And a setting unit, configured to set an execution priority of the respective RPC request according to response time constraint information of each RPC request in the service request queue, and/or allocate an execution resource of the respective RPC request.

The service processing server according to claim 15, wherein the setting unit is specifically configured to:

If the service processing server cannot simultaneously process all RPCs in the service request queue The request is not only able to separately process any one of the RPC requests in the service request queue, and the service processing server sets the execution priority of each RPC request according to the response time constraint information of each RPC request in the service request queue. Level, and allocate the execution resources of the respective RPC requests.

The service processing server according to claim 15 or 16, wherein the setting unit comprises:

a resource allocation module, configured to acquire a constraint remaining time of the processing constraint time of each RPC request according to the response time constraint information; and set an execution priority of each RPC request according to a predefined rule, where the predefined rule includes: The less the constraint remaining time, the higher the execution priority of the setting;

a priority setting module, configured to acquire, according to the response time constraint information, a remaining time of the processing constraint time of the respective RPC request; and the execution resource of each RPC request is allocated according to a predefined rule, where the predefined rule includes: The less constraint remaining time, the more execution resources are allocated.

The service processing server according to any one of claims 15 to 17, wherein the service processing server further comprises:

a processing policy selection unit, configured to select according to the response time constraint information and the emergency threshold value

The processing policy of the RPC request, the processing policy includes: a first processing policy and a second processing policy, if the constraint remaining time of the response time constraint information is greater than the emergency threshold, selecting to execute the first processing policy, And if the constraint remaining time of the response time constraint information is less than or equal to the emergency threshold, the second processing policy is selected to be executed, and the processing complexity of the first processing policy is greater than the second processing policy.

a request processing unit, configured to process the RPC request with the highest priority, record the processing time of the RPC request, and update the response time constraint information of the RPC request according to the processing time;

And an information feedback unit, configured to feed back, to the service distribution server, response time constraint information of the RPC request that is processed.

20. A service distribution server, comprising:

a user request receiving unit, configured to receive a user request;

An information distribution unit, configured to allocate response time constraint information to the user request, where the response time The inter-constraint information is used to mark the processing constraint time of the user request and the processing time has occurred; the request generating unit is configured to generate a remote procedure call RPC request according to the user request; and the request sending unit is configured to send the RPC to the service processing server Requesting, the RPC request carries the response time constraint information; causing the service processing server to set an execution priority for the RPC request according to the response time constraint information, and/or allocate an execution resource.

The service distribution server according to claim 20, wherein the information distribution unit is specifically configured to:

22. The service distribution server according to claim 21, wherein the information distribution unit comprises:

a first allocation module, configured to: if a hardware performance of the current service processing server is changed relative to a historical value of the response time constraint information, obtain a ratio of the first processing speed to the second processing speed, The historical value of the response time constraint information is multiplied by the ratio to obtain response time constraint information currently allocated for the user request; the first processing speed is processing of the RPC request by the hardware performance of the current service processing server. Speed, the second processing speed is a processing speed of the hardware processing performance of the service processing server to the RPC request when the historical value of the response time constraint information is recorded;

a second allocation module, configured to: if the hardware performance of the current service processing server is not changed when the historical value of the response time constraint information is recorded, the historical value of the response time constraint information is used as the current The response time constraint information requested by the user.

The service distribution server according to any one of claims 20 to 22, wherein the request generating unit comprises:

a parameter extraction module, configured to extract a service parameter required to complete the user request;

a logic analysis module, configured to determine, according to the preset service logic, a step that needs to be performed to complete the user request, and determine, according to the step that needs to be performed, a service processing server that needs to be invoked;

And a request generating module, configured to respectively allocate a corresponding service parameter to the service processing server that needs to be invoked, and generate an RPC request corresponding to the service processing server.

The service distribution server according to any one of claims 20 to 22, wherein the service processing server further comprises: An information receiving unit, configured to receive response time constraint information fed back by the service processing server; and an information updating unit, configured to update response time constraint information of the corresponding RPC request by using the response time constraint information.

The service distribution server according to claim 24, wherein the service distribution server further comprises:

The allocation policy selection unit is configured to select an allocation processing policy corresponding to the user request according to the response time constraint information and the emergency allocation threshold, where the allocation processing policy includes: a first allocation processing policy and a second allocation processing policy, if If the constraint remaining time of the response time constraint information is greater than the emergency processing threshold, the first allocation processing policy is selected to be executed, and if the constraint remaining time of the response time constraint information is less than or equal to the emergency processing threshold, Then, the second allocation processing policy is selected to be executed, and the processing complexity of the first allocation processing policy is greater than the second allocation processing policy.

26. A request processing system, comprising:

Service distribution server and business processing server;

The service distribution server is configured to receive a user request, and allocate response time constraint information for the user request, where the response time constraint information is used to mark a processing constraint time of the user request and a processing time that has occurred; Generating a remote procedure call RPC request; sending an RPC request to the service processing server; the RPC request carrying the response time constraint information;

The service processing server is configured to receive a remote procedure call RPC request sent by the service distribution server, add the RPC request to the service request queue, and set the RPC request according to the response time constraint information of each RPC request in the service request queue. Execution priority, and/or allocation of execution resources for the respective RPC requests.