WO2021197128A1

WO2021197128A1 - Traffic rate-limiting method and apparatus

Info

Publication number: WO2021197128A1
Application number: PCT/CN2021/082307
Authority: WO
Inventors: 邹勇
Original assignee: 阿里巴巴集团控股有限公司
Priority date: 2020-03-30
Filing date: 2021-03-23
Publication date: 2021-10-07
Also published as: CN113472681A

Abstract

The present invention relates to the technical field of Internet. Disclosed are a traffic rate-limiting method and apparatus, capable of solving the problems in the prior art that the efficiency of rate-limiting is low and the rate-limiting is inaccurate. The method in the present invention mainly comprises: receiving data to be processed that is transmitted; determining whether a total number of first tokens in a preset number of sub-token buckets is greater than a total number of second tokens in a global token bucket, different cores corresponding to different sub-token buckets, and the tokens in the global token bucket being periodically increased; if the total number of the first tokens is greater than the total number of the second tokens, discarding the data to be processed; and if the total number of the first tokens is less than or equal to the total number of the second tokens, retaining the data to be processed, and adding one token to the sub-token bucket of the current core. The present invention is mainly applicable to a scenario of performing traffic control on the basis of a token bucket under a multi-core system.

Description

Flow rate limiting method and device

This application claims the priority of the Chinese patent application with the application number 202010238135.7 and the invention title of "Flow Rate Limiting Method and Device" filed on March 30, 2020, the entire content of which is incorporated into this application by reference.

Technical field

The present invention relates to the field of Internet technology, in particular to a flow rate limiting method and device.

Background technique

Token bucket is a common flow measurement technology, which is often used for flow restriction and flow shaping. It can measure the rate and burst of flow. With the popularization of Internet technology, the number of Internet users has greatly increased. In order to process the data sent by users in a timely manner, the server often uses a multi-core concurrent system to restrict traffic.

Currently, in a multi-core concurrent system, when traffic is restricted based on token buckets, the following two methods are usually used:

(1) Establish a global token bucket and lock the global token bucket so that only one kernel is allowed to update the global token bucket at the same time, so as to avoid multiple kernels from concurrently updating the global token bucket, resulting in high Frequent data write coverage makes the speed limit inaccurate. However, due to the addition of a lock mechanism, this method has problems such as lock waiting and cache flushing waiting, which leads to a decrease in the rate-limiting efficiency.

(2) Establish a token bucket for each core respectively, and put tokens into the token bucket at a specific token addition rate for use by the corresponding core. Although this method does not have the flaws of the lock mechanism, there is no definite rule for the rate at which the data to be processed enters each core, so the core with a large amount of data to be processed has already discarded the data because there is no token, and it is waiting to be processed. The kernel with a small amount of data still has a large number of unused tokens, which results in very inaccurate speed limit at the whole machine level.

It can be seen that how to provide a speed limiting method with relatively high speed limiting efficiency and accuracy is an urgent need to be solved.

Summary of the invention

In view of this, the flow rate limiting method and device provided by the present invention aims to solve the problems of low rate limiting efficiency and inaccurate rate limiting in the prior art.

In the first aspect, the present invention provides a traffic rate limiting method, the method includes:

Receive and send pending data;

Determine whether the total number of first tokens in the preset number of sub-token buckets is greater than the total number of second tokens in the global token bucket; different cores correspond to different sub-token buckets, and the tokens in the global token bucket Is increasing periodically;

If the total number of first tokens is greater than the total number of second tokens, discarding the to-be-processed data;

If the total number of first tokens is less than or equal to the total number of second tokens, the data to be processed is retained, and a token is added to the sub-token bucket of the current kernel.

Optionally, before determining whether the total number of first tokens in the preset number of sub-token buckets is greater than the total number of second tokens in the global token bucket, the method further includes:

Determine whether the global continuous packet loss status is open;

The determining whether the total number of first tokens in the preset number of sub-token buckets is greater than the total number of second tokens in the global token bucket includes:

If the global continuous packet loss state is the off state, it is determined whether the total number of first tokens is greater than the total number of second tokens.

Optionally, if the total number of first tokens is greater than the total number of second tokens, the method further includes:

The global continuous packet loss state is set to an open state.

Optionally, if the global continuous packet loss state is an on state, the method further includes:

Discard the to-be-processed data.

If it is determined that the global token bucket enters the next cycle and the token increase operation has been completed for the next cycle, then the global continuous packet loss state is set to a closed state;

Alternatively, when the duration of the global continuous packet loss state being the on state is a preset duration, the global continuous packet loss state is set to the off state.

Determine whether the number of tokens in the sub-token bucket of the current kernel is a multiple of M; M is a positive integer;

If the number of tokens is a multiple of M, it is determined whether the total number of the first tokens is greater than the total number of the second tokens.

Optionally, if the number of tokens in the sub-token bucket of the current kernel is not a multiple of M, the method further includes:

The data to be processed is retained, and a token is added to the sub-token bucket of the current kernel.

Optionally, the method further includes:

For each cycle, determine whether the difference between the total number of second tokens and the total number of first tokens is greater than the committed burst size;

If it is greater than the committed burst size, update the global token bucket so that the total number of second tokens in the global token bucket is the sum of the total number of first tokens and the acceptable burst size;

If it is less than or equal to the committed burst size, N tokens are added to the global token bucket, where N is the product of the committed information rate and the period.

Optionally, the bucket depth of the sub-token bucket and/or the global token bucket is infinite bucket depth.

In the second aspect, the present invention provides a flow rate limiting device, the device comprising:

The receiving unit is used to receive the sent data to be processed;

The judging unit is used to judge whether the total number of first tokens in the preset number of sub-token buckets is greater than the total number of second tokens in the global token bucket; different kernels correspond to different sub-token buckets, and the global token The tokens in the bucket are periodically increased;

A discarding unit, configured to discard the to-be-processed data if the total number of first tokens is greater than the total number of second tokens;

A retention unit, configured to retain the to-be-processed data if the total number of the first tokens is less than or equal to the total number of the second tokens;

The adding unit is configured to add a token to the sub-token bucket of the current kernel if the data to be processed is retained.

Optionally, the judging unit is further configured to judge the global continuous packet loss state before judging whether the total number of first tokens in the preset number of sub-token buckets is greater than the total number of second tokens in the global token bucket Whether it is an open state; if the global continuous packet loss state is an off state, it is determined whether the total number of the first tokens is greater than the total number of the second tokens.

Optionally, the device further includes:

The first setting unit is configured to set the global continuous packet loss state to an open state if the total number of first tokens is greater than the total number of second tokens.

Optionally, the discarding unit is further configured to discard the to-be-processed data if the global continuous packet loss state is an open state.

Optionally, the device further includes:

The second setting unit is configured to: if the global continuous packet loss status is the open state, if it is determined that the global token bucket enters the next cycle and the token increase operation has been completed for the next cycle, then The global continuous packet loss state is set to the off state; or, when the duration of the global continuous packet loss state being the on state is a preset duration, the global continuous packet loss state is set to the off state.

Optionally, the determining unit is further configured to determine whether the total number of first tokens in the preset number of sub-token buckets is greater than the total number of second tokens in the global token bucket, to determine the sub-commands of the current kernel Whether the number of tokens in the bucket is a multiple of M; M is a positive integer; if the number of tokens is a multiple of M, it is determined whether the total number of first tokens is greater than the total number of second tokens.

Optionally, the retaining unit is further configured to retain the to-be-processed data if the number of tokens in the sub-token bucket of the current kernel is not a multiple of M.

Optionally, the judging unit is further configured to judge whether the difference between the total number of second tokens and the total number of first tokens is greater than the committed burst size for each cycle;

The adding unit is further configured to update the global token bucket if it is greater than the committed burst size, so that the total number of second tokens of the global token bucket is the total number of the first token and the total number of tokens. If the sum of burst sizes is assumed; if it is less than or equal to the committed burst size, then N tokens are added to the global token bucket, where N is the product of the committed information rate and the period.

In a third aspect, the present invention provides a storage medium storing a plurality of instructions, and the instructions are suitable for being loaded by a processor and executing the traffic rate limiting method as described in the first aspect.

In a fourth aspect, the present invention provides an electronic device, the electronic device including a storage medium and a processor;

The processor is suitable for implementing various instructions;

The storage medium is suitable for storing multiple instructions;

The instructions are adapted to be loaded by the processor and execute the method according to the first aspect.

With the above technical solutions, the traffic rate limiting method and device provided by the present invention can first create a global token bucket and create a sub-token bucket for each core respectively, and then periodically add commands to the global token bucket. For each sub-token bucket, when the current kernel receives the data to be processed, the total number of tokens passing through the preset number of sub-token buckets (the total number of first tokens) cannot exceed the total number of tokens in the global token bucket ( The second total number of tokens) is a limiting condition to determine whether to add tokens to the sub-token bucket of the current kernel (that is, if the total number of first tokens is greater than the total number of second tokens, the data to be processed will be discarded, and the current kernel will not Add tokens in the sub-token bucket of the, and if the total number of the first token is less than or equal to the total number of the second token, the data to be processed is retained, and a token is added to the sub-token bucket of the current kernel), so that no It may happen that there are far more unused tokens in a certain sub-token bucket than in other sub-token buckets, thereby achieving balanced control of the whole machine and improving the accuracy of the speed limit of the whole machine. And because the lock mechanism is not used, there will be no lock waiting and cache flush waiting problems caused by the lock mechanism. Compared with the prior art using the lock mechanism, the efficiency of the rate limit is greatly improved, and the system is improved. performance.

The above description is only an overview of the technical solution of the present invention. In order to understand the technical means of the present invention more clearly, it can be implemented in accordance with the content of the specification, and in order to make the above and other objectives, features and advantages of the present invention more obvious and understandable. In the following, specific embodiments of the present invention are specifically cited.

Description of the drawings

By reading the detailed description of the preferred embodiments below, various other advantages and benefits will become clear to those of ordinary skill in the art. The drawings are only used for the purpose of illustrating the preferred embodiments, and are not considered as a limitation to the present invention. Also, throughout the drawings, the same reference symbols are used to denote the same components. In the attached picture:

Figure 1 shows a flow chart of a flow rate limiting method provided by an embodiment of the present invention;

Figure 2 shows a flow chart of another method for traffic rate limiting according to an embodiment of the present invention;

FIG. 3 shows a flowchart of yet another method for traffic rate limiting according to an embodiment of the present invention;

FIG. 4 shows a flow chart of still another method for traffic rate limiting according to an embodiment of the present invention;

Figure 5 shows a block diagram of a flow rate limiting device provided by an embodiment of the present invention;

Fig. 6 shows a block diagram of another flow rate limiting device provided by an embodiment of the present invention.

Detailed ways

Hereinafter, exemplary embodiments of the present disclosure will be described in more detail with reference to the accompanying drawings. Although the drawings show exemplary embodiments of the present disclosure, it should be understood that the present disclosure can be implemented in various forms and should not be limited by the embodiments set forth herein. On the contrary, these embodiments are provided to enable a more thorough understanding of the present disclosure and to fully convey the scope of the present disclosure to those skilled in the art.

In order to improve the speed limit efficiency and the speed limit accuracy rate, the embodiment of the present invention provides a combination of a global token bucket and each kernel sub-token bucket, and uses the global token bucket to control the flow limit of the number of tokens in the sub-token bucket. Fast method, the method is mainly applied to a single core side, as shown in Figure 1, the method mainly includes:

101. Receive and send data to be processed.

The embodiments of the present invention can be applied to any kind of network project. A multi-core network device with a safe speed limit function can be set in the project, and the network device can perform a speed limit operation on the data to be processed from the outside. Among them, the data to be processed can be a message or a data packet in the data format, and can be a request or non-request in the content. The embodiment of the present invention does not deal with the specific data format and data content of the interaction. limited. For example, in the field of processing messages, the embodiments of the present invention can implement restrictions on the frequency of message sending, and can also implement restrictions on the frequency of message forwarding.

102. Determine whether the total number of first tokens in the preset number of sub-token buckets is greater than the total number of second tokens in the global token bucket; if the total number of first tokens is greater than the total number of second tokens, then Go to step 103; if the total number of first tokens is less than or equal to the total number of second tokens, go to step 104.

Among them, each core corresponds to a sub-token bucket, and different cores correspond to different sub-token buckets. The tokens in the global token bucket are added periodically (one of the kernels can be designated to perform the token increase operation on the global token bucket), and the tokens in the sub-token bucket are executed by performing steps 101-104 Way to increase. The bucket depth of the global token bucket and/or sub-token bucket can be infinite bucket depth, that is, the number of tokens can only increase without decreasing, and can increase indefinitely.

In order to use the global token bucket to control the flow of the sub-token buckets, thereby improving the rate limiting efficiency and accuracy, the total number of first tokens accumulated in the preset number of sub-token buckets can be combined with the total number of first tokens in the global token bucket. The total number of the two tokens is compared, and the discarding and retention of the data to be processed is determined through different comparison results. In practical applications, the preset number can be determined according to the requirements of speed limit efficiency and speed limit accuracy. The higher the speed limit accuracy requirement, the greater the preset number. For example, when the preset number is the sum of all sub-token buckets, the speed limit accuracy rate is higher than that of other numbers, but it is higher than other number limits. The speed efficiency may be lower. In specific implementation, the sub-token bucket that needs to be accumulated to obtain the first total number of tokens can be determined according to a preset condition. The preset conditions include, but are not limited to, sub-token buckets with a number of tokens greater than a preset threshold, and sub-token buckets with a number of tokens in the top N (N is a positive integer).

For the total number of first tokens accumulated by the preset number of sub-token buckets, a total number of first tokens can be stored in the preset storage space. If a token is added to a certain sub-token bucket, update it When the total number of first tokens stored in the preset storage space needs to be compared with the total number of second tokens, the total number of first tokens can be directly read from the preset storage space. You can also create a temporary token bucket, add a token to the sub-token bucket, add a token to the temporary token bucket, and when you need to compare with the total number of second tokens, count the temporary token bucket The number of tokens as the first total number of tokens.

The tokens of the sub-token bucket are increased according to the data to be processed. If there is no data to be processed for a long time, or the data to be processed is received for a long time, the tokens of the global token bucket will always be Increase periodically, which will cause the total number of second tokens in the global token bucket to be much larger than the total number of first tokens accumulated by the preset number of sub-token buckets. The judgment mechanism of the embodiment of the invention allows the data to be processed to be retained, which may cause a sudden increase in the load of the kernel and cause failures such as downtime. To avoid this technical problem, the following methods can be used to periodically add tokens to the global token bucket:

For each cycle, determine whether the difference between the total number of second tokens and the total number of first tokens is greater than the committed burst size (Committed Burst Size, CBS); if it is greater than the committed burst size, update the Global token bucket, so that the total number of second tokens in the global token bucket is the sum of the total number of first tokens and the acceptable burst size; if it is less than or equal to the committed burst size, then N tokens are added to the global token bucket, where N is the product of the committed information rate (Committed Information Rate, CIR) and the period.

Among them, the committed burst size is also called the committed burst volume, which refers to the total amount of data that the network allows to transmit on the virtual circuit in a normal state and within a certain time interval. In the embodiment of the present invention, by controlling the total number of second tokens in the global token bucket not to exceed the sum of the total number of first tokens and the burst size, the total amount of data to be processed that is allowed to be retained within a period of time can be controlled. It will not exceed the amount of data represented by the burst size, so that the burst size is limited, and the sudden increase will prevent the core from being unable to withstand the failure.

The committed information rate is the virtual circuit information transmission rate under the normal state agreed by the user and the network. If the difference between the total number of second tokens and the total number of first tokens is less than or equal to the committed burst size, tokens can be added to the global token bucket according to the committed information rate, that is, the increase allowed in each cycle The number of cards N can be the product of the committed information rate and the period. When the committed information rate is represented by CIR and the period is represented by T, N=CIR*T.

What needs to be added is that when allocating memory for the sub-token buckets of each kernel, it can be separated and decoupled as much as possible to minimize conflicts occupied by the cache.

103. Discard the to-be-processed data.

If the total number of first tokens is greater than the total number of second tokens, the function of controlling the sub-token bucket of the global token bucket can be achieved by discarding the data to be processed, thereby realizing the speed limit of the whole machine.

104. Keep the data to be processed, and add a token to the sub-token bucket of the current kernel.

If the total number of first tokens is less than or equal to the total number of second tokens, the total number of tokens in the preset number of sub-token buckets has not yet reached the online allowed by the whole machine, so the received tokens this time can be received The data to be processed is retained, and by adding a token to the sub-token bucket of the current kernel, the added token is allocated to the data to be processed. Wherein, the specific implementation manner of retaining the data to be processed may be to add the data to be processed into the cache queue corresponding to the current kernel, so that it can be subsequently processed.

The traffic rate limiting method provided by the embodiment of the present invention can first create a global token bucket and create a sub-token bucket for each core respectively, and then periodically add tokens to the global token bucket, and for each sub-token bucket Token bucket, when the current kernel receives data to be processed, the total number of tokens passing through the preset number of sub-token buckets (the total number of first tokens) cannot exceed the total number of tokens in the global token bucket (the total number of second tokens) As a restriction, determine whether to add tokens to the sub-token bucket of the current kernel (that is, if the total number of the first token is greater than the total number of the second token, the data to be processed will be discarded, and will not be added to the sub-token bucket of the current kernel. Add tokens, and if the total number of first tokens is less than or equal to the total number of second tokens, the data to be processed is retained, and a token is added to the sub-token bucket of the current kernel, so that a certain sub-token will not occur The phenomenon that there are far more unused tokens in the bucket than other sub-token buckets, thereby realizing the balanced control of the whole machine, and improving the accuracy of the speed limit of the whole machine. And because the lock mechanism is not used, there will be no lock waiting and cache flush waiting problems caused by the lock mechanism. Compared with the prior art using the lock mechanism, the efficiency of the rate limit is greatly improved, and the system is improved. performance.

Optionally, in order to further improve the rate limiting efficiency, the embodiment of the present invention provides the following three methods:

Method 1: As shown in Figure 2, the method mainly includes:

201. Receive and send data to be processed.

The specific implementation of this step is the same as that of step 101, and will not be repeated here.

202. Determine whether the global continuous packet loss state is an open state.

Among them, the global continuous packet loss status is a parameter used to control global packet loss, that is, when it is turned on, it is used to instruct each core to discard new data to be processed after it is received, and there is no need to send it to Add tokens to the sub-token bucket.

Since the global continuous packet loss status is a global parameter, and when it is in the open state, no matter whether the total number of first tokens in the preset number of sub-token buckets is greater than the total number of second tokens in the global token bucket, the kernel Both need to discard the newly received pending data, so after receiving the sent pending data, first judge whether the global continuous packet loss status is the open state to avoid wasting time and resources to judge the total number of tokens.

203. If the global continuous packet loss state is an open state, discard the to-be-processed data.

After the global continuous packet loss status is turned on, the newly received pending data will be discarded, so the sub-token bucket will also stop adding tokens, but the global token bucket will also periodically add tokens, so if If the global continuous packet loss status is always turned on, the total number of second tokens in the global token bucket is much larger than the total number of first tokens accumulated by the preset number of sub-token buckets, which will happen Unable to respond to the sudden increase in new data in a timely manner. In order to solve the problem, under certain conditions, the global continuous packet loss state can be set to the off state. Among them, a relatively good time is: if it is determined that the global token bucket enters the next cycle, and the token increase operation has been completed for the next cycle, then the global continuous packet loss state is set to the off state . In the case of little impact on the accuracy of the overall speed limit, you can also choose another time to set the global continuous packet loss state to the off state. For example, you can set the global continuous packet loss state to the on state for a preset duration. At this time, the global continuous packet loss state is set to the off state.

204. If the global continuous packet loss state is the off state, determine whether the total number of first tokens in the preset number of sub-token buckets is greater than the total number of second tokens in the global token bucket.

If the global continuous packet loss status is off, there is no mandatory control on packet loss globally. Therefore, it is necessary to determine whether to discard the data to be processed or keep the data to be processed by judging the total number of the first token and the total number of the second token. Data processing.

205. If the total number of first tokens is greater than the total number of second tokens, discard the to-be-processed data, and set the global continuous packet loss state to an open state.

Because when tokens are added periodically in the global token bucket, in a cycle, it may take the first part of the cycle to complete the token increase operation, and during the remaining time of the cycle, the global token The total number of second tokens in the bucket will not change, so during the remaining time, once the total number of first tokens accumulated by the preset number of sub-token buckets is greater than the total number of second tokens, then From the moment (may be called the first occurrence moment) until the next token is added to the global token bucket, no matter whether the judgment is made, the conclusion is that the total number of first tokens is greater than the total number of second tokens, and it needs to be processed The data is discarded. In order to save judgment resources and judgment time, the global continuous packet loss status can be set to the on state, so that when new data to be processed is subsequently received, the global continuous packet loss status can be directly discarded.

Wherein, when the first judgment result is that the total number of first tokens is greater than the total number of second tokens, the global continuous packet loss state may be set to the on state, or it may be set to the open state at the Xth (X is greater than 1) time The judgment result is that when the total number of first tokens is greater than the total number of second tokens, the global continuous packet loss status is set to the open state, as long as the global continuous packet loss status can be guaranteed to be set to the open state. The timing can save judgment resources and judgment time to a certain extent from the moment of the first occurrence until the next time before adding tokens to the global token bucket, thereby improving the limited efficiency to a certain extent.

206. If the total number of first tokens is less than or equal to the total number of second tokens, retain the data to be processed, and add a token to the sub-token bucket of the current kernel.

The specific implementation of this step is the same as that of step 104, and will not be repeated here.

The traffic rate limiting method provided by the embodiment of the present invention is based on the embodiment shown in FIG. , And when the global continuous packet loss status is on, the data to be processed is directly discarded, which can reduce the number of judgments for loss/retention decisions, and further improve the speed limit efficiency on the basis of improving the accuracy of the overall speed limit.

Method 2: As shown in Figure 3, the method mainly includes:

301. Receive and send data to be processed.

302. Determine whether the number of tokens in the sub-token bucket of the current kernel is a multiple of M; if the number of tokens is not a multiple of M, perform step 303; if the number of tokens is a multiple of M, perform step 304.

Among them, M is a positive integer, and the specific value of M can be determined based on historical experience.

Since the number of tokens in the sub-token bucket and global token bucket of each core is cumulative, the total number of tokens is judged every M pieces of data to be processed (that is, the decision to lose/leave), and the middle to be processed When the data is directly retained, even if Y data is received in a certain period (M data is a period), in the following periods, the new arrival will be discarded due to the limit of the total number of second tokens to the total number of first tokens. Data, that is, Y data will be reduced in the subsequent cycle, so for the whole machine, its speed limit accuracy rate is almost unchanged. Therefore, when receiving the data to be processed, you can first determine whether the number of tokens in the sub-token bucket of the current kernel is a multiple of M, if not, you can directly keep it and send it to the sub-token bucket of the current kernel. Add a token (that is, step 303), if it is, then perform a judgment of the total number of tokens, and make a decision to discard/retain the data to be processed based on the result of the judgment of the total number of tokens (that is, step 304->305->303).

303. Keep the data to be processed, and add a token to the sub-token bucket of the current kernel.

304. Determine whether the total number of first tokens in the preset number of sub-token buckets is greater than the total number of second tokens in the global token bucket; if the total number of first tokens is greater than the total number of second tokens, then Go to step 305; if the total number of first tokens is less than or equal to the total number of second tokens, go to step 303.

305. Discard the to-be-processed data.

According to the flow rate limiting method provided by the embodiment of the present invention, on the basis of the embodiment shown in FIG. 1, a drop/leave decision is made every M pieces of to-be-processed data, and the intermediate to-be-processed data can be directly reserved. Reduce the number of times of decision-making for loss/retention, which can further improve the speed limit efficiency on the basis of improving the accuracy of the whole machine speed limit.

Method 3: Since the first method is to improve the rate-limiting efficiency from a global level, and the second method is to improve the rate-limiting efficiency from a single-core level, and there is no dependence and contradiction between the two, the two can be combined to form the third method to further To improve the speed limit efficiency, as shown in Figure 4, the method includes:

401. Receive sent data to be processed.

402. Determine whether the global continuous packet loss state is an open state; if the global continuous packet loss state is an open state, perform step 403; if the global continuous packet loss state is an off state, perform step 404.

403. Discard the to-be-processed data.

404. Determine whether the number of tokens in the sub-token bucket of the current kernel is a multiple of M; if the number of tokens is not a multiple of M, perform step 405; if the number of tokens is a multiple of M, perform step 406.

405. Keep the data to be processed, and add a token to the sub-token bucket of the current kernel.

406. Determine whether the total number of first tokens in the preset number of sub-token buckets is greater than the total number of second tokens in the global token bucket; if the total number of first tokens is greater than the total number of second tokens, then Go to step 407; if the total number of first tokens is less than or equal to the total number of second tokens, go to step 405.

407. Discard the to-be-processed data, and set the global continuous packet loss state to an open state.

Further, according to the above method embodiment, another embodiment of the present invention also provides a flow rate limiting device. As shown in Fig. 5, the device includes:

The receiving unit 51 is configured to receive sent data to be processed;

The judging unit 52 is configured to judge whether the total number of first tokens in the preset number of sub-token buckets is greater than the total number of second tokens in the global token bucket; different kernels correspond to different sub-token buckets, and the global command The tokens in the bucket are increased periodically;

The discarding unit 53 is configured to discard the to-be-processed data if the total number of the first tokens is greater than the total number of the second tokens;

The retaining unit 54 is configured to retain the data to be processed if the total number of the first tokens is less than or equal to the total number of the second tokens;

The adding unit 55 is configured to add a token to the sub-token bucket of the current kernel if the data to be processed is retained.

Optionally, the judging unit 52 is further configured to judge whether the total number of first tokens in the preset number of sub-token buckets is greater than the total number of second tokens in the global token bucket, to determine the global continuous packet loss Whether the state is an open state; if the global continuous packet loss state is an off state, it is determined whether the total number of first tokens is greater than the total number of second tokens.

Optionally, as shown in FIG. 6, the device further includes:

The first setting unit 56 is configured to set the global continuous packet loss state to an open state if the total number of first tokens is greater than the total number of second tokens.

Optionally, the discarding unit 53 is further configured to discard the to-be-processed data if the global continuous packet loss state is an open state.

Optionally, as shown in FIG. 6, the device further includes:

The second setting unit 57 is configured to, if the global continuous packet loss state is the open state, if it is determined that the global token bucket enters the next cycle and the token increase operation has been completed for the next cycle, then The global continuous packet loss state is set to the off state; or, when the duration of the global continuous packet loss state being the on state is a preset duration, the global continuous packet loss state is set to the off state.

Optionally, the judging unit 52 is further configured to judge whether the total number of first tokens in the preset number of sub-token buckets is greater than the total number of second tokens in the global token bucket, before judging the sub-tokens of the current kernel. Whether the number of tokens in the token bucket is a multiple of M; M is a positive integer; if the number of tokens is a multiple of M, it is determined whether the total number of first tokens is greater than the total number of second tokens.

Optionally, the retaining unit 54 is further configured to retain the to-be-processed data if the number of tokens in the sub-token bucket of the current kernel is not a multiple of M.

Optionally, the judging unit 52 is further configured to judge whether the difference between the total number of second tokens and the total number of first tokens is greater than the committed burst size for each cycle;

The adding unit 55 is further configured to update the global token bucket if it is larger than the committed burst size, so that the total number of second tokens in the global token bucket is the total number of the first tokens and the total number of tokens. If the sum of burst sizes is assumed; if it is less than or equal to the committed burst size, then N tokens are added to the global token bucket, where N is the product of the committed information rate and the period .

The flow rate limiting device provided by the embodiment of the present invention can first create a global token bucket and create a sub-token bucket for each core respectively, and then periodically add tokens to the global token bucket, and for each sub-token bucket Token bucket, when the current kernel receives the data to be processed, the total number of tokens passing through the preset number of sub-token buckets (the total number of first tokens) cannot exceed the total number of tokens in the global token bucket (the total number of second tokens) It is a restriction condition to determine whether to add tokens to the sub-token bucket of the current kernel (that is, if the total number of the first token is greater than the total number of the second token, the data to be processed will be discarded, and will not be added to the sub-token bucket of the current kernel. Add tokens, and if the total number of first tokens is less than or equal to the total number of second tokens, the data to be processed is retained, and a token is added to the sub-token bucket of the current kernel, so that a certain sub-token will not occur The phenomenon that there are far more unused tokens in the bucket than other sub-token buckets, thereby realizing the balanced control of the whole machine, and improving the accuracy of the speed limit of the whole machine. And because the lock mechanism is not used, there will be no lock waiting and cache flush waiting problems caused by the lock mechanism. Compared with the prior art using the lock mechanism, the efficiency of the rate limit is greatly improved, and the system is improved. performance. On this basis, by enabling the global continuous packet loss state, and only when the global continuous packet loss state is off, the loss/retention decision is made, and when the global continuous packet loss state is on, the pending data is directly discarded , Which can reduce the number of judgments of the loss/retention decision. If every M pieces of data to be processed, a loss/retention decision is made, and the intermediate data to be processed is directly retained, which can also reduce the number of judgments of the loss/retention decision. Furthermore, on the basis of improving the accuracy of the speed limit of the whole machine, the speed limit efficiency can be further improved.

Further, another embodiment of the present invention also provides a storage medium, the storage medium stores a plurality of instructions, and the instructions are suitable for being loaded by a processor and executing the above-mentioned flow rate limiting method.

Further, another embodiment of the present invention also provides an electronic device, the electronic device including a storage medium and a processor;

The processor is suitable for implementing various instructions;

The storage medium is suitable for storing multiple instructions;

The instructions are adapted to be loaded by the processor and execute the traffic rate limiting method as described above.

In the above-mentioned embodiments, the description of each embodiment has its own emphasis. For parts that are not described in detail in an embodiment, reference may be made to related descriptions of other embodiments.

It can be understood that the relevant features in the above method and device can be referred to each other. In addition, the “first”, “second”, etc. in the foregoing embodiments are used to distinguish the embodiments, and do not represent the advantages and disadvantages of the embodiments.

Those skilled in the art can clearly understand that, for the convenience and conciseness of the description, the specific working process of the above-described system, device, and unit can refer to the corresponding process in the foregoing method embodiment, which will not be repeated here.

The algorithms and displays provided here are not inherently related to any particular computer, virtual system or other equipment. Various general-purpose systems can also be used with the teaching based on this. Based on the above description, the structure required to construct this type of system is obvious. In addition, the present invention is not directed to any specific programming language. It should be understood that various programming languages can be used to implement the content of the present invention described herein, and the above description of a specific language is for the purpose of disclosing an embodiment of the present invention.

In the instructions provided here, a lot of specific details are explained. However, it can be understood that the embodiments of the present invention can be practiced without these specific details. In some instances, well-known methods, structures, and technologies are not shown in detail, so as not to obscure the understanding of this specification.

Similarly, it should be understood that in order to simplify the present disclosure and help understand one or more of the various inventive aspects, in the above description of the exemplary embodiments of the present invention, the various features of the present invention are sometimes grouped together into a single embodiment, Figure, or its description. However, the disclosed method should not be interpreted as reflecting the intention that the claimed invention requires more features than those explicitly stated in each claim. More precisely, as reflected in the following claims, the inventive aspect lies in less than all the features of a single embodiment disclosed previously. Therefore, the claims following the specific embodiment are thus explicitly incorporated into the specific embodiment, wherein each claim itself serves as a separate embodiment of the present invention.

Those skilled in the art can understand that it is possible to adaptively change the modules in the device in the embodiment and set them in one or more devices different from the embodiment. The modules or units or components in the embodiments can be combined into one module or unit or component, and in addition, they can be divided into multiple sub-modules or sub-units or sub-components. Except that at least some of such features and/or processes or units are mutually exclusive, any combination can be used to compare all the features disclosed in this specification (including the accompanying claims, abstract and drawings) and any method or methods disclosed in this manner or All the processes or units of the equipment are combined. Unless expressly stated otherwise, each feature disclosed in this specification (including the accompanying claims, abstract and drawings) may be replaced by an alternative feature providing the same, equivalent or similar purpose.

In addition, those skilled in the art can understand that although some embodiments described herein include certain features included in other embodiments but not other features, the combination of features of different embodiments means that they are within the scope of the present invention. Within and form different embodiments. For example, in the following claims, any one of the claimed embodiments can be used in any combination.

The various component embodiments of the present invention may be implemented by hardware, or by software modules running on one or more processors, or by a combination of them. Those skilled in the art should understand that a microprocessor or a digital signal processor (DSP) may be used in practice to implement some or all of the functions of some or all components in the traffic rate limiting method and device according to the embodiments of the present invention. The present invention can also be implemented as a device or device program (for example, a computer program and a computer program product) for executing part or all of the methods described herein. Such a program for realizing the present invention may be stored on a computer-readable medium, or may have the form of one or more signals. Such a signal can be downloaded from an Internet website, or provided on a carrier signal, or provided in any other form.

It should be noted that the above-mentioned embodiments illustrate rather than limit the present invention, and those skilled in the art can design alternative embodiments without departing from the scope of the appended claims. In the claims, any reference signs placed between parentheses should not be constructed as a limitation to the claims. The word "comprising" does not exclude the presence of elements or steps not listed in the claims. The word "a" or "an" preceding an element does not exclude the presence of multiple such elements. The invention can be implemented by means of hardware comprising several different elements and by means of a suitably programmed computer. In the unit claims listing several devices, several of these devices may be embodied in the same hardware item. The use of the words first, second, and third, etc. do not indicate any order. These words can be interpreted as names.

Claims

A flow rate limiting method, characterized in that the method includes:

Receive and send pending data;

Determine whether the total number of first tokens in the preset number of sub-token buckets is greater than the total number of second tokens in the global token bucket; different cores correspond to different sub-token buckets, and the tokens in the global token bucket Is increasing periodically;

If the total number of first tokens is greater than the total number of second tokens, discarding the to-be-processed data;

If the total number of first tokens is less than or equal to the total number of second tokens, the data to be processed is retained, and a token is added to the sub-token bucket of the current kernel.
The method according to claim 1, wherein before determining whether the total number of first tokens in the preset number of sub-token buckets is greater than the total number of second tokens in the global token bucket, the method further comprises :

Determine whether the global continuous packet loss status is open;

The determining whether the total number of first tokens in the preset number of sub-token buckets is greater than the total number of second tokens in the global token bucket includes:

If the global continuous packet loss state is the off state, it is determined whether the total number of first tokens is greater than the total number of second tokens.
The method according to claim 2, wherein if the total number of the first tokens is greater than the total number of the second tokens, the method further comprises:

The global continuous packet loss state is set to an open state.
The method according to claim 2, wherein if the global continuous packet loss state is an open state, the method further comprises:

Discard the to-be-processed data.
The method according to claim 2, wherein if the global continuous packet loss state is an open state, the method further comprises:

If it is determined that the global token bucket enters the next cycle and the token increase operation has been completed for the next cycle, then the global continuous packet loss state is set to a closed state;

Alternatively, when the duration of the global continuous packet loss state being the on state is a preset duration, the global continuous packet loss state is set to the off state.
The method according to claim 1, wherein before determining whether the total number of first tokens in the preset number of sub-token buckets is greater than the total number of second tokens in the global token bucket, the method further comprises :

Determine whether the number of tokens in the sub-token bucket of the current kernel is a multiple of M; M is a positive integer;

The determining whether the total number of first tokens in the preset number of sub-token buckets is greater than the total number of second tokens in the global token bucket includes:

If the number of tokens is a multiple of M, it is determined whether the total number of the first tokens is greater than the total number of the second tokens.
The method according to claim 6, wherein if the number of tokens in the sub-token bucket of the current kernel is not a multiple of M, the method further comprises:

The data to be processed is retained, and a token is added to the sub-token bucket of the current kernel.
The method according to claim 1, wherein the method further comprises:

For each cycle, determine whether the difference between the total number of second tokens and the total number of first tokens is greater than the committed burst size;

If it is greater than the committed burst size, update the global token bucket so that the total number of second tokens in the global token bucket is the sum of the total number of first tokens and the supported burst size;

If it is less than or equal to the committed burst size, then N tokens are added to the global token bucket, where N is the product of the committed information rate and the period.
The method according to any one of claims 1-8, wherein the bucket depth of the sub-token bucket and/or the global token bucket is an infinite bucket depth.
A flow rate limiting device, characterized in that the device includes:

The receiving unit is used to receive the sent data to be processed;

The judging unit is used to judge whether the total number of first tokens in the preset number of sub-token buckets is greater than the total number of second tokens in the global token bucket; different kernels correspond to different sub-token buckets, and the global token The tokens in the bucket are periodically increased;

A discarding unit, configured to discard the to-be-processed data if the total number of first tokens is greater than the total number of second tokens;

A retention unit, configured to retain the to-be-processed data if the total number of the first tokens is less than or equal to the total number of the second tokens;

The adding unit is configured to add a token to the sub-token bucket of the current kernel if the data to be processed is retained.
A storage medium, characterized in that the storage medium stores a plurality of instructions, and the instructions are suitable for being loaded by a processor and executing the flow rate limiting method according to any one of claims 1 to 9.
An electronic device, characterized in that, the electronic device includes a storage medium and a processor;

The processor is suitable for implementing various instructions;

The storage medium is suitable for storing multiple instructions;

The instructions are adapted to be loaded by the processor and execute the method according to any one of claims 1-9.