CN120711012A - Cluster management method, device, computer equipment and storage medium - Google Patents

Abstract

The present application relates to a cluster management method, apparatus, computer device, storage medium and computer program product. The method comprises the steps of responding to the global master node to group the node clusters to obtain a plurality of node groups, determining a target node group to which the target node belongs from the node groups, determining a target machine subgroup associated with the target node group from machine groups, associating different machine subgroups in the machine groups with each node group, enabling a local master node to be used for sending load information of each machine in the machine subgroup associated with the node group to routing equipment, obtaining the load information of each machine in the target machine subgroup, and sending the load information of each machine in the target machine subgroup to the routing equipment when the target node is the local master node in the target node group, wherein each load information is used for carrying out load balancing on the machines in the machine group by the routing equipment. By adopting the method, the system stability can be improved.

Description

Cluster management method, device, computer equipment and storage medium

Technical Field

The present application relates to the field of computer technology, and in particular, to a cluster management method, apparatus, computer device, storage medium, and computer program product.

Background

With the rapid development of internet technology, various internet applications are increasing, and the amount of data on a network is increasing. For the background system of the application or the platform, when the access amount of the background system is large, the load of each machine of the background system is increased, so that the stability of the background system is affected, and therefore, the load balancing is needed for the machine.

In the traditional technology, load balancing is performed by collecting load information of machines through a control center of a background system, however, when the number of machines is large, the load of the control center is increased, so that the processing capacity of the control center is reduced, request blocking is caused, and the stability of the background system is affected.

Disclosure of Invention

In view of the foregoing, it is desirable to provide a cluster management method, apparatus, computer device, computer-readable storage medium, and computer program product that can provide system stability.

In one aspect, the present application provides a cluster management method, which is executed by a target node in a node cluster, where the node cluster includes a global master node and at least two global standby nodes, and the target node is one of the global master node or the global standby nodes, and includes:

responding to the global master node to group the node clusters to obtain a plurality of node groups, and determining a target node group to which the target node belongs from the plurality of node groups, wherein each node group comprises at least one global standby node;

determining a target machine subgroup associated with the target node group from machine groups, wherein each node group is associated with a different machine subgroup in the machine group;

Acquiring load information of each machine in the target machine subset;

And when the target node is a local master node in the target node group, sending load information of each machine in the target machine subgroup to the routing equipment, wherein each load information is used for carrying out load balancing on the machines in the machine group by the routing equipment.

On the other hand, the application also provides a cluster management device, which comprises:

The node group determining module is used for responding to the global master node to group the node clusters to obtain a plurality of node groups, and determining a target node group to which a target node belongs from the plurality of node groups; each node group comprises at least one global standby node; for each node group, one node in the node group is a local master node, and the rest nodes are local standby nodes serving as standby local master nodes; the node cluster comprises a global main node and at least two global standby nodes, and the target node is one of the global main node or the global standby nodes;

The machine subgroup determining module is used for determining a target machine subgroup associated with the target node group from machine groups, wherein each node group is associated with a different machine subgroup in the machine groups;

The load information acquisition module is used for acquiring the load information of each machine in the target machine subset;

and the load information sending module is used for sending the load information of each machine in the target machine subset to the routing equipment when the target node is a local master node in the target node group, wherein each load information is used for carrying out load balancing on the machines in the machine set by the routing equipment.

In some embodiments, the apparatus further includes an update processing module, configured to, when the target node is a global master node, receive a first update instruction sent by an operation device and indicating that configuration information including the machine group has been updated, compare, in response to the first update instruction, a locally stored version number of the machine group with a version number of the machine group in the configuration information, and when the comparison is inconsistent, determine that the machine group has been updated, and execute the step of determining, from the slave machine groups, that the target machine group associated with the target node group is again to update the target machine group.

In some embodiments, the configuration information further includes a machine group configuration item related to each machine in the machine group, and the update processing module is further configured to compare the locally stored machine group configuration item with the machine group configuration item in the configuration information, obtain the machine group configuration item in the configuration information when the comparison is inconsistent, update the locally stored machine group configuration item with the machine group configuration item in the configuration information to obtain an updated machine group configuration item, and update the configuration of each machine in the target machine group with the updated machine group configuration item.

In some embodiments, the load information obtaining module is further configured to send a machine heartbeat packet carrying the updated machine group configuration information to each machine in the target machine subset when the target node is a local master node in the target node group, where the machine heartbeat packet is configured to instruct each machine to return the respective load information and perform configuration update with the updated machine group configuration information, and receive load information returned by each machine in the target machine subset in response to the machine heartbeat packet.

In some embodiments, the machine subgroup determining module is further configured to obtain a subgroup identifier corresponding to the target node group determined by the global master node, where the subgroup identifier indicates a machine subgroup in the machine group, and obtain, from the machine group, a machine subgroup indicated by a subgroup identifier corresponding to the target node group, so as to obtain a target machine subgroup associated with the target node group.

In some embodiments, the apparatus further includes an identification processing module, where the identification processing module is configured to generate a plurality of subgroup identifications based on the machine group when the target node is a global master node, establish a one-to-one correspondence between each subgroup identification and each node group, determine, for each global backup node, a matching node group to which the global backup node belongs, send, to the global backup node, a subgroup identification corresponding to the matching node group to indicate the global backup node, and obtain, from the machine group, a machine subgroup indicated by the subgroup identification corresponding to the matching node group, so as to obtain a machine subgroup associated with the matching node group.

In some embodiments, the identifier processing module is further configured to send, when it is determined that the machine group has been updated, a second update instruction carrying a subgroup identifier corresponding to the matching node group to the global standby node, where the update processing module is further configured to, when it is determined that the machine group has not been updated, the configuration information further includes a machine group configuration item related to each machine in the machine group, send, to the global standby node, a third update instruction carrying no subgroup identifier, where the third update instruction is configured to instruct the global standby node to update a locally stored machine group configuration item using the machine group configuration item in the configuration information.

In some embodiments, the load information sending module is further configured to send, to the routing device, a route heartbeat packet carrying load information of each machine in the target machine subset and the target machine subset, in a case that the target machine subset has been updated, so that the routing device updates the target machine subset, and obtain the load information of each machine in the target machine subset.

In some embodiments, the apparatus further includes a master node determining module, where the master node determining module is configured to query, from a storage device, time information associated with a query word by using a preset query word, determine that the target node is a global master node when the time information associated with the query word is not queried, and store the time information of the current time point and the time information associated with the query word to the storage device.

In some embodiments, the master node determining module is further configured to determine, when time information associated with the query word is queried, a time interval between a time point represented by the queried time information and a current time point, determine, when the time interval does not reach a time interval threshold, that the target node is a global standby node, determine, when the time interval reaches the time interval threshold, that the target node is a global master node, and update, as time information of the current time point, the time information associated with the query word stored in the storage device.

In some embodiments, the apparatus further includes a grouping module configured to group the node cluster to obtain a plurality of node groups when the target node is a global master node, and execute the step of grouping the node cluster to obtain the plurality of node groups again to update the plurality of node groups when it is determined that the node cluster has changed.

In some embodiments, the node group determining module is further configured to share, with each global standby node, grouping information for indicating the plurality of node groups, where the grouping information is used to indicate the global standby node, determine the node group from the plurality of node groups to which the global standby node belongs, and send, when the global standby node is a local master node in the node group to which the global standby node belongs, load information of each machine in a subset of machines associated with the node group to which the global standby node belongs to the routing device.

In another aspect, the present application further provides a computer device, including a memory and a processor, where the memory stores a computer program, and the processor implements the steps in the cluster management method when executing the computer program.

In another aspect, the present application also provides a computer readable storage medium, on which a computer program is stored, which when executed by a processor, implements the steps of the cluster management method described above.

In another aspect, the present application also provides a computer program product comprising a computer program which, when executed by a processor, implements the steps of the cluster management method described above.

The cluster management method, apparatus, computer device, storage medium and computer program product, because the node cluster includes a global master node and at least two global standby nodes, the master node and the standby nodes are globally distinguished, the global master node groups the node cluster into a plurality of node groups, one node in the node groups is a local master node in the node groups, the rest nodes are local standby nodes standby for the local master node, thus locally distinguishing the master node and the standby nodes, namely realizing multi-stage master-slave division, under the condition of multi-stage master-slave division, each node group is associated with different machine subgroups in the same machine group, determining the target node group to which the target node belongs from the plurality of node groups, the method comprises the steps of determining a target machine subgroup associated with a target node group, acquiring load information of each machine in the target machine subgroup, and sending the load information of each machine in the target machine subgroup to the routing equipment when the target node is a local main node in the target node group so as to instruct the routing equipment to carry out load balancing on the machines in the target machine subgroup, so that when the overall flow of each machine in the machine group is large, the problem of large flow can be relieved by adding the machines and adding the nodes, and when the local main node in the node group has a problem, the starting of standby local standby nodes can be supported, thereby improving the stability of a system to which the machines, the nodes and the routing equipment belong.

Drawings

In order to more clearly illustrate the embodiments of the present application or the technical solutions in the related art, the drawings that are required to be used in the embodiments or the related technical descriptions will be briefly described, and it is apparent that the drawings in the following description are only some embodiments of the present application, and other drawings may be obtained according to the drawings without inventive effort for those skilled in the art.

FIG. 1 is a diagram of an application environment for a cluster management method in one embodiment;

FIG. 2 is a flow diagram of a cluster management method in one embodiment;

FIG. 3 is a schematic diagram of a cyclic sequence of nodes in one embodiment;

FIG. 4 is a block diagram of a cluster management method according to one embodiment;

FIG. 5 is a flow diagram of determining a global master node in one embodiment;

FIG. 6 is a schematic diagram of load balancing using a cluster management approach in one embodiment;

FIG. 7 is a flow chart of a cluster management method according to another embodiment;

FIG. 8 is a flow chart of a cluster management method in yet another embodiment;

FIG. 9 is a block diagram of a cluster management device in one embodiment;

FIG. 10 is an internal block diagram of a computer device in one embodiment;

FIG. 11 is an internal block diagram of a computer device in one embodiment.

Detailed Description

The present application will be described in further detail with reference to the drawings and examples, in order to make the objects, technical solutions and advantages of the present application more apparent. It should be understood that the specific embodiments described herein are for purposes of illustration only and are not intended to limit the scope of the application.

The cluster management method provided by the embodiment of the application can be applied to an application environment shown in figure 1. The application environment includes a terminal 102, a node cluster 104, a machine cluster 106, and a routing device 108. Wherein the terminals 102, the node clusters 104, the machine clusters 106, and the routing devices 108 communicate over a network. The data storage system may store data that the machines in the machine cluster 106 need to process. The data storage system may be integrated on the machine or may be located on the cloud or other network server. The node cluster 104 is configured to manage the machine cluster 106, and may obtain load information of the machine, send the load information to the routing device 108, and the routing device 108 may load balance the machine according to the load information. When the terminal 102 requests to establish a communication connection with a machine in the machine cluster 106, the routing device 108 determines a target machine from the machine cluster 106 according to load information of the machine in the machine cluster 106, and returns a network address, e.g., an IP address, of the target machine to the terminal 102, and the terminal 102 may establish a communication connection, e.g., a long connection, with the target machine according to the network address of the target machine, so that the terminal 102 may communicate with the target machine according to the established connection. Machines in the machine cluster 106 require load balancing.

Specifically, the node cluster 104 distinguishes between a master node and a standby node, where the standby node is a node that is standby for the master node, and when a problem occurs in the master node, the standby node switches to the master node to take on the role of the master node. The node group also distinguishes between primary and backup nodes. The master nodes in the node group are called local master nodes, and the standby nodes in the node group are called local standby nodes. The master node in the node cluster 104 is called a global master node, the standby nodes in the node cluster 104 are called global standby nodes, and one global master node and at least two global standby nodes exist in the node cluster 104. To load balance the machine cluster 106, the global master node may group the node clusters to obtain a plurality of node groups, each node group including at least one global backup node. For any node in the node cluster, the any node may be one of a global master node or a global standby node, and in the present application, in order to distinguish different nodes, the any node is referred to as a target node, that is, the target node is any node in the node cluster. The cluster management method provided by the application can be applied to each node in the node cluster, and each node in the node cluster can execute the cluster management method provided by the application. The target node responds to the global master node to group the node clusters to obtain a plurality of node groups, the target node group to which the target node belongs can be determined from the plurality of node groups, the target machine subgroup associated with the target node group is determined, and each node group is associated with a different machine subgroup in the same machine group. The target node may obtain load information for each machine in the target subset of machines. And when the target node is a local master node in the target node group, the target node sends load information of each machine in the target machine subgroup to the routing equipment. The routing device may load balance the machines in the target machine subset according to the load information.

The terminal 102 may be, but not limited to, various personal computers, notebook computers, smart phones, tablet computers, intelligent voice interaction devices, intelligent home appliances, vehicle-mounted terminals, aircrafts, internet of things devices and portable wearable devices, and the internet of things devices may be intelligent sound boxes, intelligent televisions, intelligent air conditioners, intelligent vehicle-mounted devices and the like. The portable wearable device may be a smart watch, smart bracelet, headset, or the like. Each machine and each node can be implemented by a server, the nodes, the machines and the routing equipment can be implemented by independent servers or a server cluster formed by a plurality of servers, and the nodes, the machines and the routing equipment can also be cloud servers, wherein the cloud servers are used for providing cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communication, middleware services, domain name services, security services, CDNs (Content Delivery Network, content distribution networks), basic cloud computing services such as big data and artificial intelligent platforms and the like. The terminals 102, node clusters 104, machine clusters 106, and routing devices 108 may be directly or indirectly connected via wired or wireless communication, as the application is not limited in this regard. The machine may also be an instance of a cloud server, the instance being a base unit in a cloud computing infrastructure, the instance may be a virtual machine or a virtual server, for example.

The cloud storage (cloud storage) is a new concept that extends and develops in the concept of cloud computing, and the distributed cloud storage system (hereinafter referred to as a storage system for short) refers to a storage system that provides data storage and service access functions for the outside through aggregation of a large number of storage devices (storage devices are also referred to as storage nodes) of different types in a network through application software or application interfaces by means of functions such as cluster application, grid technology, and distributed storage file systems.

At present, a storage method of a storage system is that logical volumes are created, and when the logical volumes are created, a physical storage space is allocated to each logical volume, where the physical storage space may be a disk of a certain storage device or a plurality of storage devices. The client stores data on a certain logical volume, that is, the data is stored on a file system, the file system divides the data into a plurality of parts, each part is an object, the object not only contains the data but also contains additional information such as a data Identification (ID) and the like, the file system writes each object into a physical storage space of the logical volume, and the file system records storage position information of each object, so that when the client requests to access the data, the file system can enable the client to access the data according to the storage position information of each object.

The storage system allocates physical storage space for a logical volume, specifically, the physical storage space is divided into stripes in advance according to the set of capacity estimation of objects stored in the logical volume (the estimation often has a large margin with respect to the capacity of the objects actually to be stored) and redundant array of independent disks (RAID, redundant Array of INDEPENDENT DISK), and one logical volume can be understood as one stripe, so that the physical storage space is allocated for the logical volume.

Cloud computing (cloud computing) refers to the delivery and usage mode of an IT infrastructure, to obtaining required resources in an on-demand, easily scalable manner over a network, and generalized cloud computing refers to the delivery and usage mode of a service, to obtaining required services in an on-demand, easily scalable manner over a network. Such services may be IT, software, internet related, or other services. Cloud Computing is a product of fusion of traditional computer and network technology developments such as Grid Computing (Grid Computing), distributed Computing (DistributedComputing), parallel Computing (Parallel Computing), utility Computing (Utility Computing), network storage (Network Storage Technologies), virtualization (Virtualization), load balancing (Load Balance), and the like.

With the development of the internet, real-time data flow and diversification of connected devices, and the promotion of demands of search services, social networks, mobile commerce, open collaboration and the like, cloud computing is rapidly developed. Unlike the previous parallel distributed computing, the generation of cloud computing will promote the revolutionary transformation of the whole internet mode and enterprise management mode in concept.

In some embodiments, as shown in fig. 2, a cluster management method is provided, which may be performed by any node in a node cluster, and the method is applied to a target node in the node cluster, for example, and includes the following steps 202 to 210. Wherein:

step 202, responding to a global master node to group node clusters to obtain a plurality of node groups, determining a target node group to which a target node belongs from the plurality of node groups, wherein each node group comprises at least one global standby node, one node in each node group is a local master node, and the rest nodes are local standby nodes serving as local master nodes.

Each node group comprises at least two nodes, and each node group comprises at least one global standby node because the node cluster comprises only one global main node. At least one node of the different node groups is different, or the same node does not exist in the different node groups. Each node group may include the same number of nodes. The number of node groups may be consistent with the number of nodes, or the number of node groups may be less than or greater than the number of nodes.

Specifically, the global master node may group the node clusters in two cases, one in which the node clusters have not been grouped, and another in which the node clusters have been grouped into multiple node groups, but the node clusters have changed, e.g., the nodes have been added or subtracted from the node clusters.

In some embodiments, where the target node is a global master node, the plurality of node groups are generated by the target node. The global master node may arrange the nodes in the node cluster to obtain a node sequence, segment the node sequence to obtain a plurality of subsequences, and determine each subsequence as a node group. The nodes are ordered according to the network address of the nodes or randomly. When cutting, the cutting can be uniformly cut or randomly cut. Taking the example of uniform segmentation and the length of each sub-sequence being 2, when the node cluster is (node A, node B, node C, node D), the node sequence is [ node A, node B, node C, node D ], the segmented sub-sequences are [ node A, node B ] and [ node C, node D ].

In some embodiments, the global master node may select a specified number of nodes from the node cluster to form a node group, and select the node group from the node cluster multiple times until no new node group is selected. Taking the designated number of 2 as an example, when the node cluster is (node a, node B, node C, node D), the obtained node groups are (node a, node B), (node a, node C), (node a, node D), (node B, node C), (node B, node D), (node C, node D). When the node cluster is (node A, node B, node C), the obtained node groups are (node A, node B), (node A, node C), (node B, node C).

In some embodiments, after the global master node obtains the node sequence, subsequences with specified lengths can be obtained from the node sequence at specified intervals, so as to obtain a plurality of subsequences, and each subsequence is used as a node group. The designated interval and the designated length can be set according to the requirement, taking the designated interval as 1 and the designated length as 2 as an example, when the node cluster is (node A, node B and node C), the obtained subsequences are [ node A, node B ] and [ node B, node C ].

The target node group refers to a node group containing target nodes. At least one target node group may be present in the plurality of node groups, for example, 1 or 2 target node groups may be present. For example, the plurality of node groups are (node a, node B), (node a, node C), (node B, node C), respectively, and if the target node is node a, there are 2 target node groups (node a, node B), (node a, node C), respectively.

The global master node can arrange all nodes in the node cluster to obtain a ring-shaped node sequence, and obtain a plurality of subsequences from the node sequence to obtain a plurality of node groups. The ring-shaped node sequence refers to nodes adjacent from end to end, and takes node clusters as (node A, node B and node C) as an example, and as shown in fig. 3, a schematic diagram of the ring-shaped node sequence is shown.

In some embodiments, the server may select all sub-sequences of a specified length from the ring-shaped node sequence to obtain a plurality of node groups, e.g., a specified length of 2, and the node sequence in fig. 3 may obtain 3 node groups, (node a, node B), (node B, node C), and (node C, node a), respectively. By adopting the method, the number of the obtained node groups is consistent with the number of the nodes of the node cluster, the occurrence frequency of the nodes in the node groups is consistent, the load balance of each node is facilitated, and the system stability is improved. For example, if the node cluster has 3 nodes, 3 node groups can be obtained, and each node appears 2 times.

Step 204, determining target machine sub-groups associated with the target node groups from the machine groups, wherein each node group is associated with a different machine sub-group in the machine group.

Wherein the machine group is recorded in a database. The machine group records a plurality of machines, for example, the machine group records each machine in the machine cluster in fig. 1. A machine subset refers to a portion of a machine set, e.g., the machine set may be a list of machines, and the machine subset is a sub-list of the list of machines. Each machine in the same machine group is configured to perform the same type of task, such as performing a data transfer or forwarding task, e.g., the machine may be a media proxy device. And the local master node in each node group is used for sending the load information of each machine in the machine subgroup associated with the node group to the routing equipment.

Specifically, for each node group, each node in the node group may periodically obtain load information of each machine in the machine subgroup associated with the node group, and when the node is a local master node, the load information of each machine in the machine subgroup associated with the node group may also be sent to the routing device.

In some embodiments, the machine group may be a machine list, when the target node is a global master node, the target node may determine size information of the machine list, where the size information includes at least one of a number of rows or a number of columns of the machine list, and the target node may generate a plurality of sub-list identifiers according to the size information, where each sub-list identifier indicates one sub-list in the machine list, and one sub-list is a subgroup, and the sub-list identifiers are subgroup identifiers. The number of subgroup identifications corresponds to the number of node groups. The target node may establish a one-to-one correspondence between the respective subgroup identifiers and the respective node groups, i.e. each subgroup identifier corresponds to a node group and one node group corresponds to a subgroup identifier. The subgroup identification is used to indicate a subgroup of the machine group, the subgroup being part of the machine group. For example, when one machine is recorded per row or one machine is recorded per column in a machine group, the subgroup identification may include a start position and an offset, the start position being the start row or the start column. The target node may determine a subgroup identifier corresponding to the target node group, and form the target node into a subgroup indicated by the corresponding subgroup identifier as a machine subgroup associated with the target node group.

Step 206, obtaining the load information of each machine in the target machine subset.

Wherein the load information includes at least one of CPU (Central Processing Unit ) usage, memory usage, network bandwidth, network throughput, network delay, traffic, etc.

Specifically, the target node may periodically send load acquisition requests to each machine in the target machine subset, and the machines may return load information acquired during the request interval to the target node. The request interval refers to an interval between a time when one load acquisition request is received and a time when a next load acquisition request is received. The load request may be implemented using a heartbeat packet, and the request interval may be a heartbeat interval. The heartbeat interval refers to an interval between a time when one heartbeat packet is received and a time when the next heartbeat packet is received. Each node can acquire the load information of each machine in the machine subgroup associated with the node group by adopting the method provided by the embodiment.

In some embodiments, the load information may be stored after the load information is obtained by the target node. For example, the load information may be stored in a memory. The step of obtaining load information is performed for each node in the node group, but only the local master node in the node group will send load information to the routing device.

And step 208, when the target node is a local master node in the target node group, sending load information of each machine in the target machine subgroup to the routing equipment, wherein each load information is used for carrying out load balancing on the machines in the machine group by the routing equipment.

The routing equipment is used for carrying out load balancing on all machines in the machine group. The routing device, the machine and the node cluster belong to the same system. The system may be any system including, but not limited to, a system that is an online meeting platform, a system that is a social platform, or a system that is an instant messaging platform. The online conference platform is an online conference tool, provides remote communication and cooperation platform for enterprises and individuals, and provides functions of audio and video conference, online cooperation, conference recording, simultaneous interpretation and the like. The social platform is a platform for social contact, provides a platform for content sharing, and can receive shared content for sharing. The instant communication platform is used for providing instant communication service and also can provide a content sharing function, and can receive the shared content for sharing.

Specifically, for each node group, the local master node in the node group sends load information of each machine in the machine subgroup associated with the node group to the routing device. When the target node is a local master node in the target node group, the target node may periodically send load information of each machine in the target machine subset to the routing device. For example, the destination node may periodically send a heartbeat packet to the routing device, where the heartbeat packet carries load information for each machine in the subset of destination machines. It should be noted that, the heartbeat packet sent by the node to the machine and the heartbeat packet sent by the node routing device are two different heartbeat packets. The heartbeat packet sent by the node to the machine may be referred to as a machine heartbeat packet, and the heartbeat packet sent by the node to the routing device may be referred to as a routing heartbeat packet. Since for each node group the local master node in that node group will send the load information of each machine in the subgroup of machines associated with that node group to the routing device. So that the routing device can obtain load information for each machine in the set of machines.

In some embodiments, the routing device receives a machine connection request sent by the terminal, and the routing device selects a target machine from each machine in the machine group according to the load information of each machine in the machine group, and returns a network address of the target machine to the terminal. The terminal receives the network address, sends a connection request to the target machine according to the network address, and the target machine responds to the connection request to establish communication connection with the terminal. In the process of selecting the target machine, in order to balance the load of each machine, the target machine may be selected from the machines according to the order of low load to high load, for example, the machine with the lowest load may be selected as the target machine.

In some embodiments, the communication network employed by the machines in the machine group may be an operator provided network or a CAP (Content Acceleration Platform ) provided network. Operators include, but are not limited to, at least one of telecommunications, mobile, or Unicom. Machines employing different communication networks may be present in each machine. As shown in fig. 4, there is illustrated a machine employing different communication networks, [ telecom ] representing a network provided by a telecom operator, [ UNICOM ] representing a network provided by a UNICOM operator, [ mobile ] representing a network provided by a mobile operator, [ CAP ] representing a network provided by a CAP platform. The terminal request 1 obtains the network address of the machine, namely sends the machine connection request, after authentication by the authentication module, the routing equipment responds to the machine connection request, and the routing equipment selects the target machine from all the machines. The routing device may determine a network type adopted by the terminal, select at least one machine having a network type consistent with the network type adopted by the terminal from the machines, and then determine a target machine from the selected at least one machine, for example, may determine the target machine according to respective load information of the at least one machine. Of course, the routing device may determine the target machine from the machines recorded in the machine group directly according to the load information without distinguishing the network type adopted by the terminal. The system comprises an authentication module, a signaling access module, routing equipment, a node cluster and a machine, and belongs to a background system. The system may be any system including, but not limited to, a system that is an online meeting platform, a system that is a social platform, or a system that is an instant messaging platform. The routing device returns the network address of the target machine to the terminal, which connects the target machine according to the network address, e.g. "2. Connect machine" in fig. 4.

In the cluster management method, since the node cluster comprises a global main node and at least two global standby nodes, the main node and the standby nodes are globally distinguished, the global main node groups the node cluster into a plurality of node groups, one node in the node groups is a local main node in the node groups, and the rest node is a local standby node for standby of the local main node, so that the main node and the standby nodes are locally distinguished, namely, multi-stage main and standby division is realized, under the condition of multi-stage main and standby division, each node group is associated with different machine subgroups in the same machine group, the target node group to which the target node belongs is determined from the plurality of node groups, the target machine subgroup associated with the target node group is determined, the load information of each machine in the target machine subgroup is acquired, when the target node is the local main node in the target node group, the load information of each machine in the target machine subgroup is transmitted to the routing equipment, so as to instruct the routing equipment to perform load balancing on the machines in the target machine subgroup, namely, each node group is associated with one machine subgroup, in the node groups, namely, each node group is managed by the plurality of nodes, and the number of machines in the same machine group can be increased to the local machine subgroup, and the number of the backup machine group can be increased, and the number of the machines can be increased to the local machine group is increased, and the number of the machines can be increased to the local and the machine has a large number.

In some embodiments, the method further comprises the steps of receiving a first update instruction sent by the operation device and indicating that the configuration information containing the machine group has been updated when the target node is a global master node, comparing the locally stored machine group version number with the version number of the machine group in the configuration information in response to the first update instruction, determining that the updated machine group is inconsistent in comparison, and executing the step of determining a target machine subgroup associated with the target node group from the machine group again to update the target machine subgroup.

The target node is a global master node, which is any node, and can be understood as a master angle of the target node in the node cluster, and the target node is a global standby node, which can be understood as a slave role of the target node in the node cluster. Likewise, the target node is a local master node in the target node group, which can be understood as the target node being the master angle in the target node group. The target node is a local master node in the target node group, which can be understood as a slave role in the target node group.

The updating of the machine set includes at least one of removing machines from the machine set or adding machines to the machine set. Under the condition that the machine group is not changed, after each node in the node cluster determines the machine subgroup associated with the node group to which the node group belongs, the repeated process of acquiring the machine subgroup is not needed. However, when a change has occurred to the machine group, for example, a machine is added or deleted from the machine group, each node in the node cluster needs to perform the process of acquiring the machine subgroup from the machine group again, so as to update the machine subgroup with the update of the machine group. The configuration information is stored in a database.

In particular, the identity of each machine in the set of machines may be stored in a database in the form of a table, and the identity of the machine may be the name of the machine or the IP (Internet Protocol) address of the machine. When the operation device updates the machine group, the operation device may update the machine group recorded in the database, and then notify the global master node that the machine group has been updated, that is, send a first update instruction to the global master node. Before updating the machine group in the database, the operation equipment can copy the machine group to obtain a backup machine group, generate a corresponding version number for the backup machine group, store the backup machine group corresponding to the version number in the database, and update the machine group after backup to obtain the current latest machine group. When rollback is required, the current up-to-date machine set may be rolled back using the backup machine set.

In this embodiment, when the machine set has been changed, the step of determining the target machine subset associated with the target node set from the machine set is performed again, so that the machine subset can be updated along with the update of the machine set, and the correctness of the machine subset is ensured.

In some embodiments, the configuration information further includes machine group configuration items associated with each machine in the machine group, and the method further includes comparing the locally stored machine group configuration items with the machine group configuration items in the configuration information, obtaining the machine group configuration items in the configuration information when the comparison is inconsistent, updating the locally stored machine group configuration items with the machine group configuration items in the configuration information to obtain updated machine group configuration items, and updating the updated machine group configuration items for configuring and updating each machine in the target machine subset.

The machine group configuration item is related to each machine in the machine group and is used for uniformly configuring each machine in the machine group, and the machine group configuration item comprises, but is not limited to, time for restarting or time for upgrading the machine and the like.

Specifically, for each node in each node group, the locally stored machine group configuration items can be compared with the machine group configuration items in the configuration information, when the comparison is inconsistent, the machine group configuration items in the configuration information are obtained, and the machine group configuration items in the configuration information are adopted to update the locally stored machine group configuration items so as to obtain updated machine group configuration items. For each node group, the local backup node in the node group is not responsible for sending the updated machine group configuration item to each machine in the machine subgroup associated with the node group. And the local master node in the node group is responsible for sending the updated machine group configuration items to all machines in the machine subgroup associated with the node group, so that the consistency of the machine group configuration items can be ensured.

In this embodiment, under the condition that the configuration information is changed, whether the configuration item of the machine set is changed or not can be quickly determined, and the configuration item of the machine set is updated in time.

In some embodiments, obtaining load information of each machine in the target machine subset includes sending a machine heartbeat packet carrying updated machine set configuration information to each machine in the target machine subset when the target node is a local master node in the target node set, the machine heartbeat packet being used for indicating each machine to return respective load information and performing configuration update by using the updated machine set configuration information, and receiving load information returned by each machine in the target machine subset in response to the machine heartbeat packet.

Specifically, for each node group, the local master node in the node group may send a machine heartbeat packet carrying updated machine group configuration information to each machine in the machine subgroup associated with the node group. After the machine receives the machine heartbeat packet, the machine updates the locally stored machine group configuration information by adopting the updated machine group configuration information, obtains the current load information, and returns the load information to the local main node.

In this embodiment, since the heartbeat packet is periodically sent, the data processing efficiency can be improved compared with the data packet with a new type of reconfigurability by sending updated machine group configuration information of the heartbeat packet and obtaining the load information.

In some embodiments, determining the target machine subgroup associated with the target node group from the machine group comprises obtaining a subgroup identifier corresponding to the target node group determined by the global master node, wherein the subgroup identifier indicates a machine subgroup in the machine group, and obtaining the machine subgroup indicated by the subgroup identifier corresponding to the target node group from the machine group to obtain the target machine subgroup associated with the target node group.

Specifically, the global master node may divide the machine group into a plurality of machine subgroups, determine a subgroup identifier of each machine subgroup, and establish a correspondence between the subgroup identifiers and the node groups. And for each global standby node, sharing the subgroup identifier corresponding to the node group to which the global standby node belongs to the global standby node, so that the global standby node can acquire the machine subgroup indicated by the subgroup identifier corresponding to the node group from the machine group to obtain the machine subgroup associated with the node group.

In some embodiments, the machine group is stored in a database, and the target node may query the machine subgroup indicated by the subgroup identifier corresponding to the target node group from the machine group stored in the database, as the target machine subgroup associated with the target node group.

In this embodiment, the machine subset is obtained according to the subset identifier, so that the amount of transmission data can be reduced when the machine subset is longer than when the machine subset is directly transmitted.

In some embodiments, the method further comprises the steps of generating a plurality of subgroup identifications based on the machine group when the target node is a global master node, establishing a one-to-one correspondence between each subgroup identification and each node group, determining a matched node group to which the global backup node belongs for each global backup node, sending the subgroup identifications corresponding to the matched node group to the global backup node to indicate the global backup node, and acquiring the machine subgroup indicated by the subgroup identifications corresponding to the matched node group from the machine group to obtain the machine subgroup associated with the matched node group.

The matching node group refers to a node group to which the global standby node belongs. The machines in the machine group may be ordered, for example, a list of machines, and the subgroup identification may be represented by a starting position and an offset.

Specifically, for each global standby node, the target node may determine a node group including the global standby node from the node groups, and determine the determined node group as a matching node group to which the global standby node belongs. For example, the node cluster is (node a, node B, node C), and the node cluster is grouped into 3 node groups, namely (node a, node B), (node B, node C) and (node C, node a), and if the node a is a global standby node, (node a, node B) and (node C, node a) have node a, so that the (node a, node B) and (node C, node a) are matching node groups to which the node a belongs. The global standby node may use the subgroup indicated by the subgroup identifier corresponding to the matching node group as a machine subgroup associated with the matching node group. After the machine group is updated, the global main node generates a plurality of subgroup identifications again based on the machine group, and sends the subgroup identifications corresponding to the matched node group to the global standby node again to inform the global standby node of acquiring the machine subgroup indicated by the subgroup identifications corresponding to the matched node group from the machine group again based on the subgroup identifications corresponding to the matched node group, so as to obtain the machine subgroup associated with the matched node group, and update the machine subgroup.

In some embodiments, the target node may generate the plurality of subgroup identifications in a manner that evenly divides the machine groups such that the machine subgroup indicated by each subgroup identification is uniform in size. For example, assuming that 3000 machines are recorded in a machine group, a node cluster is (node a, node B, node C), and there are 3 node groups, respectively (node a, node B), (node B, node C), and (node C, node a), 3 subgroup identifications are generated in such a manner that the machine group is uniformly divided into 3 machine subgroups. The subgroup mark 1 is that the initial position is 0, the offset is 1000, the subgroup mark 2 is that the initial position is 1000, the offset is 1000, the subgroup mark 3 is that the initial position is 2000, and the offset is 1000. Of course, the target node may also generate multiple subgroup identifications in a manner that non-uniformly divides the machine groups. I.e. the size of the machine sub-groups may not be uniform.

In this embodiment, after each subgroup identifier is generated, the target node sends the subgroup identifier corresponding to the matched node group to the global standby node, so that the global standby node can quickly obtain the machine subgroup according to the subgroup identifier.

In some embodiments, the method comprises the steps of sending a subgroup identifier corresponding to a matched node group to a global standby node, when the updated machine group is determined, sending a second updating instruction carrying the subgroup identifier corresponding to the matched node group to the global standby node, when the un-updated machine group is determined, the configuration information further comprises machine group configuration items related to all machines in the machine group, sending a third updating instruction carrying no subgroup identifier to the global standby node, and the third updating instruction is used for indicating the global standby node to update the locally stored machine group configuration items by adopting the machine group configuration items in the configuration information.

Specifically, after receiving the first update instruction, the global master node may determine that the configuration information is changed, and if the configuration information includes a machine group and a machine group configuration item, and if the machine group in the configuration information is changed, the global master node regroups the machine group according to the new machine group, generates a plurality of subgroup identifiers, and sends a second update instruction carrying subgroup identifiers corresponding to the matched node group to the global standby node. If the machine group is not changed in the configuration information, the machine group configuration item can be determined to be changed, and the global master node does not need to regroup the machine group again, so that a third updating instruction without carrying the sub-group identification is sent to the global standby node.

In some embodiments, after receiving the third update instruction sent by the global master node, the global standby node updates the locally stored machine group configuration item by using the machine group configuration item in the configuration information, and obtains the updated machine group configuration item. When the global standby node is the local main node in the affiliated node group, the updated machine group configuration item can be sent to each machine in the machine subgroup associated with the affiliated node group.

In this embodiment, the global standby node is notified to change along with the change of the configuration information by updating the assignment, so that the correctness of the machine group configuration item and the machine subgroup recorded by the global standby node can be ensured.

In some embodiments, sending the load information of each machine in the target machine subset to the routing device includes sending, to the routing device, a routing heartbeat packet carrying the load information of each machine in the target machine subset and the target machine subset, in the event that the target machine subset has been updated, so that the routing device updates the target machine subset, and obtaining the load information of each machine in the target machine subset.

Specifically, for each node group, a local master node in the node group is responsible for sending load information and machine subgroups to the routing device, so that the problem of inconsistent data can be avoided.

In this embodiment, under the condition that the target machine subset is updated, the routing device may update the locally stored machine subset associated with the target node set by sending the routing heartbeat packet, and may collect the load information at the same time, thereby improving the processing efficiency.

In some embodiments, the method further comprises the steps of inquiring time information associated with the inquiry words from the storage device by adopting preset inquiry words, determining that the target node is a global master node when the time information associated with the inquiry words is not inquired, and storing the time information of the current time point and the inquiry words in the storage device in an associated mode.

Wherein, the query words are fixed words and can be preset. The storage device has a function of storing data, and the storage device may store data in any form, for example, may store data in the form of key-value pairs, for example, the storage device may realize a function of storing data based on etcd (distributed key-value storage system). The target node may query the corresponding key from the storage device for instant information with the query term as a key. The association storage means that the time information of the current time point and the query word are stored in the storage device, and the time information of the current time point and the query word are associated in the storage device. For example, the target node may use the query word as a key, use the time information of the current time point as a value, generate a key value pair, and store the generated key value pair in the storage device.

Specifically, when the time information associated with the query word does not exist in the storage device, it is stated that no node is yet the global master node, and the target node can make the target node the global master node by writing the time information associated with the query word into the storage device.

In some embodiments, the target node may determine whether the target node is a local master node in the group of nodes to which the target node belongs using the same method. The difference is that the query terms employed are different. For example, it is determined whether the query term that is the global master node is the first query term. And the target node adopts the preset first query word to query time information associated with the first query word from the storage equipment, and when the time information associated with the first query word is not queried, the target node is determined to be a global master node. Each node in the node cluster can adopt the embodiment to determine whether the node is a global master node.

In some embodiments, each node group corresponds to a second query term and a different node group corresponds to a different second query term. Each second query term is different from the first query term. The second query term may be generated according to information of nodes in the node group, for example, may be generated according to identifiers of nodes in the node group, for example, the identifiers of the nodes in the node group may be combined to obtain the second query term. For example, if node group 1 is (node a, node B), and node group 2 is (node B, node C), then the second query term corresponding to node group 1 may be "node a_node B", and the second query term corresponding to node group 2 may be "node b_node C". For any node group to which the target node belongs, the target node can adopt a second query word corresponding to the node group to which the target node belongs, time information associated with the second query word is stored in the storage device, and when the time information associated with the second query word is not queried, the target node is determined to be a local master node in the node group to which the target node belongs.

In the embodiment, whether the target node is the global master node is judged by adopting a mode of inquiring the time information related to the inquiry words, so that the efficiency of determining the global master node is improved.

In some embodiments, the method further comprises determining a time interval between a time point represented by the queried time information and a current time point when the time information associated with the query word is queried, determining that the target node is a global standby node when the time interval does not reach a time interval threshold, determining that the target node is a global main node when the time interval reaches a time interval threshold, and updating the time information associated with the query word stored in the storage device to the time information of the current time point.

The time interval threshold may be set as required, for example, 5 seconds or 3 seconds.

Specifically, after the target node becomes the global master node, the "query time information associated with the query word from the storage device by using the preset query word" may be executed according to a preset time interval, where the preset time interval is smaller than or equal to the time interval threshold, so that the time interval between the time point represented by the queried time information and the current time point does not reach the time interval threshold, and the target node is continuously used as the global master node.

In some embodiments, when the target node is not the global master node, the "use preset query word" may be executed according to a preset time interval, and the time information associated with the query word is queried from the storage device, where the time interval does not reach the time interval threshold, which indicates that the existing node becomes the global master node, and where the time interval has reached the time interval threshold, which indicates that the global master node has not timely updated the time information corresponding to the query word in the storage device due to its own cause, so that it may indicate that a problem occurs in the global master node, and a new global master node is required, so that the target node may update the time information associated with the query word in the storage device to the time information of the current time point, thereby becoming the global master node, and then update the time information associated with the query word at a preset time interval to continue to serve as the global master node.

It should be noted that, each node in the node cluster may determine whether the node is a global master node by using the method of determining whether the node is a global master node provided in this embodiment. As shown in fig. 5, a flowchart of determining whether any node is a global master node is shown, and each node in the node cluster may execute the steps in fig. 5 according to a preset time interval to implement switching between the global master node and the global standby node.

In some embodiments, for each node in the node cluster, a method similar to that of fig. 5 may be used to determine itself as a local master node or a local backup node in the belonging node group. Specifically, the node adopts a second query word corresponding to the affiliated node group, queries time information associated with the second query word from the storage device, determines itself as a local master node when the time information associated with the second query word is not queried, and stores the time information of the current time point and the second query word in the storage device in an associated manner. When the time information related to the second query word is queried, determining the time interval between the time point represented by the queried time information and the current time point, determining that the time interval does not reach a time interval threshold, determining that the time interval is a local standby node, determining that the time interval reaches the time interval threshold, determining that the time interval is a local main node, and updating the time information related to the second query word stored in the storage device into the time information of the current time point.

In this embodiment, if the time interval does not reach the time interval threshold, it indicates that the global master node timely updates the time information associated with the query word or that the global standby node becomes the global master node by updating the time information, thereby determining that the global master node is the global standby node, if the time interval reaches the time interval threshold, it indicates that the global master node does not timely update the time information, and indicates that the global master node has a problem, thereby making the original global master node become the local master node by updating the time information, and improving the switching efficiency and the switching accuracy of the global master node and the local master node.

In some embodiments, the cluster management method further comprises the steps of grouping the node clusters to obtain a plurality of node groups when the target node is a global master node, and performing the step of grouping the node clusters to obtain the plurality of node groups again to update the plurality of node groups when the node clusters are determined to have changed.

Specifically, grouping the node clusters is performed by a global master node in the node clusters. If the nodes in the node cluster are unchanged, the global master node is replaced, and the node cluster is not triggered to be regrouped. For example, in the current period, node 1 in the node cluster is a global master node, and in the next period, node 1 in the node cluster is a global standby node, and node 2 is a global master node, so that the node cluster is not triggered to be regrouped.

In some embodiments, the global master node performs anomaly detection on each node in the node cluster, and when it is determined that an anomaly node exists in the node cluster, the anomaly node is removed from the node cluster, so that modification of the node cluster is achieved. After the change, the global master node groups the node clusters again to obtain a plurality of node groups, and each node executes the steps in each embodiment of the application according to the plurality of node groups which are grouped again.

In this embodiment, when it is determined that the node cluster has been changed, the step of grouping the node cluster to obtain a plurality of node groups is performed again, so as to update the plurality of node groups, thereby ensuring the correctness of the node groups.

In some embodiments, the method further comprises sharing grouping information for indicating a plurality of node groups to each global standby node, wherein the grouping information is used for indicating the global standby node, determining the node group from the plurality of node groups, and sending load information of each machine in a machine subgroup associated with the node group to the routing equipment when the global standby node is a local master node in the node group.

Specifically, the global master node may generate packet information indicating a plurality of node groups and send the packet information to each global standby node in the node cluster. For example, the global master node may broadcast the packet information to each node in the node cluster. For each node in the node cluster, the node may determine the group of nodes to which it belongs from the groups of nodes indicated by the grouping information.

In some embodiments, the cluster of nodes is used to control a set of machines for performing different tasks in a backend system, which may be a backend system of any platform. The cluster management method provided by the application can be applied to load balancing of any machine group in a background system. As shown in fig. 6, a schematic diagram of load balancing a machine group is provided, where a machine group 1 and a machine group 2 exist in a background system, a node cluster includes a node a, a node B, and a node C, and there are 3 node groups, the node group 1 includes the node a and the node B, the node group 2 includes the node B and the node C, and the node group 3 includes the node C and the node a. The node cluster can manage the machine group 1 and the machine group 2 at the same time, taking the machine group 2 as an example, the machine group 2 comprises n machines, the machine group 2 is uniformly divided into 3 parts to obtain 3 machine subgroups, the machines 1 to n/3 are 1 st machine subgroup, the machines n/3+1 to 2n/3 are 2 nd machine subgroup, the machines 2n/3+1 to n are 3 rd machine subgroup, the node group 1 is associated with the 1 st machine subgroup, the node group 2 is associated with the 2 nd machine subgroup, and the node group 3 is associated with the 3 rd machine subgroup. Firstly, each node in a node cluster performs 'node selection' or determines whether the node is a global master node, wherein the node B is assumed to be the global master node, the node A and the node C are all global standby nodes, after the operation equipment updates configuration information in a database, namely '1. Configuration landing', the operation equipment informs the global master node, namely the node B to update according to the configuration information, namely '2. Notification updating', and then the node B informs the node A and the node C to update according to the configuration information, namely '3. Notification updating', so that the node A, the node B and the node C can determine the latest configuration information from the database, and update, for example update a machine subset or update a machine set configuration item, by adopting the latest configuration information. For each node group, each node in the node group acquires load information of each machine in the associated machine subgroup, namely 'heartbeat load', by sending a heartbeat packet, and sends an updated machine group configuration item, namely 'configuration update', to the associated machine subgroup by sending the heartbeat packet.

As shown in fig. 7, a flowchart of a cluster management method is provided, in fig. 7, a "node selection" is performed first, that is, whether a node is a global master node is determined, after the global master node has been determined, the "global master node groups the nodes, that is, the nodes are selected in a group," that is, the global master node groups obtain a plurality of node groups, the nodes in the node groups determine whether the nodes are local master nodes in the node groups, then, "operating equipment updates configuration information," and notifies the global master node of updating configuration, "and" the global master node notifies each group of updating configuration, "that is, the operating equipment updates configuration information in a database, and updates according to the latest configuration information through the global master node, and the global master node further notifies the global standby node of updating, for example, updating a machine subgroup or a machine group configuration item according to the latest configuration information. "whether or not it is an intra-packet master" refers to the node determining whether or not it is a local master in the node group. If yes, the latest machine group configuration item is sent to the machine in a heartbeat packet sending mode by 'configuration issuing, heartbeat and load acquisition', load information is obtained, and then the latest machine group configuration item is sent to the routing equipment in a heartbeat packet sending mode by 'heartbeat routing equipment and load information is issued', and the latest machine group can be sent to the routing equipment. If not, the heartbeat and load information acquisition is to acquire the load information by adopting a heartbeat packet sending mode.

In this embodiment, the grouping information for indicating the plurality of node groups is shared with each global standby node, so that the global standby node can quickly determine the node group to which the global standby node belongs from the plurality of node groups.

In some embodiments, as shown in fig. 8, a cluster management method is provided, which may be performed by any node in the node cluster in fig. 1, and the method is applied to a target node in the node cluster, which is illustrated by way of example, and includes the following steps 802 to 828. Wherein:

step 802, determining whether the target node is a global master node, if so, executing step 804, and if not, executing step 810.

The target node is any node in the node cluster. The node cluster comprises a global main node and at least two global standby nodes, the global main node groups the node cluster into a plurality of node groups, one node in the node groups is a local main node in the node groups, and the rest nodes are local standby nodes for standby of the local main node.

Step 804, grouping the node clusters to obtain a plurality of node groups, and determining a target node group to which the target node belongs from the plurality of node groups.

Step 806, generating a plurality of subgroup identifications according to the machine group, establishing a one-to-one correspondence between each node group and each subgroup identification, obtaining a machine subgroup indicated by the subgroup identification corresponding to the target node group from the machine group, obtaining a target machine subgroup associated with the target node group, and sending, for each global standby node, a subgroup identification corresponding to the node group to which the global standby node belongs to the global standby node.

Step 808, it is determined whether the machine set has been changed, if so, the process returns to step 806, and if not, the process proceeds to step 816.

Step 810, determining a target node group to which the target node belongs from a plurality of node groups obtained by grouping global master nodes.

Step 812, receiving the subgroup identifier corresponding to the target node group sent by the global master node, and obtaining the machine subgroup indicated by the subgroup identifier corresponding to the target node group from the machine group, thereby obtaining the target machine subgroup associated with the target node group.

Wherein each node group is associated with a different subset of machines in the same machine group.

Step 814, it is determined whether the global master node notifies the machine group that it has been updated, if so, it returns to step 812, if not, it proceeds to step 816.

Step 816, it is determined whether the machine set configuration item has been changed, if so, step 818 is executed, and if not, step 820 is executed.

Step 818, update the locally stored machine group configuration item.

Step 820, it is determined whether the node cluster has been changed, if not, step 822 is entered, and if so, step 802 is returned.

Step 822, determining whether the target node is a local master node in the target node group, if not, executing step 824, and if so, proceeding to step 826.

Step 824, send a machine heartbeat packet to each machine in the target machine subset, where the machine heartbeat packet does not carry the updated machine group configuration item, and receive load information returned by each machine in the target machine subset in response to the machine heartbeat packet.

Step 826, sending a machine heartbeat packet carrying the updated machine group configuration item to each machine in the target machine subset if the machine group configuration item is changed, and sending a machine heartbeat packet not carrying the machine group configuration item to each machine in the target machine subset if the machine group configuration item is not changed.

Step 828, receiving load information returned by each machine in the target machine subset in response to the machine heartbeat packet, and sending the route heartbeat packet carrying the load information of each machine in the target machine subset and the target machine subset to the route device.

In this embodiment, since the node cluster includes a global master node and at least two global backup nodes, the master node and the backup nodes are globally distinguished, and the global master node groups the node cluster into multiple node groups, and one node in the node groups is a local master node in the node groups, and the remaining nodes are local backup nodes for backup of the local master node, so that multi-stage master-backup division is implemented in local distinction of the master node and the backup nodes, in the case of multi-stage master-backup division, each node group is associated with a different machine subgroup in the same machine group, a target node group to which the target node belongs is determined from the multiple node groups, a target machine subgroup associated with the target node group is determined, load information of each machine in the target machine subgroup is obtained, when the target node is a local master node in the target node group, load information of each machine in the target machine subgroup is sent to the routing device, so as to instruct the routing device to perform load balancing on the machines in the target machine subgroup, thereby supporting a mode of increasing the machine flow and increasing the node in the whole, and the problem of local backup node and the backup node can be solved when the local node group is enabled.

The cluster management method provided by the application can be applied to application scenes such as online conferences, social contact or instant messaging, and the stability of respective systems of an online conference platform, a social contact platform or an instant messaging platform can be improved.

In an online conference scene, a node cluster, a machine and a routing device belong to a background system of an online conference platform, and a terminal can send a machine connection request to the background system of the online conference platform through a webpage or a client provided by the online conference platform. For any node in the node cluster, the node responds to the node cluster grouping of the global master node to obtain a plurality of node groups, the node group to which the node belongs is determined from the plurality of node groups, a machine subgroup associated with the node group is determined from the machine groups, the load information of each machine in the machine subgroup is obtained, and when the node is a local master node in the node group to which the node belongs, the load information of each machine in the machine subgroup is sent to the routing equipment. The routing equipment responds to the machine connection request, selects a target machine from the machines according to the load information of the machines, returns the network address of the target machine to the terminal, and establishes communication connection with the target machine according to the network address, so that the terminal can send contents in the conference to the target machine through the established connection or acquire the contents generated in the conference from the target machine.

In a social scene, a node cluster, a machine and a routing device belong to a background system of a social platform, and a terminal can send a machine connection request to the background system of the social platform through a network or a client provided by the social platform. For any node in the node cluster, the node responds to the global master node to group the node cluster to obtain a plurality of node groups, the node group to which the node belongs is determined from the plurality of node groups, a machine subgroup associated with the node group is determined from the machine groups, the load information of each machine in the machine subgroup is obtained, and when the node is a local master node in the node group to which the node belongs, the load information of each machine in the machine subgroup is sent to the routing equipment. The routing equipment responds to the machine connection request, selects a target machine from the machines according to the load information of the machines, returns the network address of the target machine to the terminal, and establishes communication connection with the target machine according to the network address, so that a session can be established through the established connection, session contents can be sent to the target machine, or contents generated in the session can be acquired from the target machine. Or the social platform can also pass through a content sharing function, and the terminal can respond to the content sharing operation triggered in the social platform and send the media content indicated to be shared by the content sharing operation to the target machine through the established connection.

It should be understood that, although the steps in the flowcharts related to the embodiments described above are sequentially shown as indicated by arrows, these steps are not necessarily sequentially performed in the order indicated by the arrows. The steps are not strictly limited to the order of execution unless explicitly recited herein, and the steps may be executed in other orders. Moreover, at least some of the steps in the flowcharts described in the above embodiments may include a plurality of steps or a plurality of stages, which are not necessarily performed at the same time, but may be performed at different times, and the order of the steps or stages is not necessarily performed sequentially, but may be performed alternately or alternately with at least some of the other steps or stages.

Based on the same inventive concept, the embodiment of the application also provides a cluster management device for realizing the above related cluster management method. The implementation of the solution provided by the device is similar to the implementation described in the above method, so the specific limitation in one or more embodiments of the cluster management device provided below may refer to the limitation of the cluster management method hereinabove, and will not be repeated herein.

In some embodiments, as shown in FIG. 9, a cluster management apparatus is provided, comprising a node group determination module 902, a machine subset determination module 904, a load information acquisition module 906, and a load information transmission module 908, wherein:

A node group determining module 902, configured to obtain a plurality of node groups in response to grouping the node clusters by the global master node, and determine, from the plurality of node groups, a target node group to which the target node belongs; each node group comprises at least one global standby node; for each node group, one node in the node group is a local main node, and the rest nodes are local standby nodes serving as local main nodes;

A machine subset determining module 904 configured to determine a target machine subset associated with a target node group from among the machine groups, each node group being associated with a different machine subset of the machine groups;

a load information obtaining module 906, configured to obtain load information of each machine in the target machine subset;

The load information sending module 908 is configured to send, to the routing device, load information of each machine in the target machine subset when the target node is a local master node in the target node group, where each load information is used for load balancing of the machines in the machine group by the routing device.

In some embodiments, the apparatus further includes an update processing module, configured to, when the target node is a global master node, receive a first update instruction sent by the operation device and indicating that configuration information including the machine group has been updated, compare, in response to the first update instruction, a locally stored version number of the machine group with a version number of the machine group in the configuration information, determine that the updated machine group is inconsistent, and, when the comparison is inconsistent, execute the step of determining, from the machine group, a target machine subset associated with the target node group again, so as to update the target machine subset.

In some embodiments, the configuration information further includes a machine group configuration item associated with each machine in the machine group, the update processing module is further configured to compare the locally stored machine group configuration item with the machine group configuration item in the configuration information, obtain the machine group configuration item in the configuration information when the comparison is inconsistent, update the locally stored machine group configuration item with the machine group configuration item in the configuration information to obtain an updated machine group configuration item, and update the updated machine group configuration item for configuring and updating each machine in the target machine group.

In some embodiments, the load information obtaining module 906 is further configured to send a machine heartbeat packet carrying updated machine group configuration information to each machine in the target machine subset when the target node is a local master node in the target node group, the machine heartbeat packet is configured to instruct each machine to return respective load information and perform configuration update with the updated machine group configuration information, and receive load information returned by each machine in the target machine subset in response to the machine heartbeat packet.

In some embodiments, the machine subgroup determining module 904 is further configured to obtain a subgroup identifier corresponding to the target node group determined by the global master node, where the subgroup identifier indicates a machine subgroup in the machine group, and obtain, from the machine group, a machine subgroup indicated by the subgroup identifier corresponding to the target node group, and obtain a target machine subgroup associated with the target node group.

In some embodiments, the device further comprises an identification processing module, wherein the identification processing module is used for generating a plurality of subgroup identifications based on the machine group when the target node is a global main node, establishing a one-to-one correspondence between each subgroup identification and each node group, determining a matched node group to which the global standby node belongs for each global standby node, sending a subgroup identification corresponding to the matched node group to the global standby node to indicate the global standby node, and acquiring a machine subgroup indicated by the subgroup identification corresponding to the matched node group from the machine group to obtain a machine subgroup associated with the matched node group.

In some embodiments, the identification processing module is further configured to, when it is determined that the machine group has been updated, send a second update instruction carrying a subgroup identification corresponding to the matched node group to the global standby node, the update processing module is further configured to, when it is determined that the machine group has not been updated, send a third update instruction carrying no subgroup identification to the global standby node, when the configuration information further includes machine group configuration items related to each machine in the machine group, and instruct the global standby node to update the locally stored machine group configuration items using the machine group configuration items in the configuration information.

In some embodiments, the load information sending module 908 is further configured to send, to the routing device, a route heartbeat packet carrying the load information of each machine in the target machine subset and the target machine subset, in a case where the target machine subset has been updated, so that the routing device updates the target machine subset, and obtains the load information of each machine in the target machine subset.

In some embodiments, the apparatus further includes a master node determining module, where the master node determining module is configured to query, from the storage device, time information associated with the query word by using a preset query word, determine that the target node is a global master node when the time information associated with the query word is not queried, and store the time information of the current time point and the query word in association with the storage device.

In some embodiments, the master node determining module is further configured to determine, when time information associated with the query word is queried, a time interval between a time point represented by the queried time information and a current time point, determine that the target node is a global standby node when the time interval does not reach a time interval threshold, determine that the target node is a global master node when the time interval reaches the time interval threshold, and update the time information associated with the query word stored in the storage device to the time information of the current time point.

In some embodiments, the cluster management device further includes a grouping module, where the grouping module is configured to group the node cluster to obtain a plurality of node groups when the target node is a global master node, and when it is determined that the node cluster has been changed, execute the step of grouping the node cluster again to obtain the plurality of node groups, so as to update the plurality of node groups.

In some embodiments, the node group determining module 902 is further configured to share, with each global standby node, grouping information for indicating a plurality of node groups, where the grouping information is used to indicate the global standby node, determine the node group to which the global standby node belongs from the plurality of node groups, and send, when the global standby node is a local master node in the node group to which the global standby node belongs, load information of each machine in a subset of machines associated with the node group to which the global standby node belongs to the routing device.

The modules in the cluster management device may be implemented in whole or in part by software, hardware, or a combination thereof. The above modules may be embedded in hardware or may be independent of a processor in the computer device, or may be stored in software in a memory in the computer device, so that the processor may call and execute operations corresponding to the above modules.

In some embodiments, a computer device is provided, which may be a server, the internal structure of which may be as shown in fig. 10. The computer device includes a processor, a memory, an Input/Output interface (I/O) and a communication interface. The processor, the memory and the input/output interface are connected through a system bus, and the communication interface is connected to the system bus through the input/output interface. Wherein the processor of the computer device is configured to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, computer programs, and a database. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage media. The database of the computer device is used for storing data involved in the cluster management method. The input/output interface of the computer device is used to exchange information between the processor and the external device. The communication interface of the computer device is used for communicating with an external terminal through a network connection. The computer program is executed by a processor to implement a cluster management method.

In some embodiments, a computer device is provided, which may be a terminal, and the internal structure of which may be as shown in fig. 11. The computer device includes a processor, a memory, an input/output interface, a communication interface, a display unit, and an input means. The processor, the memory and the input/output interface are connected through a system bus, and the communication interface, the display unit and the input device are connected to the system bus through the input/output interface. Wherein the processor of the computer device is configured to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage media. The input/output interface of the computer device is used to exchange information between the processor and the external device. The communication interface of the computer device is used for carrying out wired or wireless communication with an external terminal, and the wireless mode can be realized through WIFI, a mobile cellular network, NFC (near field communication) or other technologies. The computer program is executed by a processor to implement a cluster management method. The display unit of the computer device is used for forming a visual picture, and can be a display screen, a projection device or a virtual reality imaging device. The display screen can be a liquid crystal display screen or an electronic ink display screen, and the input device of the computer equipment can be a touch layer covered on the display screen, can also be a key, a track ball or a touch pad arranged on the shell of the computer equipment, and can also be an external keyboard, a touch pad or a mouse and the like.

It will be appreciated by those skilled in the art that the structures shown in fig. 10 and 11 are merely block diagrams of portions of structures associated with aspects of the application and are not intended to limit the computer device to which aspects of the application may be applied, and that a particular computer device may include more or fewer components than those shown, or may combine certain components, or may have a different arrangement of components.

In some embodiments, a computer device is provided, comprising a memory having a computer program stored therein and a processor implementing the steps of the cluster management method described above when the computer program is executed.

In one embodiment, a computer readable storage medium is provided, on which a computer program is stored which, when executed by a processor, implements the steps of the cluster management method described above.

In an embodiment, a computer program product is provided comprising a computer program which, when executed by a processor, implements the steps of the cluster management method described above.

It should be noted that, the user information (including but not limited to user equipment information, user personal information, etc.) and the data (including but not limited to data for analysis, stored data, presented data, etc.) related to the present application are both information and data authorized by the user or sufficiently authorized by each party, and the collection, use and processing of the related data are required to meet the related regulations.

Those skilled in the art will appreciate that implementing all or part of the above described methods may be accomplished by way of a computer program stored on a non-transitory computer readable storage medium, which when executed, may comprise the steps of the embodiments of the methods described above. Any reference to memory, database, or other medium used in embodiments provided herein may include at least one of non-volatile and volatile memory. The nonvolatile Memory may include Read-Only Memory (ROM), magnetic tape, floppy disk, flash Memory, optical Memory, high density embedded nonvolatile Memory, resistive random access Memory (ReRAM), magneto-resistive random access Memory (Magnetoresistive Random Access Memory, MRAM), ferroelectric Memory (Ferroelectric Random Access Memory, FRAM), phase change Memory (PHASE CHANGE Memory, PCM), graphene Memory, and the like. Volatile memory can include random access memory (Random Access Memory, RAM) or external cache memory, and the like. By way of illustration, and not limitation, RAM can be in various forms such as static random access memory (Static Random Access Memory, SRAM) or dynamic random access memory (Dynamic Random Access Memory, DRAM), etc. The databases referred to in the embodiments provided herein may include at least one of a relational database and a non-relational database. The non-relational database may include, but is not limited to, a blockchain-based distributed database, and the like. The processor referred to in the embodiments provided in the present application may be a general-purpose processor, a central processing unit, a graphics processor, a digital signal processor, a programmable logic unit, a data processing logic unit based on quantum computing, or the like, but is not limited thereto.

The technical features of the above embodiments may be arbitrarily combined, and all possible combinations of the technical features in the above embodiments are not described for brevity of description, however, as long as there is no contradiction between the combinations of the technical features, they should be considered as the scope of the description.

The foregoing examples illustrate only a few embodiments of the application and are described in detail herein without thereby limiting the scope of the application. It should be noted that it will be apparent to those skilled in the art that several variations and modifications can be made without departing from the spirit of the application, which are all within the scope of the application. Accordingly, the scope of the application should be assessed as that of the appended claims.

Claims (16)

1. A method of cluster management, performed by a target node in a node cluster, the node cluster comprising a global master node and at least two global standby nodes, the target node being one of the global master node or the global standby nodes, the method comprising:

responding to the global master node to group the node clusters to obtain a plurality of node groups, and determining a target node group to which the target node belongs from the plurality of node groups, wherein each node group comprises at least one global standby node;

determining a target machine subgroup associated with the target node group from machine groups, wherein each node group is associated with a different machine subgroup in the machine group;

Acquiring load information of each machine in the target machine subset;

And when the target node is a local master node in the target node group, sending load information of each machine in the target machine subgroup to the routing equipment, wherein each load information is used for carrying out load balancing on the machines in the machine group by the routing equipment.

2. The method according to claim 1, wherein the method further comprises:

When the target node is a global master node, a first updating instruction which is sent by operation equipment and indicates that the configuration information containing the machine group is updated is received;

Responding to the first updating instruction, and comparing the version number of the locally stored machine group with the version number of the machine group in the configuration information;

And when the comparison is inconsistent, determining that the machine group is updated, and executing the step of determining the target machine subgroup associated with the target node group from the machine group again to update the target machine subgroup.

3. The method of claim 2, wherein the configuration information further comprises a machine group configuration item associated with each machine in the machine group, the method further comprising:

Comparing the locally stored machine group configuration items with the machine group configuration items in the configuration information;

When the comparison is inconsistent, acquiring a machine group configuration item in the configuration information;

And updating the locally stored machine set configuration items by adopting the machine set configuration items in the configuration information to obtain updated machine set configuration items, wherein the updated machine set configuration items are used for carrying out configuration updating on all machines in the target machine subset.

4. A method according to claim 3, wherein said obtaining load information for each machine in said target subset of machines comprises:

When the target node is a local master node in the target node group, sending a machine heartbeat packet carrying the updated machine group configuration information to each machine in the target machine subgroup, wherein the machine heartbeat packet is used for indicating each machine to return the respective load information and carrying out configuration update by adopting the updated machine group configuration information;

And receiving load information returned by each machine in the target machine subset in response to the machine heartbeat packet.

5. The method of claim 2, wherein the determining the target subset of machines associated with the target node group from the set of machines comprises:

the method comprises the steps of acquiring a subgroup identifier corresponding to a target node group determined by a global master node, wherein the subgroup identifier indicates a machine subgroup in the machine group;

And acquiring a machine subgroup indicated by the subgroup identification corresponding to the target node group from the machine group to obtain a target machine subgroup associated with the target node group.

6. The method of claim 5, wherein the method further comprises:

when the target node is a global master node, generating a plurality of subgroup identifications based on the machine group;

Establishing a one-to-one correspondence between each subgroup identifier and each node group;

determining a matching node group to which the global standby node belongs for each global standby node;

And sending a subgroup identifier corresponding to the matched node group to the global standby node to indicate the global standby node, and acquiring a machine subgroup indicated by the subgroup identifier corresponding to the matched node group from the machine group to obtain a machine subgroup associated with the matched node group.

7. The method of claim 6, wherein the sending the subgroup identifier corresponding to the matching node group to the global standby node comprises:

when the machine group is determined to be updated, a second updating instruction carrying a subgroup identifier corresponding to the matched node group is sent to the global standby node;

The method further comprises the steps of:

And when the machine group is not updated, the configuration information also comprises machine group configuration items related to all machines in the machine group, a third updating instruction without carrying the sub-group identification is sent to the global standby node, and the third updating instruction is used for indicating the global standby node to update the locally stored machine group configuration items by adopting the machine group configuration items in the configuration information.

8. The method of claim 1, wherein said sending load information for each machine in the subset of target machines to the routing device comprises:

And under the condition that the target machine subset is updated, sending a routing heartbeat packet carrying the load information of each machine in the target machine subset and the target machine subset to the routing equipment so that the routing equipment updates the target machine subset and acquires the load information of each machine in the target machine subset.

9. The method according to claim 1, wherein the method further comprises:

inquiring time information associated with a preset inquiry word from a storage device by adopting the preset inquiry word;

And when the time information associated with the query word is not queried, determining that the target node is a global master node, and storing the time information of the current time point and the query word in association with the storage device.

10. The method according to claim 9, wherein the method further comprises:

When the time information related to the query word is queried, determining the time interval between the time point represented by the queried time information and the current time point;

When the time interval does not reach the time interval threshold, determining that the target node is a global standby node;

And when the time interval reaches the time interval threshold, determining that the target node is a global master node, and updating the time information associated with the query word stored in the storage device into the time information of the current time point.

11. The method according to any one of claims 1 to 10, further comprising:

When the target node is a global master node, grouping the node clusters to obtain a plurality of node groups;

And when the node cluster is determined to be changed, the step of grouping the node cluster to obtain a plurality of node groups is executed again so as to update the plurality of node groups.

12. The method of claim 11, wherein the method further comprises:

sharing grouping information for indicating the plurality of node groups to each global standby node;

The grouping information is used for indicating the global standby node, determining the affiliated node group from the plurality of node groups, and sending the load information of each machine in the machine subgroup associated with the affiliated node group to the routing equipment when the global standby node is a local main node in the affiliated node group.

13. A cluster management device, the device comprising:

The node group determining module is used for responding to the global master node to group the node clusters to obtain a plurality of node groups, and determining a target node group to which a target node belongs from the plurality of node groups; each node group comprises at least one global standby node; for each node group, one node in the node group is a local master node, and the rest nodes are local standby nodes serving as standby local master nodes; the node cluster comprises a global main node and at least two global standby nodes, and the target node is one of the global main node or the global standby nodes;

The machine subgroup determining module is used for determining a target machine subgroup associated with the target node group from machine groups, wherein each node group is associated with a different machine subgroup in the machine groups;

The load information acquisition module is used for acquiring the load information of each machine in the target machine subset;

and the load information sending module is used for sending the load information of each machine in the target machine subset to the routing equipment when the target node is a local master node in the target node group, wherein each load information is used for carrying out load balancing on the machines in the machine set by the routing equipment.

14. A computer device comprising a memory and a processor, the memory storing a computer program, characterized in that the processor implements the steps of the method of any one of claims 1 to 12 when the computer program is executed.

15. A computer readable storage medium, on which a computer program is stored, characterized in that the computer program, when being executed by a processor, implements the steps of the method of any of claims 1 to 12.

16. A computer program product comprising a computer program, characterized in that the computer program, when being executed by a processor, implements the steps of the method of any one of claims 1 to 12.