WO2021151293A1

WO2021151293A1 - Hash ring load balancing method and apparatus, and electronic device and storage medium

Info

Publication number: WO2021151293A1
Application number: PCT/CN2020/117659
Authority: WO
Inventors: 乐伟
Original assignee: 平安科技（深圳）有限公司
Priority date: 2020-08-27
Filing date: 2020-09-25
Publication date: 2021-08-05
Also published as: CN112015552A

Abstract

A hash ring load balancing method and apparatus, and an electronic device and a storage medium. The method comprises: virtualizing a preset number of virtual service nodes for each real service node, so as to increase the number of nodes on a hash ring; forming virtual service nodes, with the same serial number, virtualized for different real service nodes into a group, and performing sequencing according to the corresponding serial numbers of the real service nodes, such that the real service nodes corresponding to the virtual service nodes in each group are spaced; and then, connecting each group of virtual service nodes to form a full virtual service node queue according to the sequence of the serial numbers of the virtual service nodes, and uniformly distributing, according to the queue, all the virtual service nodes on the hash ring, such that not only are all the virtual service nodes uniformly distributed on the hash ring, but the various virtual service nodes virtualized for each real service node are also uniformly distributed on the hash ring.

Description

Hash ring balancing load method, device, electronic equipment and storage medium

This application claims the priority of a Chinese patent application filed with the Chinese Patent Office on August 27, 2020, the application number is 202010879098.8, and the invention title is "hash ring balancing load method, device, electronic equipment, and storage medium". The entire content is approved The reference is incorporated in this application.

Technical field

This application relates to the field of computer technology, and in particular to a method, device, electronic device, and storage medium for hash ring load balancing.

Background technique

The inventor found that the problem of unbalanced load exists when a consistent hash is currently used as load balancing. The main reason is that the hash value of the service node cannot be distributed evenly on the hash ring, which causes the request message sent to the service node to not be evenly distributed to each service node.

Summary of the invention

In order to solve the above uneven load problem, this application provides a hash ring load balancing method, including:

Virtualize each real service node into a number of equal number of virtual service nodes, and obtain the virtual service node corresponding to each real service node;

A virtual service node is selected from the virtual service nodes corresponding to multiple real service nodes to form a virtual service node group to obtain multiple virtual service node groups, each of the virtual service nodes in the virtual service node group The order of the corresponding real service nodes is the same;

Connect the virtual service nodes in each virtual service node group end to end to obtain a virtual service node queue;

The virtual service node queue is evenly distributed on the hash ring.

This application also provides a hash ring load balancing device including:

The virtual unit is used to virtualize each real service node into a number of equal number of virtual service nodes to obtain the virtual service node corresponding to each real service node;

The grouping unit is used to select a virtual service node from the virtual service nodes corresponding to the multiple real service nodes to form a virtual service node group to obtain multiple virtual service node groups, each of the virtual service node groups The order of the virtual service nodes corresponding to the real service nodes in the virtual service node is the same;

The connecting unit is used to connect the virtual service nodes in each virtual service node group end to end to obtain a virtual service node queue;

The distribution unit is configured to evenly distribute the virtual service node queue on the hash ring.

The present application also provides an electronic device, including a memory and a processor, the memory stores a computer program, and the processor implements the following steps when executing the computer program:

A virtual service node is selected from virtual service nodes corresponding to multiple real service nodes to form a virtual service node group to obtain multiple virtual service node groups, each of the virtual service nodes in the virtual service node group The order of the corresponding real service nodes is the same;

The virtual service node queue is evenly distributed on the hash ring.

This application also provides a readable storage medium on which a computer program is stored, and when the computer program is executed by a processor, the following steps are implemented:

The virtual service node queue is evenly distributed on the hash ring.

This application increases the number of nodes on the hash ring by virtualizing each real service node with an equal number of virtual service nodes; the virtual service nodes virtualized corresponding to different real service nodes form a group, In this way, the real service nodes corresponding to the virtual service nodes in each group are spaced; then, each group is connected end to end to obtain a virtual service node queue containing all virtual service nodes, and all virtual service nodes are sorted according to the queue Evenly distributed on the hash ring, so that not only all virtual service nodes can be evenly distributed on the hash ring, but also the virtual service nodes virtualized by each real service node are evenly distributed on the hash ring , So that the load of each service node on the hash ring is more balanced.

Description of the drawings

Figure 1 is a schematic diagram of the current distribution and load of service nodes on the consistent hash ring.

FIG. 2 is a schematic diagram of the steps of a method for balancing a load of a hash ring in an embodiment of the present application.

Fig. 3 is a schematic diagram of the distribution and load of virtual service nodes on the hash ring in an embodiment of the present application.

FIG. 4 is a schematic diagram of the distribution and load of virtual service nodes on the hash ring after capacity expansion based on the distribution shown in FIG. 3.

FIG. 5 is a schematic block diagram of the structure of a hash ring load balancing device in an embodiment of the present application.

FIG. 6 is a schematic block diagram of the structure of an electronic device in an embodiment of the present application.

Fig. 7 is a schematic diagram of a readable storage medium in an embodiment of the present application.

Detailed ways

The following will clearly and completely describe the technical solutions in the embodiments of the present application with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in this application, all other embodiments obtained by those of ordinary skill in the art without creative work shall fall within the protection scope of this application.

In the description of this application, it should be understood that the terms “first” and “second” are only used for description purposes, and cannot be understood as indicating or implying relative importance or implicitly indicating the number of indicated technical features. Thus, the features defined with "first" and "second" may explicitly or implicitly include one or more of these features. In the description of this application, unless otherwise specified, "plurality" means two or more.

In addition, the described features, structures, or characteristics may be combined in one or more embodiments in any suitable manner. In the following description, many specific details are provided to give a sufficient understanding of the embodiments of the present application. However, those skilled in the art will realize that the technical solutions of the present application can be practiced without one or more of the specific details, or other methods, components, devices, steps, etc. can be used. In other cases, well-known methods, devices, implementations or operations are not shown or described in detail in order to avoid obscuring various aspects of the present application.

The block diagrams shown in the drawings are merely functional entities, and do not necessarily correspond to physically independent entities. That is, these functional entities can be implemented in the form of software, or implemented in one or more hardware modules or integrated circuits, or implemented in different networks and/or processor devices and/or microcontroller devices. entity.

The flowchart shown in the drawings is only an exemplary description, and does not necessarily include all contents and operations/steps, nor does it have to be performed in the described order. For example, some operations/steps can be decomposed, and some operations/steps can be combined or partially combined, so the actual execution order may be changed according to actual conditions.

The technical solution of the present application can be applied to the field of blockchain and/or big data technology to achieve load balancing.

The design goal of the consistent Hash algorithm is to solve the hotspot problem in the Internet, and its original intention is very similar to CARP. The consistent Hash algorithm is a special Hash algorithm. Due to its balanced and durable mapping characteristics, it is widely used in the load balancing field. For example, nginx and memcached both use a consistent Hash algorithm as a cluster load balancing solution. . At present, when the consistent hash algorithm is used as a load balancing method, there are still some load imbalance problems. The main reasons are: first, the hash value of the service node cannot be distributed evenly on the hash ring; second, there are too few service nodes, and the hash value of the source request is more likely to be distributed on the same service node.

Referring to Figure 1, in the hash ring shown in Figure 1, there are three service nodes Node1, Node2, and Node3. When performing load balancing, you can use FNV1_64_HASH to calculate the service node hash value; a, b, and c are three respectively Request the corresponding request, also calculate the corresponding hash value through FNV1_64_HASH, and then a clockwise to find the nearest Node service node, as can be seen from Figure 1, because the interval between Node3 and Node2 is larger, the greater the probability of the request will be loaded to Node1 node Above, both requests c and a will be loaded on the Node1 node, the request b will be loaded on the Node2 node, and there is no request from the Node3 node. It can be found that this kind of consistent hash algorithm design will lead to unbalanced load.

2, in some embodiments of the present application, a method for balancing load of a hash ring is provided, and the method includes:

Step S1, virtualize each real service node into a plurality of equal number of virtual service nodes, and obtain a virtual service node corresponding to each real service node;

Step S2, selecting a virtual service node from the virtual service nodes corresponding to the multiple real service nodes to form a virtual service node group to obtain multiple virtual service node groups, each of the virtual service node groups The order of the real service nodes corresponding to the virtual service nodes is the same;

Step S3, connecting the virtual service nodes in each virtual service node group end to end to obtain a virtual service node queue;

Step S4: Distribute the virtual service node queue evenly on the hash ring.

As described in step S1 above, each real service node is virtualized into multiple virtual service nodes, and the number of virtual service nodes virtualized by each real service node is equal, so that the number of virtual service nodes corresponding to each real service node is obtained. A virtual service node.

For example, there are a total of three real service nodes, Node1, Node2, and Node3. Each real service node virtualizes three virtual service nodes. For example, the real service node Node1 virtualizes three virtual service nodes, V1Node1, V2Node1, and V3Node1. The real service node Node2 virtualizes three virtual service nodes, V1, Node2, V2, Node2, and V3, Node2. The real service node Node3 virtualizes three virtual service nodes, V1 Node3, V2 Node3, and V3 Node3.

Therefore, each virtual service node is distributed on the hash ring instead of the real service node to increase the number of service nodes on the hash ring. Among them, each real service node can be understood as a server.

As described in step S2 above, a virtual service node is selected from the virtual service nodes corresponding to each real service node, for example:

Select the virtual service node V1 Node1 from the three virtual service nodes virtualized by the real service node Node1, select the virtual service node V1 Node2 from the three virtual service nodes virtualized by the real service node Node2, and select the virtual service node V1 Node2 from the real service node Node3. The virtual service node V1 Node3 is selected from the three virtual service nodes to form a virtual service node group V1 Node1, V1 Node2, and V1 Node3.

Select the virtual service node V2Node1 from the three virtual service nodes virtualized by the real service node Node1, select the virtual service node V2Node2 from the three virtual service nodes virtualized by the real service node Node2, and virtualize it from the real service node Node3 The virtual service node V2Node3 is selected from the three virtual service nodes in, so as to form a virtual service node group V2Node1, V2Node2, and V2Node3.

Select the virtual service node V3Node1 from the three virtual service nodes virtualized by the real service node Node1, select the virtual service node V3Node2 from the three virtual service nodes virtualized by the real service node Node2, and virtualize it from the real service node Node3 The virtual service node V3Node3 is selected from the three virtual service nodes in, so as to form a virtual service node group V3Node1, V3Node2, and V3Node3.

Thus, three virtual service node groups are obtained. The order of the virtual service nodes corresponding to the real service nodes in the three virtual service node groups is the same. For example, in the above three virtual service node groups, each virtual service node group is a virtual service node corresponding to the real service node Node1 At the top, the virtual service node corresponding to the real service node Node2 is ranked in the middle, and the virtual service node corresponding to the real service node Node3 is ranked last.

As described in step S3 above, the virtual service nodes in each virtual service node group are connected end to end to obtain a virtual service node queue, for example:

After connecting the above three virtual service node groups end to end, the virtual service node queues V1 Node1, V1 Node2, V1 Node3, V2 Node1, V2 Node2, V2 Node3, V3 Node1, V3 Node2, V3 Node3 are obtained.

As described in the above step S4, the virtual service node queues obtained through the above steps are evenly distributed on the hash ring. As shown in Figure 3, after the virtual service node queues V1, Node1, V1, Node2, V1, Node3, V2, Node1, V2, Node2, V2, Node3, V3, Node1, V3, Node2, V3, Node3 are obtained, the virtual service nodes in the queue are aligned according to the The sequence in the column is evenly distributed on the hash ring, so that request a, request b, and request c can be more evenly loaded on each real service node.

Through the above steps, it can be ensured that the two adjacent virtual service nodes on the hash ring correspond to different real service nodes, so that all virtual service nodes are evenly distributed on the hash ring, and each real service node is The virtual service nodes that are virtualized are also evenly distributed on the hash ring.

Through the above scheme, not only can the load of each service node on the hash ring be more balanced, but also the characteristics of the hash ring can be retained to the greatest extent.

For example, add a new node Node4 and load it between the virtual nodes of Node3 and Node1, so that the node will be expanded. At the same time, most of the load clusters will not change, which will only affect part of the data load between Node3 and Node1 in the newly added Node4. On the virtual node.

Influence ratio formula: (V_Node'-V_Node)/(2×V_Node);

Among them, V_Node is the original virtual node; V_Node' is the virtual node after capacity expansion.

As shown in Figure 4, according to the above scheme, there are nine original virtual nodes. If a real service node Node4 is added to expand the capacity, three virtual service nodes V1 Node4, V2 Node4, and V3 Node4 will be added. The total number of virtual service nodes is Twelve, the impact ratio is 16.667%.

If you use traditional hash modulus to load, when a new node is added, for example, there are four nodes, hash%4 is modulo. If another service node is added, hash%5 is modulated, so that only the first 25 The% hash modulus has not changed, and the rest will change. That is, the influence ratio is 75%.

Therefore, compared with the traditional hash modulus load, the expansion of the cluster will have a much smaller proportion of the load node change, and the hash ring characteristics are retained to the greatest extent.

In some embodiments of the present application, the above will select a virtual service node from the virtual service nodes corresponding to multiple real service nodes to form a virtual service node group, so as to obtain multiple virtual service node groups, and each virtual service node Step S2 in which the order of the virtual service nodes corresponding to the real service nodes in the group is the same, includes:

Step S21, respectively numbering each of the real service nodes and the virtual service nodes corresponding to each of the real service nodes, wherein the numbering sequence of the virtual service nodes corresponding to each of the real service nodes is the same;

In step S22, the virtual service nodes with the same number are formed into one virtual service node group, and the virtual service node groups are arranged according to the corresponding number of the real service node in each virtual service node group.

As described in step S21 above, each real service node may be numbered first, for example, the three real service nodes are respectively numbered Node1, Node2, and Node3, so as to facilitate the distinction between different real service nodes.

Then, the virtual service nodes virtualized by each real service node are numbered, and the numbering sequence of the virtual service nodes corresponding to each real service node is the same.

For example, the three virtual service nodes virtualized by one real service node Node are numbered V1, V2, and V3, respectively, and the three virtual service nodes virtualized by each of the remaining real service nodes are numbered V1, V2, and V3, respectively .

As described in step S22 above, the virtual service nodes with the same number are formed into a virtual service node group.

For example, the virtual service node numbered V1 virtualized by the real service node Node1 is recorded as V1Node1, and the virtual service node numbered V1 virtualized by the real service node Node2 is recorded as V1 Node2 and the virtual service node virtualized by the real service node Node3 The virtual service node numbered V1 is recorded as V1 and Node3 forms a virtual service node group, which is recorded as the first virtual service node group.

The virtual service node numbered V2 virtualized by the real service node Node1 is recorded as V2Node1, the virtual service node numbered V2 virtualized by the real service node Node2 is recorded as V2 Node2, and the number virtualized by the real service node Node3 is The virtual service section of V2 is marked as V2 Node3 points to form a virtual service node group, which is recorded as the second virtual service node group.

The virtual service node numbered V3 virtualized by the real service node Node1 is recorded as V3Node1, the virtual service node numbered V3 virtualized by the real service node Node2 is recorded as V3 Node2 and the virtual service node numbered by the real service node Node3 is The virtual service node of V3 is recorded as V3 Node3 forms a virtual service node group, which is recorded as the third virtual service node group.

In this way, three virtual service node groups are obtained, and then the three virtual service node groups are arranged according to the numbers of the corresponding real service nodes.

For example, in the first virtual service node group, the three virtual service nodes V1, Node1, V1, Node2, and V1, Node3 are arranged according to the numbers of the corresponding three real service nodes Node1, Node2, and Node3, namely V1, Node1, V1, Node2, and V1 Node3.

Similarly, in the second virtual service node group, the three virtual service nodes V2, Node1, V2, Node2, and V2, Node3 are arranged according to the numbers of the corresponding three real service nodes, Node1, Node2, and Node3, namely V2, Node1, V2, Node2, and V2 Node3, in the third virtual service node group, the three virtual service nodes V3 Node1, V3 Node2, and V3 Node3 are arranged according to the numbers of the corresponding three real service nodes Node1, Node2, and Node3, namely V3, Node1, V3, Node2, and V3Node3.

Through the above method, the virtual service nodes in each virtual service node group are sorted, so that the corresponding relationship between the virtual service node and the real service node is more intuitive, which is convenient for analysis and comparison and load verification.

In some embodiments of the present application, the above step S4 of evenly distributing the virtual service node queue on the hash ring includes:

Step S41, taking the arc length between two adjacent virtual service nodes on the hash ring as a tolerance, and calculating the value of the equal difference common term based on the common term formula of the arithmetic sequence;

Step S42, based on the correspondence between the virtual service node in the virtual service node queue and the equal difference pass item, map the virtual service node queue on the hash ring according to the value of the equal difference pass item.

As described in step S41 above, the arc length between two adjacent virtual service nodes on the hash ring is used as the tolerance, and the value of the equal difference common term is calculated based on the common term formula of the arithmetic sequence, for example:

The general formula of the arithmetic sequence is

a _n ＝a ₁ +(n-1)×d

Among them, a _n represents the general term. A _n general term representative of the position of the virtual queue service node n th hash virtual service node on the ring. a ₁ represents the first item. n represents the number of items. Let the number of items n be equal to the number of all virtual nodes in the virtual service node queue. d stands for tolerance.

Assuming that the arc length between two adjacent virtual service nodes on the hash ring is l, the tolerance d is replaced by the arc length l to obtain

a _n ＝a ₁ +(n-1)×l

According to this formula, combining the arc length of l and the number of terms n, the value of each term in the equidistant general term can be calculated.

As described in the above step S42, based on the correspondence between the virtual service nodes in the virtual service node queue and the equal difference pass item, the virtual service node queue is mapped on the hash ring according to the value of the equal difference pass item. Wherein the correspondence relationship between the virtual service node through the arithmetic term refers to the position on behalf of a _n term virtual service node queue n-th hash virtual service node on the ring to construct general term virtual service and a _n Correspondence of nodes.

For example, a ₁ represents the position of the first virtual service node in the virtual service node queue on the hash ring, a ₁ corresponds to the first virtual service node in the virtual service node queue; a ₂ represents the second virtual service node in the virtual service node queue hash location service node on the ring, the second virtual queue service node corresponding to a ₂ virtual service node, and so on, through a _n item corresponding to the n-th virtual service node.

Whereby, after the above step S41 is calculated through the numerical entries contained a _n, according to the above correspondence relationship, the virtual service node queue mapping each virtual service node on the ring a _n hash values corresponding to the position .

For example, if a ₁ =1073741823, a ₂ =2147483647, and a ₃ =3221225471 are calculated, then V1 Node1, V1 Node2, and V1 Node3 in the virtual service node queue are mapped to positions 1073741823, 2147483647, and 3221225471 on the hash ring, respectively. By analogy, all virtual service nodes in the queue of all virtual service nodes are mapped on the hash ring, so that all virtual service nodes are evenly distributed on the hash ring.

In some embodiments of the present application, the arc length between two adjacent virtual service nodes on the hash ring is used as the tolerance, and the value of the equal difference common term is calculated based on the common term formula of the arithmetic sequence before step S21 , The method further includes:

Step S21: Obtain the value range of the hash value of the virtual service node, and obtain the perimeter of the hash ring according to the value range;

Step S22: Calculate the arc length according to the perimeter of the hash ring and the number of virtual service nodes on the hash ring.

As described in the above step S21 and the above step S22, the perimeter of the hash ring can be obtained according to the value range of the hash value of the virtual service node. The specific method for obtaining the perimeter of the hash ring may be to use the largest value in the range of the hash value of the virtual service node as the perimeter of the hash ring, so as to facilitate the calculation of the radius of the hash ring according to the perimeter of the hash ring in the next step.

For example, through step S21, the perimeter c of the hash ring is 2 ³² -1, and the virtual service node on the hash ring divides the hash ring into n equal parts, so that the value of any two adjacent virtual service nodes on the hash ring can be calculated. The arc length between l=(2 ³² -1)/n.

In the above embodiment, the tolerance d=1, and then d=(2 ³² -1)/n is obtained.

In some embodiments, it can also be obtained according to the arc length formula l=a×r and the conversion formula of radians and angles 1°=π/180°

l=m×π×r/180

Among them, α represents the radian value, r represents the radius of the hash, and m represents the angle of the center of the circle;

Let the arc length be equal to the tolerance value, that is, the length l=d, so d=m×π×r/180;

Since the hash ring is equally divided by the virtual service node n, the central angle corresponding to each arc is m=360/n;

And then get,

From the perimeter of the hash ring c=2 ³² -1, combined with the perimeter formula of the circle c=2×π×r, we get

2 ³² -1=2×π×r

Then we get r=(2 ³² -1)/2π, and substituting into the above formula can also get

d=(2 ³² -1)/n.

In some embodiments of the present application, the above step S21 of obtaining the value range of the hash value of the virtual service node, and obtaining the perimeter of the hash ring according to the value range includes:

Step S211: Use the maximum value corresponding to the value range as the circumference of the hash ring.

For example, if the value of the hash of the virtual service node ranges from 0 to 2 ³² -1, the maximum value of the hash of the virtual service node is 2 ³² -1, and the perimeter of the hash ring can be recorded as c=232- 1. Substituting the perimeter c of the hash ring into the above embodiment can be used to calculate the tolerance of the arithmetic sequence.

In some embodiments of the present application, the arc length between two adjacent virtual service nodes on the hash ring is used as the tolerance, and after the step S41 of calculating the value of the equal difference common term based on the common term formula of the arithmetic sequence, so The method also includes:

In step S411, the value of the equal difference pass term is taken as an integer.

In step S411, each value calculated by the equal difference general term is taken as an integer, for example:

a ₁ =1×(2 ³² -1)/4=1073741823, a ₂ ＝2×(2 ³² -1)/4=2147483647

a ₃ =3×(2 ³² -1)/4=3221225471, a ₄ =4×(2 ³² -1)/4=4294967295, which is convenient for statistics and calculations.

In some embodiments of the present application, the number of virtual service nodes in the virtual service node queue is an even number, so that the load is more balanced. By setting up different real service nodes and virtual service nodes, the load of each real service node is verified as follows:

Setting plan 1: There are four real service nodes, and each real service node virtualizes two virtual service nodes. Through computer verification, the load results are as follows:

The real service node Node1, load ratio: 0.2527;

The real service node Node2, load ratio: 0.2486;

The real service node Node3, load ratio: 0.2441;

The real service node Node4, load ratio: 0.2546.

Setting plan 2: There are four real service nodes, and each real service node virtualizes five virtual service nodes. Through computer verification, the load results are as follows:

The real service node Node1, load ratio: 0.2907;

The real service node Node2, load ratio: 0.3029;

The real service node Node3, load ratio: 0.2001;

The real service node Node4, load ratio: 0.2063.

Comparing the load results of setting plan 1 and setting plan 2, it can be seen that when the virtual node is an even number, the load of each real service node is more balanced.

Setting scheme 3: There are four real service nodes, and each real service node virtualizes six virtual service nodes. Through computer verification, the load results are as follows:

The real service node Node1, load ratio: 0.2506;

The real service node Node2, load ratio: 0.2505;

The real service node Node3, load ratio: 0.2508;

The real service node Node4, load ratio: 0.2481.

Comparing the load results of setting plan 2 and setting plan 3, it can be seen that when there are more virtual nodes, the load of each real service node is more balanced.

Setting plan 4: Expand the capacity of real service nodes to eight, and each real service node virtualizes six virtual service nodes. Through computer verification, the load results are as follows:

The real service node Node1, load ratio: 0.1258;

The real service node Node2, load ratio: 0.1229;

The real service node Node3, load ratio: 0.1244;

The real service node Node4, load ratio: 0.1223;

The real service node Node5, load ratio: 0.1232;

The real service node Node6, load ratio: 0.1235;

The real service node Node7, load ratio: 0.1279;

The real service node Node8, load ratio: 0.13.

Comparing the load results of setting plan 3 and setting plan 4, it can be seen that the load is still relatively balanced after the node is expanded according to the plan of this application.

Through the above embodiments, after the virtual service nodes are evenly distributed on the hash ring, the hash value of the requested information can be calculated according to the hash algorithm, and then the hash value of the requested information and the hash value of each virtual service node will be requested. Assigned to the closest virtual service node in the clockwise direction.

For example, the hash function uses prime numbers (unless 1) to avoid overflow and cause loss of information. Taking into account the performance of the virtual machine, the virtual machine uses 2<<5-1 by default to improve the performance of computing hash. In summary, the hash algorithm is:

hash=31*hash+ascii;

Among them, the hash initialization value is 0. The initial value of ascii is 32. ascii represents the ascii code of the string that needs to be hashed. For example, it needs to be hashed according to the ip. The ascii here is the ascii code of the ip character array. For example, if ip is 114.114.114.114, the calculated value is -1009865594, and then the absolute value needs to be 1009865594.

In this step, the request information is hashed through the hash algorithm to improve the discrete probability.

Then, according to the calculated hash value of the request information, the request is mapped to the hash ring. For example, the hash value of the request information obtained after hash calculation is 1009865594. Then the request information is mapped to the 1009865594 location on the hash.

It can be seen that the location 1009865594 of the requested information is relatively close to the location of the virtual service node V1Node1 (a ₁ =1073741823). According to the clockwise direction, the request information is loaded on the real service node Node1 corresponding to the virtual service node V1Node1.

Based on the above embodiments, the present application distributes the virtual service nodes obtained after virtualization evenly on the hash ring, and the virtual nodes virtualized by each real service node are also evenly distributed on the hash ring, so that each The probability of a data request load on each service node is as uniform as possible. Through this solution, the characteristics of the hash ring can be retained to the greatest extent. After a certain server/service node fails, or a new service node is added for expansion, the impact ratio is greatly reduced, making the server cluster more stable and reliable.

Referring to FIG. 5, in some embodiments of the present application, a hash ring load balancing device is also provided, which includes a virtual unit 10, a grouping unit 20, a connection unit 30, and a distribution unit 40. The virtual unit 10 is used to virtualize each real service node into a plurality of virtual service nodes of equal number, and then obtain the virtual service node corresponding to each real service node. The grouping unit 20 is configured to select one virtual service node from the virtual service nodes corresponding to the multiple real service nodes to form a virtual service node group, thereby obtaining multiple virtual service node groups. Wherein, the order of the real service nodes corresponding to the virtual service nodes in each virtual service node group is the same. The connecting unit 30 is used to connect the virtual service nodes in each virtual service node group end to end to obtain a virtual service node queue. The distribution unit 40 is used to evenly distribute the virtual service node queue on the hash ring.

Optionally, the hash ring load balancing device further includes a hash value calculation unit and a distribution unit for calculating request information. The hash value calculation unit of the request information is used to calculate the hash value of the request information. The allocation unit is configured to allocate the request information to the virtual service node closest in a clockwise direction according to the hash value of the request information.

Optionally, the distribution unit is configured to: use the arc length between two adjacent virtual service nodes on the hash ring as a tolerance, and calculate the value of the equal difference common term based on the general term formula of the arithmetic sequence Based on the correspondence between the virtual service node in the virtual service node queue and the equal difference pass item, the virtual service node queue is mapped to the hash ring according to the value of the equal difference pass item.

Optionally, the distribution unit is configured to: obtain the value range of the hash value of the virtual service node, and obtain the perimeter of the hash ring according to the value range; The arc length is calculated from the number of the virtual service nodes on the hash ring.

Optionally, the distribution unit is configured to: use the arc length between two adjacent virtual service nodes on the hash ring as a tolerance, and calculate the value of the equal difference common term based on the general term formula of the arithmetic sequence After that, the value of the equal difference general term is taken as an integer.

Optionally, the grouping unit is configured to: respectively number each real service node and the virtual service node corresponding to each real service node, wherein the virtual service corresponding to each real service node The numbering sequence of the nodes is the same; virtual service nodes with the same number are formed into a virtual service node group, and the virtual service node groups are arranged according to the corresponding real service node number.

For the implementation process of the functions and roles of each module in the above-mentioned device, refer to the implementation process of corresponding steps in the above-mentioned hash ring load balancing method for details, which will not be repeated here.

It should be noted that although in the above detailed description it is mentioned for several modules or units, this division is not mandatory. In fact, according to the embodiments disclosed in the present application, the features and functions of two or more modules or units described above can be embodied in one module or unit. Conversely, the features and functions of a module or unit described above can be further divided into multiple modules or units to be embodied.

In addition, although the various steps of the method in the present application are described in a specific order in the drawings, this does not require or imply that these steps must be performed in the specific order, or that all the steps shown must be performed to achieve the desired result. Additionally or alternatively, some steps may be omitted, multiple steps may be combined into one step for execution, and/or one step may be decomposed into multiple steps for execution, etc.

Through the description of the above embodiments, those skilled in the art can easily understand that the example embodiments described here can be implemented by software, or can be implemented by combining software with necessary hardware. Therefore, the technical solution according to the embodiments of the present application can be embodied in the form of a software product, which can be stored in a non-volatile storage medium (which can be a CD-ROM, U disk, mobile hard disk, etc.) or on the network , Including several instructions to make a computing device (which can be a personal computer, a server, a mobile terminal, or a network device, etc.) execute the method according to the embodiment of the present application.

Referring to FIG. 6, some embodiments of the present application also provide an electronic device for implementing the above method for balancing load of the hash ring. The electronic device 500 shown in FIG. 6 is only an example, and should not bring any limitation to the functions and scope of use of the embodiments of the present application.

The electronic device 500 is represented in the form of a general-purpose computing device. The components of the electronic device 500 may include but are not limited to: at least one processing unit 510, at least one storage unit 520, and a bus 530 connecting different system components (including the storage unit 520 and the processing unit 510).

Wherein, the storage unit stores program code, and the program code can be executed by the processing unit 510, so that the processing unit 510 executes the various exemplary methods described in the “Exemplary Method” section of this specification. Steps of implementation. For example, the processing unit 510 may execute steps S1 to S4 as shown in FIG. 2.

The storage unit 520 may include a readable medium in the form of a volatile storage unit, such as a random access storage unit (RAM) 5201 and/or a cache storage unit 5202, and may further include a read-only storage unit (ROM) 5203.

The storage unit 520 may also include a program/utility tool 5204 having a set (at least one) program module 5205. Such program module 5205 includes but is not limited to: an operating system, one or more application programs, other program modules, and program data, Each of these examples or some combination may include the implementation of a network environment.

The bus 530 may represent one or more of several types of bus structures, including a storage unit bus or a storage unit controller, a peripheral bus, a graphics acceleration port, a processing unit, or a local area using any bus structure among multiple bus structures. bus.

The electronic device 500 may also communicate with one or more external devices 1000 (such as keyboards, pointing devices, Bluetooth devices, etc.), and may also communicate with one or more devices that enable a user to interact with the electronic device 800, and/or communicate with Any device (such as a router, modem, etc.) that enables the electronic device 500 to communicate with one or more other computing devices. Such communication may be performed through an input/output (I/O) interface 540. In addition, the electronic device 500 may also communicate with one or more networks (for example, a local area network (LAN), a wide area network (WAN), and/or a public network, such as the Internet) through the network adapter 560. As shown in the figure, the network adapter 560 communicates with other modules of the electronic device 500 through the bus 530. It should be understood that although not shown in the figure, other hardware and/or software modules can be used in conjunction with the electronic device 500, including but not limited to: microcode, device drivers, redundant processing units, external disk drive arrays, RAID systems, tape drives And data backup storage system, etc.

Through the description of the above embodiments, those skilled in the art can easily understand that the example embodiments described here can be implemented by software, or can be implemented by combining software with necessary hardware. Therefore, the technical solution according to the embodiments of the present application can be embodied in the form of a software product, which can be stored in a non-volatile storage medium (which can be a CD-ROM, U disk, mobile hard disk, etc.) or on the network , Including several instructions to make a computing device (which can be a personal computer, a server, a terminal device, or a network device, etc.) execute the method according to the embodiment of the present application.

The embodiment of the present application also provides a readable storage medium, such as a computer readable storage medium, on which a computer program is stored, and when the computer program is executed by a processor, some or all of the steps of the above hash ring load balancing method are realized. Optionally, the media involved in this application, such as computer-readable storage media, may be non-volatile or volatile. For example, referring to FIG. 7, in an exemplary embodiment of the present application, a computer-readable storage medium is also provided, on which a program product capable of implementing the above-mentioned hash ring load balancing method of this specification is stored. In some possible implementation manners, various aspects of the present application can also be implemented in the form of a program product, which includes program code. When the program product runs on a terminal device, the program code is used to make the The electronic device executes the steps described in the above-mentioned "Exemplary Method" section of this specification according to various exemplary embodiments of the present application.

Fig. 7 is a computer-readable storage medium for realizing the above-mentioned hash ring load balancing method according to an exemplary embodiment of the present application. FIG. 7 depicts a program product 600 for implementing the above-mentioned hash ring load balancing method according to an embodiment of the present application, which can adopt a portable compact disk read-only memory (CD-ROM) and include program code, and can be installed in an electronic device, For example, running on a personal computer. However, the program product of this application is not limited to this. In this document, the readable storage medium can be any tangible medium that contains or stores a program, and the program can be used by or in combination with an instruction execution system, device, or device.

The program product can use any combination of one or more readable media. The readable medium may be a readable signal medium or a readable storage medium. The readable storage medium may be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device, or device, or a combination of any of the above. More specific examples (non-exhaustive list) of readable storage media include: electrical connections with one or more wires, portable disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable Type programmable read only memory (EPROM or flash memory), optical fiber, portable compact disk read only memory (CD-ROM), optical storage device, magnetic storage device, or any suitable combination of the above.

The computer-readable signal medium may include a data signal propagated in baseband or as a part of a carrier wave, and readable program code is carried therein. This propagated data signal can take many forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination of the foregoing. The readable signal medium may also be any readable medium other than a readable storage medium, and the readable medium may send, propagate, or transmit a program for use by or in combination with the instruction execution system, apparatus, or device.

The program code contained on the readable medium can be transmitted by any suitable medium, including but not limited to wireless, wired, optical cable, RF, etc., or any suitable combination of the foregoing.

The program code used to perform the operations of the present application can be written in any combination of one or more programming languages. The programming languages include object-oriented programming languages—such as Java, C++, etc., as well as conventional procedural programming languages. Programming language-such as "C" language or similar programming language. The program code can be executed entirely on the user's computing device, partly on the user's device, executed as an independent software package, partly on the user's computing device and partly executed on the remote computing device, or entirely on the remote computing device or server Executed on. In the case of a remote computing device, the remote computing device can be connected to a user computing device through any kind of network, including a local area network (LAN) or a wide area network (WAN), or it can be connected to an external computing device (for example, using Internet service providers). Business to connect via the Internet).

In addition, the above-mentioned drawings are merely schematic illustrations of the processing included in the method according to the exemplary embodiments of the present application, and are not intended for limitation. It is easy to understand that the processing shown in the above drawings does not indicate or limit the time sequence of these processings. In addition, it is easy to understand that these processes can be executed synchronously or asynchronously in multiple modules.

Although this application has been described with reference to several exemplary embodiments, it should be understood that the terms used are illustrative and exemplary rather than restrictive. Since this application can be implemented in various forms without departing from the spirit or essence of the application, it should be understood that the above-mentioned implementations are not limited to any of the foregoing details, but should be broadly within the spirit and scope defined by the appended claims. Interpretation, therefore, all changes and modifications falling within the scope of the claims or their equivalents shall be covered by the appended claims.

Claims

A method for load balancing with a hash ring, which includes:

Virtualize each real service node into a number of equal number of virtual service nodes, and obtain the virtual service node corresponding to each real service node;

A virtual service node is selected from multiple virtual service nodes corresponding to multiple real service nodes to form a virtual service node group to obtain multiple virtual service node groups, and the virtual service node group in each virtual service node group The order of the real service nodes corresponding to the service nodes is the same;

Connect the virtual service nodes in each virtual service node group end to end to obtain a virtual service node queue;

The virtual service node queue is evenly distributed on the hash ring.
The hash ring load balancing method according to claim 1, wherein evenly distributing the virtual service node queue on the hash ring comprises:

Taking the arc length between two adjacent virtual service nodes on the hash ring as a tolerance, and calculating the value of the equal difference common term based on the common term formula of the arithmetic sequence;

Based on the corresponding relationship between the virtual service node in the virtual service node queue and the equal difference pass item, the virtual service node queue is mapped on the hash ring according to the value of the equal difference pass item.
The hash ring load balancing method according to claim 2, wherein the arc length between two adjacent virtual service nodes on the hash ring is used as a tolerance, and the equal difference is calculated based on the general term formula of the arithmetic sequence Before the value of the item, the method also includes:

Obtaining a value range of the hash value of the virtual service node, and obtaining the perimeter of the hash ring according to the value range;

The arc length is calculated according to the perimeter of the hash ring and the number of virtual service nodes on the hash ring.
The hash ring load balancing method according to claim 2, wherein the arc length between two adjacent virtual service nodes on the hash ring is used as a tolerance, and the equal difference is calculated based on the general term formula of the arithmetic sequence After the value of the item, the method further includes:

The numerical value of the equal difference general term is taken as an integer.
The hash ring load balancing method according to claim 1, wherein a virtual service node is selected from virtual service nodes corresponding to multiple real service nodes to form a virtual service node group to obtain multiple virtual service node groups , The order of the real service nodes corresponding to the virtual service nodes in each virtual service node group is the same, including:

Respectively number each of the real service nodes and the virtual service nodes corresponding to each of the real service nodes, wherein the numbering sequence of the virtual service nodes corresponding to each of the real service nodes is the same;

The virtual service nodes with the same number are formed into one virtual service node group, and the virtual service node groups are arranged according to the number of the corresponding real service node.
The hash ring load balancing method according to claim 1, wherein after the virtual service node queue is evenly distributed on the hash ring, the method further comprises:

Calculate the hash value of the requested information;

According to the hash value of the request information, the request information is allocated to the virtual service node closest in a clockwise direction.
The hash ring load balancing method according to claim 1, wherein:

The number of the virtual service nodes in the virtual service node queue is an even number.
A hash ring load balancing device, which includes:

The virtual unit is used to virtualize each real service node into a number of equal number of virtual service nodes to obtain the virtual service node corresponding to each real service node;

The grouping unit is configured to select one of the virtual service nodes corresponding to the multiple real service nodes to form a virtual service node group to obtain multiple virtual service node groups, each The order of the real service nodes corresponding to the virtual service nodes in the virtual service node groups is the same;

The connecting unit is used to connect the virtual service nodes in each virtual service node group end to end to obtain a virtual service node queue;

The distribution unit is configured to evenly distribute the virtual service node queue on the hash ring.
An electronic device includes a memory and a processor, and a computer program is stored in the memory, wherein the processor implements the following steps when executing the computer program:

Virtualize each real service node into a number of equal number of virtual service nodes, and obtain the virtual service node corresponding to each real service node;

A virtual service node is selected from multiple virtual service nodes corresponding to multiple real service nodes to form a virtual service node group to obtain multiple virtual service node groups, and the virtual service node group in each virtual service node group The order of the real service nodes corresponding to the service nodes is the same;

Connect the virtual service nodes in each virtual service node group end to end to obtain a virtual service node queue;

The virtual service node queue is evenly distributed on the hash ring.
The electronic device according to claim 9, wherein when the virtual service node queue is evenly distributed on the hash ring, the following steps are specifically implemented:

Taking the arc length between two adjacent virtual service nodes on the hash ring as a tolerance, and calculating the value of the equal difference common term based on the common term formula of the arithmetic sequence;

Based on the corresponding relationship between the virtual service node in the virtual service node queue and the equal difference pass item, the virtual service node queue is mapped on the hash ring according to the value of the equal difference pass item.
The electronic device according to claim 10, wherein the arc length between two adjacent virtual service nodes on the hash ring is used as a tolerance, and the value of the equal difference common term is calculated based on the general term formula of the arithmetic sequence Previously, the processor was also used to execute the computer program to implement the following steps:

Obtaining a value range of the hash value of the virtual service node, and obtaining the perimeter of the hash ring according to the value range;

The arc length is calculated according to the perimeter of the hash ring and the number of virtual service nodes on the hash ring.
The electronic device according to claim 9, wherein one virtual service node is selected from the virtual service nodes corresponding to the multiple real service nodes to form a virtual service node group to obtain multiple virtual service node groups, each When the order of the real service nodes corresponding to the virtual service nodes in the virtual service node group is the same, the following steps are specifically implemented:

Respectively number each of the real service nodes and the virtual service nodes corresponding to each of the real service nodes, wherein the numbering sequence of the virtual service nodes corresponding to each of the real service nodes is the same;

The virtual service nodes with the same number are formed into one virtual service node group, and the virtual service node groups are arranged according to the number of the corresponding real service node.
The electronic device according to claim 9, wherein after the virtual service node queue is evenly distributed on the hash ring, the processor is further configured to execute the computer program to implement the following steps:

Calculate the hash value of the requested information;

According to the hash value of the request information, the request information is allocated to the virtual service node closest in a clockwise direction.
The electronic device according to claim 9, wherein:

The number of the virtual service nodes in the virtual service node queue is an even number.
A readable storage medium having a computer program stored thereon, wherein the computer program implements the following steps when being executed by a processor:

Virtualize each real service node into multiple equal number of virtual service nodes, and obtain the virtual service node corresponding to each real service node;

A virtual service node is selected from multiple virtual service nodes corresponding to multiple real service nodes to form a virtual service node group to obtain multiple virtual service node groups, and the virtual service node group in each virtual service node group The order of the real service nodes corresponding to the service nodes is the same;

Connect the virtual service nodes in each virtual service node group end to end to obtain a virtual service node queue;

The virtual service node queue is evenly distributed on the hash ring.
The readable storage medium according to claim 15, wherein when the virtual service node queue is evenly distributed on a hash ring, the specific implementation is as follows:

Taking the arc length between two adjacent virtual service nodes on the hash ring as a tolerance, and calculating the value of the equal difference common term based on the common term formula of the arithmetic sequence;

Based on the corresponding relationship between the virtual service node in the virtual service node queue and the equal difference pass item, the virtual service node queue is mapped on the hash ring according to the value of the equal difference pass item.
The readable storage medium according to claim 16, wherein the arc length between two adjacent virtual service nodes on the hash ring is used as a tolerance, and the equal difference common term is calculated based on the general term formula of the arithmetic sequence Before the value of, the computer program also implements the following steps when being executed by the processor:

Obtaining a value range of the hash value of the virtual service node, and obtaining the perimeter of the hash ring according to the value range;

The arc length is calculated according to the perimeter of the hash ring and the number of virtual service nodes on the hash ring.
15. The readable storage medium according to claim 15, wherein a virtual service node is selected from the virtual service nodes corresponding to the multiple real service nodes to form a virtual service node group to obtain multiple virtual service node groups, When the sequence of the real service nodes corresponding to the virtual service nodes in each virtual service node group is the same, the specific implementation is as follows:

Respectively number each of the real service nodes and the virtual service nodes corresponding to each of the real service nodes, wherein the numbering sequence of the virtual service nodes corresponding to each of the real service nodes is the same;

The virtual service nodes with the same number are formed into one virtual service node group, and the virtual service node groups are arranged according to the number of the corresponding real service node.
The readable storage medium according to claim 15, wherein after the virtual service node queue is evenly distributed on the hash ring, the following steps are further implemented when the computer program is executed by the processor:

Calculate the hash value of the requested information;

According to the hash value of the request information, the request information is allocated to the virtual service node closest in a clockwise direction.
The readable storage medium according to claim 15, wherein:

The number of the virtual service nodes in the virtual service node queue is an even number.