KR101944594B1 - Server for distributed network and method of consensus thereof - Google Patents

Abstract

The present invention relates to a server participating in a distributed network-based consensus and a method of participating in a consensus of the server in a first group, and a server participating in a distributed network-based consensus according to an embodiment of the present invention, A memory for storing log and consensus modules, and a processor for performing an action in accordance with the log, wherein the processor executes the consensus module such that the status of the server is changed in a follower mode, a candidate mode, A leader mode, and a connection state of a network to a plurality of second servers existing in a first group to which the server belongs and operating in a follower mode or a candidate mode as the server operates in a reader mode Transmitting a heartbeat periodically, and transmitting a response to the heartbeat to the respective devices 2 from the server, wherein the processor deletes, in the first group, a third server that has not received a response to the heartbeat among the plurality of second servers, The server information and the changed log information existing in the first group are transmitted and updated.

Description

{SERVER FOR DISTRIBUTED NETWORK AND METHOD OF CONSENSUS THEREOF} [0001] The present invention relates to a server and consensus participating in a consensus in a distributed network system,

The present invention relates to a server participating in consensus in a distributed network and a method for participating in consensus.

With the development of mobile devices and communication networks, the utilization of virtual space has been increasing. In recent years, distributed network systems have been increasing in which a large number of users collect, store, process and process data together in a virtual space.

The distributed network system in the virtual space can increase the efficiency of work and overcome time and space constraints because many users can work together without gathering in a space in reality. It has attracted much attention.

One of the major problems in developing a robust distributed network system is the question of how to achieve a distributed consensus that ensures that participants in a distributed group agree on the data they share. Therefore, many algorithms have been introduced to achieve the distributed consensus up to now, and there are problems in that the algorithm is used in various systems, but the contents of the algorithm are complicated and it is difficult to verify the accuracy of implementation.

In particular, the algorithm proposed by Diego Ongaro and John Ousterhout is a consensus algorithm based on a majority consensus. In order to solve the problem of distributed consensus, the same log maintenance, state transition, and message exchange of each node constituting the consensus group are used. In addition, the Raft algorithm performs the log update operation with a majority of consent under the leadership of the leader, so that it is possible to safely manage the log even if a problem occurs in the middle.

However, in the case of the Raft algorithm, there is a problem in that the dynamic extensibility of the consensus group can not be provided because the addition or removal of participating nodes is not considered. Therefore, in such a case, there may arise a problem that the group itself needs to be reconfigured when adding or deleting nodes to the consensus group.

Disclosure of Invention Technical Problem [8] Accordingly, the present invention has been made in view of the above problems occurring in the prior art, and it is an object of the present invention to provide a method and system for participating in consensus in a distributed network.

In particular, the present invention provides a server of a distributed network system and a method of participating in a consensus of the server, wherein the number of participating nodes in the consensus group can be dynamically changed by supporting participation nodes in a consensus group in collaboration.

It should be understood, however, that the technical scope of the present invention is not limited to the above-described technical problems, and other technical problems may exist.

According to an aspect of the present invention, there is provided a server for participating in a distributed network-based consensus system, comprising: a memory for storing action information and a consensus module; Wherein the processor is configured to define a follower mode, a candidate mode, and a leader mode as the server executes the consensus module, and the server operates in a reader mode A heartbeat for checking the connection status of the network to a plurality of second servers existing in a first group to which the server belongs and operating in a follower mode or a candidate mode is periodically transmitted, Receiving a response from each of the second servers, wherein the processor, And deletes the third server which has not received the response to the first bit in the first group and transmits the server information and the changed log information existing in the first group to the remaining second servers except for the third server Update it.

The server further includes a processor for participating in a distributed network-based consensus in a first group according to another aspect of the present invention, a memory storing a log and a consensus module in which action information is recorded, and a processor for performing an action according to the log (A) defining a state of the server as a leader mode as the processor executes the consensus module; (B) transmitting a heartbeat for confirming a connection state of the network to a plurality of second servers, the processor residing in the first group and operating in a follower mode or a candidate mode; (C) receiving, by the processor, a response to the heartbeat from each of the second servers; (D) deleting, in the first group, a third server among the plurality of second servers, the processor not receiving a response to the heartbeat; And (e) transmitting and updating the server information and the changed log information existing in the first group to each of the remaining second servers except for the third server.

According to one of the above-mentioned tasks, a server of a distributed network system and a method of participating in a consensus of a server according to an embodiment of the present invention can support addition and deletion of participating nodes in a consensus group in collaboration The number of participating nodes of the consensus group can be dynamically changed.

It should be understood, however, that the effects obtained by the present invention are not limited to the above-mentioned effects, and other effects not mentioned may be clearly understood by those skilled in the art to which the present invention belongs It will be possible.

1 is a block diagram schematically illustrating a configuration of a server participating in consensus in a distributed network according to an embodiment of the present invention.
2 is a block diagram schematically illustrating a consensus method according to the state of each server existing in a first group according to an embodiment of the present invention.
FIG. 3 illustrates a process of dynamically changing the number of servers belonging to a first group in a leader mode server participating in consensus in a distributed network according to an exemplary embodiment of the present invention.
FIG. 4 illustrates variables added to the consensus algorithm when the server status is defined as a reader mode according to an embodiment of the present invention.
5 is a diagram for explaining in detail a process of checking a network connection state of second servers existing in a first group to which a server operating in a reader mode according to an embodiment of the present invention belongs.
6 is a flowchart illustrating a process of a server in a reader mode according to an exemplary embodiment of the present invention deletes a third server in a first group that has no response to a heartbeat.
7 is a diagram for explaining an operation procedure of a reader mode in which a group addition request is received from a new server that is not present in the first group according to an embodiment of the present invention.
8 is a flowchart illustrating a process of adding a fourth server in a first group to a server in a reader mode according to an embodiment of the present invention.

Hereinafter, embodiments of the present invention will be described in detail with reference to the accompanying drawings, which will be readily apparent to those skilled in the art. The present invention may, however, be embodied in many different forms and should not be construed as limited to the embodiments set forth herein. In order to clearly explain the present invention in the drawings, parts not related to the description are omitted.

Throughout the specification, when a part is referred to as being "connected" to another part, it includes not only "directly connected" but also "electrically connected" with another part in between . Also, when an element is referred to as "comprising ", it means that it can include other elements as well, without departing from the other elements unless specifically stated otherwise.

In this specification, the term " part " includes a unit realized by hardware, a unit realized by software, and a unit realized by using both. Further, one unit may be implemented using two or more hardware, or two or more units may be implemented by one hardware.

In this specification, some of the operations or functions described as being performed by the terminal or the device may be performed in the server connected to the terminal or the device instead. Similarly, some of the operations or functions described as being performed by the server may also be performed on a terminal or device connected to the server.

Hereinafter, a server participating in a consensus and a method of participating in a consensus in a distributed network according to an embodiment of the present invention will be described in detail with reference to the drawings.

Here, the network means a connection structure in which information can be exchanged between each node such as a terminal and a server. An example of such a network is a 3rd generation partnership project (3GPP) network, a long term evolution (LTE) network, a WIMAX world interoperability for microwave access networks, internet, local area network (LAN), wireless local area network (WLAN), wide area network (WAN), personal area network (PAN), bluetooth network, Broadcast networks, analog broadcast networks, digital multimedia broadcasting (DMB) networks, and the like.

1 is a block diagram schematically illustrating a configuration of a server participating in consensus in a distributed network according to an embodiment of the present invention.

2 is a block diagram schematically illustrating a consensus method according to the state of each server existing in a first group according to an embodiment of the present invention.

Meanwhile, a server according to an embodiment of the present invention may be understood as being the same as a node participating in a network terminal or a network.

Referring to FIG. 1, a server participating in a consensus in a distributed network according to an embodiment of the present invention includes a network communication module 110, a memory 120, and a processor 130.

The server performs consensus with the plurality of second servers existing in the first group to which the server belongs through the network communication module 110. [

The memory 120 may store log and consensus algorithms in which action information is recorded. However, the present invention is not limited thereto, and a server participating in consensus in a distributed network according to an embodiment of the present invention may further include a network device and a memory, and may include a separate consensus module.

At this point, the memory 120 is one of the computer storage media and may be any volatile and nonvolatile, removable and nonvolatile memory implemented in any method or technology for storage of information such as computer readable instructions, data structures, program modules or other data And a removable medium. For example, the storage device may be a ROM (Read Only Memory), a RAM (Random Access Memory), a magnetic tape, a magnetic disk, a flash memory, an optical data storage device, a compact flash Such as a NAND flash memory, a hard disk drive (HDD), etc., such as a memory card, a memory stick, a solid-state drive (SSD) Magnetic computer storage devices, and the like.

Processor 130 performs consensus with each of the second servers in the first group as it executes the consensus algorithm. That is, the processor 130 may share the log through the consensus, and may store and update the shared log in the memory 120. At this time, the log may store information about an action to be performed by the server.

The processor 130 performs an action in accordance with the log. Accordingly, a plurality of servers existing in the first group participate in consensus, share resources, i.e., log, and perform actions included in the log. Accordingly, a plurality of servers participating in the consensus in the first group can perform the same action.

In particular, as the processor 130 executes the consensus algorithm, the server may define its state as a follower mode, a candidate mode, or a leader mode. At this time, the state of the server can be defined through voting with a plurality of servers participating in the consensus in the first group.

At this time, one of the plurality of servers existing in the first group may be selected as a candidate mode to be a leader. If an error occurs in the server operating in the leader mode, if the server operating in the candidate mode agrees to a majority, the server can be selected as a new leader and operate as a leader mode.

As the state of the server is defined as the leader mode, the server can either vote for the log duplication with the second servers residing in the first group or transmit the results to each of the second servers.

On the other hand, as the state of the server is defined as the follower mode or the candidate mode, the server can send an action request to the second server or receive an action to be forwarded by the second server.

2, a server participating in a consensus in a first group may poll its status through a plurality of second servers belonging to a first group in a follower mode, a candidate, Mode, or a leader mode, and each server operates in a corresponding mode according to a defined state. That is, the plurality of servers included in the first group operate as different modes and can participate in the consensus.

In other words, the plurality of servers included in the first group can select a server to be operated in the reader mode as the processor 130 performs the consensus algorithm, and the server 230 in the reader mode can select the server The server 210 may transmit the vote for the log duplication with the server 210 in the follower mode or the server 220 in the candidate mode and may transmit the vote result to the server 210 in the follower mode or the server 220 in the candidate mode, respectively.

Meanwhile, the plurality of second servers not selected as the leaders operate as the candidate mode or the follower mode, and the server 210 in the follower mode or the server 220 in the candidate mode sends an action request to the server 230 in the leader mode Or an action transmitted by the server 230 in the reader mode.

In addition, the server participating in the consensus in the distributed network according to the embodiment of the present invention can remove the second server belonging to the first group to which the corresponding server belongs as it operates in the reader mode. Or receive a group addition request from a third server not belonging to the first group, and add the third server to the first group to perform consensus.

FIG. 3 illustrates a process of dynamically changing the number of servers belonging to a first group in a leader mode server participating in consensus in a distributed network according to an exemplary embodiment of the present invention.

Referring to FIGS. 3A and 3B, a server participating in a consensus in a distributed network according to an exemplary embodiment of the present invention performs a consensus algorithm on a first group By selecting a leader through voting, you can define the server's status as a leader mode.

3, the server 230 operating in the leader mode according to an embodiment of the present invention receives a group addition request from the new server 240, 1 group so that they can participate in the consensus together with the existing follower mode servers 212 to 218 existing in the first group.

The server 230 operating in the reader mode as shown in FIG. 3B or FIG. 3B may be a follower server in which the server is determined to be disconnected from the network among the followers 212-218 existing in the first group participating in the consensus The server 218 can be deleted.

Also, according to one embodiment of the present invention, as the processor 130 defines the status of the server as a candidate mode or a follower mode, it receives a heartbeat from a server in the leader mode existing in the first group, To the server in the reader mode.

In addition, a group addition request may be transmitted to the server of the reader mode in the first group, or the log update information may be received from the server of the reader mode to update the log stored in the memory.

4 to 8, a server operating in a reader mode according to an embodiment of the present invention changes the number of servers participating in consensus dynamically in a distributed network in a first group to which the server operates Will be described in detail.

FIG. 4 illustrates variables added to the consensus algorithm when the server status is defined as a reader mode according to an embodiment of the present invention.

Referring to FIG. 4, in order to dynamically change the number of servers, i.e., nodes, participating in consensus in the first group by performing the above-described consensus algorithm, the processor 130 according to an exemplary embodiment of the present invention As shown in FIG. 4, the consensus algorithm includes a peer variable (peers []), an invalid variable (invalid []), a non-response count variable (noResponseCnt []), a major variable (isMajor [ Variable (totalPeerNum []) can be added. In this case, the peer variable manages the list of participant nodes participating in the consensus group, the invalid variable manages the list for checking the connection status of the participant nodes, and the non-response count variable manages the number of the participant nodes not responding to the heartbeat of the leader do. Then, the major variable is a variable for checking whether the leader is a major or not in the group to which the reader belongs, and the number of peers is a variable for managing the number of participating nodes.

5 is a diagram for explaining in detail a process of checking a network connection state of second servers existing in a first group to which a server operating in a reader mode according to an embodiment of the present invention belongs.

Referring to FIG. 5, according to an embodiment of the present invention, a server 230 in a reader mode periodically receives a heartbeat for confirming a connection state of a network to a second server existing in a first group to which the server belongs (Sh1). Thereafter, the server 230 in the reader mode counts time (Sh2) and receives a response to the heartbeat from each second server within a predetermined time (Sh3). At this time, the second server may be the server 210 of the follower mode or the server 220 of the candidate mode, as described above.

On the other hand, as shown in (b) of FIG. 5, if a response to the heartbeat is not received from any of the second servers within the set time, the server 230 in the leader mode generates a timeout event, It determines that the third server that has not received the network connection is disconnected (Sh4).

6 is a flowchart illustrating a process of a server in a reader mode according to an exemplary embodiment of the present invention deletes a third server in a first group that has no response to a heartbeat.

6, the server 230 in the reader mode periodically transmits a heartbeat to a plurality of second servers (Sc1), and then counts time to check if a timeout event has occurred (Sc2 ). Thereafter, if a timeout event does not occur, it is determined that the network connection state is normal and the operation is terminated (Sc3). On the other hand, when a timeout event occurs, the number of non-response to the heartbeat response is increased (Sc4), it is determined whether the number of non-response times exceeds the predetermined number (C5) (Sc5) You can go back to transferring the heartbeat again and repeat that step. However, if the number of non-responses exceeds the predetermined number, it is possible to check whether the connection to the second server is in a non-connection state (Sc6), and delete the server.

Hereinafter, for convenience of description, a server whose number of non-responders exceeds a preset number, that is, a server disconnected from the network is defined as a third server for convenience of description.

At this time, the server 230 operating in the reader mode according to an embodiment of the present invention determines the number of invalid servers (nodes) that do not respond (Sc7) and confirms whether the first group is a major group or a minor group, The third server which is the non-response server can be deleted only when the first group is the major group (Sc8). On the other hand, when the first group is a minor group, the server 230 operating in the leader mode determines that there is no right to delete a server existing in the group, and terminates the process. In this case, in the embodiment of the present invention, as shown in FIG. 5, the number of the invalid nodes is judged by judging the major group and the minor group. When the number of invalid nodes is less than the majority, If the number of invalid nodes is greater than or equal to a majority, the first group is determined as a minor group.

In other words, the server 230 in the reader mode participating in the consensus in the distributed network according to the embodiment of the present invention can not access the servers (nodes) existing in the first server only when the first group to which the server belongs is a major group, Can be dynamically changed or deleted by adding or deleting the number.

Next, with reference to FIGS. 7 and 8, a process of adding a new server to a first group of servers in a reader mode participating in a consensus in a distributed network according to an embodiment of the present invention will be described in more detail.

7 is a diagram for explaining an operation procedure of a reader mode in which a group addition request is received from a new server that is not present in the first group according to an embodiment of the present invention.

8 is a flowchart illustrating a process of adding a fourth server in a first group to a server in a reader mode according to an embodiment of the present invention.

Referring to FIGS. 7 and 8, a fourth server 240 operating in the follower mode, which is not present in the first group 240, is a server operating in a reader mode, a candidate mode, and a follower mode, The group addition request message can be transmitted to any one (Sa1).

That is, any one of the plurality of servers existing in the first group can receive the group addition request message from the fourth server 240 that is not present in the first group.

At this time, according to an embodiment of the present invention, when the server operates in the reader mode and receives the group addition message from the fourth server (Sa2), the server 230 in the leader mode transmits And if the first group is a major group, an algorithm for adding a fourth server to the first group may be performed.

On the other hand, when the server according to the embodiment of the present invention operates in the candidate mode or the follower mode and receives the group addition request message from the fourth server (Sa2), the group addition request message received from the fourth server To the server operating in the reader mode existing in the group (Sa3).

The server 230 in the leader mode receiving the group addition request message transmitted from the fourth server 240 can determine whether the server 230 is a leader belonging to the major group or a leader belonging to the minor group (Sa4). That is, it is determined whether the first group to which the user belongs is a major group or a minor group.

Thereafter, if the server 230 in the reader mode is a reader belonging to the minor group, it is terminated (Sa8). If it is a reader belonging to the major group, the fourth server is added to the first group and the information is updated in the log To the server in the follower mode and to the fourth server (Sa5).

According to an embodiment of the present invention, the server 230 operating in the reader mode adds a new node (i.e., fourth server information) to its log (Sl1), commits it (Sl2) And manages a plurality of servers belonging thereto. Here, the node (server) management adds a fourth server to the node list (Sl3), updates the node information (Sl4), and manages the information by including it in the log. At this time, in the process of entering and committing node information in the log, when the server in the reader mode transmits the node addition information to the follower mode server (Ss1), the server 210 in the follower mode updates the node addition information to the log (Ss2 (Ss3). If the log update is not successful, the update is retried. If the update is successful, it is determined whether the first group is a major group or a minor group (Ss4) If the first group is the major group, the commit is performed (Sl 2).

On the other hand, the fourth server receives the updated log information from the server 230 in the reader mode and updates the corresponding log data to its own log (Sn1). Then, the fourth server determines whether the information about the fourth server is added to the node list (Sa2) node management is performed (Sa3), and the own state is changed from the free-follow state to the follower mode (Sn4, Sa7).

Similarly, the server 220 or the follower mode server 210 in the first group can receive the updated log from the server 230 in the reader mode and manage the node through the log list included in the log (Sf1).

One embodiment of the present invention may also be embodied in the form of a recording medium including instructions executable by a computer, such as program modules, being executed by a computer. Computer readable media can be any available media that can be accessed by a computer and includes both volatile and nonvolatile media, removable and non-removable media. In addition, the computer-readable medium may include both computer storage media and communication media. Computer storage media includes both volatile and nonvolatile, removable and non-removable media implemented in any method or technology for storage of information such as computer readable instructions, data structures, program modules or other data. Communication media typically includes any information delivery media, including computer readable instructions, data structures, program modules, or other data in a modulated data signal such as a carrier wave, or other transport mechanism.
In particular, the recording medium may refer to only a computer storage medium.

While the methods and systems of the present invention have been described in connection with specific embodiments, some or all of those elements or operations may be implemented using a computer system having a general purpose hardware architecture.

It will be understood by those skilled in the art that the foregoing description of the present invention is for illustrative purposes only and that those of ordinary skill in the art can readily understand that various changes and modifications may be made without departing from the spirit or essential characteristics of the present invention. will be. It is therefore to be understood that the above-described embodiments are illustrative in all aspects and not restrictive. For example, each component described as a single entity may be distributed and implemented, and components described as being distributed may also be implemented in a combined form.

The scope of the present invention is defined by the appended claims rather than the detailed description and all changes or modifications derived from the meaning and scope of the claims and their equivalents are to be construed as being included within the scope of the present invention do.

110: Network communication module
120: Memory
130: Processor
210: Server in follower mode
220: Server in Candidate Mode
230: Server in Reader mode

Claims (14)

Within a first group, in a server participating in distributed network based consensus,
A log in which the action information is recorded and a memory in which the consensus module is stored, and
And a processor for performing an action according to the log,
Wherein the processor is operative to execute the consensus module
A state of the server is defined as a follower mode, a candidate mode, and a leader mode,
As the server operates in the reader mode
Periodically transmits a heartbeat for confirming a connection state of the network to a plurality of second servers existing in a first group to which the server belongs and operating in a follower mode or a candidate mode, From each of the second servers,
The processor deletes from the plurality of second servers a third server that fails to receive a response to the heartbeat in the first group,
The server information and the changed log information existing in the first group are transmitted to the remaining second servers except for the third server,
Servers participating in distributed network-based consensus. The method according to claim 1,
As the server operates in the reader mode,
The processor
Determining whether the first group is a major group or a minor group,
When the number of responses to the heartbeats received from the plurality of second servers is equal to or more than half of the number of servers belonging to the first group,
Determining the first group as the major group,
When the number of responses to the heartbeats received from the plurality of second servers is less than a majority of the number of servers belonging to the first group,
The first group is determined as a minor group,
Servers participating in distributed network-based consensus. 3. The method of claim 2,
The processor
And delete the third server in the first group only if the first group is a major group.
Servers participating in distributed network-based consensus. 3. The method of claim 2,
The first group is a major group,
When receiving an add request from a fourth server existing outside the first group,
The processor
And transmits information of the fourth server to the plurality of second servers.
Servers participating in distributed network-based consensus. 3. The method of claim 2,
The processor
Repeatedly transmitting the heartbeat to the third server according to a predetermined period,
When the number of responses to the heartbeat that has not been received from the third server exceeds a predetermined number,
And deletes the third server in the first group.
Servers participating in distributed network-based consensus. 6. The method of claim 5,
The third server
Wherein a response time to the heartbeat of the plurality of second servers exceeds a predetermined time,
Servers participating in distributed network-based consensus. The method according to claim 1,
As the server operates in Candidate mode or Follower mode,
And receives server information and modified log information existing in the first group from a server in the reader mode existing in the first group.
Servers participating in distributed network-based consensus. A network method of a server including a processor participating in a distributed network-based consensus in a first group, a memory in which action information is recorded and a consensus module are stored, and a processor performing an action according to the log,
(A) defining a state of the server as a leader mode as the processor executes the consensus module;
(B) transmitting a heartbeat for confirming a connection state of the network to a plurality of second servers, the processor residing in the first group and operating in a follower mode or a candidate mode;
(C) receiving, by the processor, a response to the heartbeat from each of the second servers;
(D) deleting, in the first group, a third server among the plurality of second servers, the processor not receiving a response to the heartbeat; And
And (e) transmitting and updating the server information and the changed log information existing in the first group to each of the remaining second servers except for the third server.
Network method of servers participating in distributed network based consensus. 9. The method of claim 8,
After the step (c)
(F) determining whether the first group is a major group or a minor group,
The step (f)
When the number of responses to the heartbeats received from the plurality of second servers is equal to or more than half of the number of servers belonging to the first group,
Determining the first group as the major group,
When the number of responses to the heartbeats received from the plurality of second servers is less than a majority of the number of servers belonging to the first group,
The first group is determined as the minor group,
Network method of servers participating in distributed network based consensus. 10. The method of claim 9,
The step (d)
And delete the third server in the first group only if the first group is a major group.
Network method of servers participating in distributed network based consensus. 10. The method of claim 9,
Further comprising: (g) receiving an add request from a fourth server existing outside the first group,
If it is determined that the first group is a major group or a minor group in the step (f)
Further comprising: (h) after the step (g), the processor transmits information of the fourth server to the plurality of second servers.
Network method of servers participating in distributed network based consensus. 10. The method of claim 9,
The step (f)
(I) repeatedly transmitting the heartbeat to the third server according to a predetermined period,
When the number of responses to the heartbeat that has not been received from the third server exceeds a predetermined number,
(D). ≪ RTI ID = 0.0 >
Network method of servers participating in distributed network based consensus. 10. The method of claim 9,
In the step (f)
Wherein the third server defines a server whose response time to the heartbeat exceeds a preset time among the plurality of second servers.
Network method of servers participating in distributed network based consensus. A computer storage medium on which a program for executing a network method of a server according to any one of claims 8 to 13 is recorded.

Images