Communication system

Abstract

This application provides a communication system applied in a cloud cluster switch. The cloud cluster switch includes a first node, which includes an admin component and a role component. The admin component is used to obtain the cluster configuration of the cloud cluster switch; determine the role of the first node in the cloud cluster switch from the cluster configuration; and start the corresponding role component according to the role of the first node in the cloud cluster switch. The role component is used to receive a container description file; and create and start a container locally according to the container description file. The container description file is sent by an NCC node with the role of manager and the identity of leader, or by the admin component included in the NCC node with the role of manager and the identity of leader.

Description

Technical Field

[0001] This application relates to the field of communication technology, and more particularly to a communication system. Background Technology

[0002] A Distributed Data Center (DDC) is a facility that centrally stores, processes, and distributes data. A DDC typically includes servers, storage devices, network equipment, etc., to provide various computing and storage services. Data Center Interconnect (DCI) networking refers to the network infrastructure connecting multiple data centers. DCI networking is used to achieve high-speed, reliable, and secure communication between data centers to meet the high bandwidth and low latency requirements of data and applications. DCI networking includes network transmission technologies, fiber optic connections, routers, and switches.

[0003] With the continuous development of big data, artificial intelligence (AI), and cloud services, the service traffic carried by DCI networks is growing rapidly, increasing at a rate of 50% to 100% in recent years. This puts enormous pressure on network equipment in DCI networks. In particular, the development and application of generative artificial intelligence (AIGC) technology, characterized by high computational demands, presents several challenges to DDC and DCI networks. These include increased demands for computing resources, high-speed data transmission, and storage capacity.

[0004] Currently, high computing resources can be addressed and managed using dedicated chips such as GPUs. To adapt to the development of AIGC and meet the aforementioned challenges, DDC and DCI networks need continuous innovation and upgrades to provide more powerful, reliable, and efficient computing, storage, and communication capabilities. Within existing DDC and DCI networks, traditional chassis-based switches are also unable to meet the aforementioned challenges of AIGC. Summary of the Invention

[0005] In view of this, this application provides a communication system to solve the problems that the switches in existing DDC and DCI networks cannot adapt to the development of AIGC, nor can they meet the requirements of high-speed data transmission and increased storage capacity.

[0006] In a first aspect, this application provides a communication system applied in a cloud cluster switch, the cloud cluster switch including a first node, the first node including an admin component and a role component;

[0007] The admin component is used to obtain the cluster configuration of the cloud cluster switch; determine the role of the first node in the cloud cluster switch from the cluster configuration; and start the corresponding role component according to the role of the first node in the cloud cluster switch.

[0008] The role component is used to receive a container description file; create and start a container locally according to the container description file; the container is used to implement the business processing of the cloud cluster switch.

[0009] The container description file is sent by an NCC node with the role of manager and the identity of leader, or the container description file is sent by the admin component included in the NCC node with the role of manager and the identity of leader.

[0010] In a second aspect, this application provides a network device including a processor and a machine-readable storage medium storing machine-executable instructions that can be executed by the processor, which in turn cause the processor to execute the first node provided in the first aspect of this application.

[0011] Therefore, the communication system provided in this application is applied in a cloud cluster switch. The cloud cluster switch includes a first node, which includes an admin component and a role component. The admin component is used to obtain the cluster configuration of the cloud cluster switch; determine the role of the first node in the cloud cluster switch from the cluster configuration; and start the corresponding role component according to the role of the first node in the cloud cluster switch. The role component is used to receive a container description file; and create and start a container locally according to the container description file. The container description file is sent by an NCC node with the role of manager and the identity of leader, or the container description file is sent by the admin component included in the NCC node with the role of manager and the identity of leader.

[0012] Therefore, this application proposes a communication system implemented using a cloud cluster switch, where various types of nodes are aggregated into a single cloud cluster switch. Based on container description files, different types of nodes create containers within themselves, and these containers handle the business processing of the cloud cluster switch. Thus, through container technology, the cloud cluster switch can be dynamically managed, satisfying the switch's scalability and robustness requirements, and also adapting well to the current development of AIGC (AI Generic Communication). The aforementioned cloud cluster switch can be deployed within existing DDC (Distributed Data Center) and DCI (Distributed Communication Interface) networks, solving the problems of traditional switches being unable to adapt to the development of AIGC, meet the demands of high-speed data transmission, and address the increasing storage capacity requirements. Attached Figure Description

[0013] Figure 1 This is a network diagram of the communication system provided in an embodiment of this application;

[0014] Figure 2 A schematic diagram of the internal components of a node provided in an embodiment of this application;

[0015] Figure 3 The network device hardware structure provided in the embodiments of this application. Detailed Implementation

[0016] Exemplary embodiments will now be described in detail, examples of which are illustrated in the accompanying drawings. When the following description relates to the drawings, unless otherwise indicated, the same numerals in different drawings denote the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with this application. Rather, they are merely examples of apparatuses and methods consistent with some aspects of this application as detailed in the appended claims.

[0017] The terminology used in this application is for the purpose of describing particular embodiments only and is not intended to be limiting of the application. The singular forms “a,” “the,” and “the” used in this application and the appended claims are also intended to include the plural forms unless the context clearly indicates otherwise. It should also be understood that the term “and / or” as used herein refers to and includes any or all possible combinations of one or more of the corresponding listed items.

[0018] It should be understood that although the terms first, second, third, etc., may be used in this application to describe various information, such information should not be limited to these terms. These terms are only used to distinguish information of the same type from one another. For example, without departing from the scope of this application, first information may also be referred to as second information, and similarly, second information may also be referred to as first information. Depending on the context, the word "if" as used herein may be interpreted as "when," "when," or "in response to determination."

[0019] The communication system provided in the embodiments of this application will be described in detail below. See also Figure 1 , Figure 1 This is a network topology diagram of a communication system provided in an embodiment of this application. The communication system is applied to a cloud cluster switch.

[0020] A cloud cluster switch comprises multiple nodes, such as Network Cloud Controller (NCC) nodes, Network Cloud Fabric (NCF) nodes, and Network Cloud Packet-Forwarder (NCP) nodes, with multiple of each node. These nodes can be independent physical devices, such as switches or white-box switches. Multiple nodes are physically connected via a Network Management Transport (MGT, a Layer 3 switch supporting common routing protocols, multicast protocols, etc.) network and aggregated into a cloud cluster switch (or cloud cluster). Externally, the cloud cluster switch appears as a single network device.

[0021] exist Figure 1 In the diagram, solid lines between NCC nodes, NCF nodes, NCP nodes, and MGT represent physical connections. The nodes establish a cluster management network through the MGT, which uses the LIPC protocol for communication. The LIPC protocol resides in the same location as the TCP / IP protocol stack, encapsulated through a socket layer, and provides a unified external interface to applications.

[0022] In the embodiments of this application, the LIPC protocol can provide the following functions for NCC nodes, NCF nodes, and NCP nodes, including but not limited to addressing functions transparent to the protocol stack; topology management functions; location-transparent unicast; location-opaque unicast; connection-based transport layer reliable unicast; multicast group management functions; location-transparent multicast; connection-based transport layer reliable multicast; connectionless unicast datagram transmission, etc.

[0023] The dashed line between NCF nodes and NCP nodes represents the cell forwarding data network, which forwards the service traffic sent and received by the NCP node. The dotted horizontal line between the NCP node and the server or switch represents the panel port service network, which connects to various servers or switches and routers in the DCI network.

[0024] In a cloud cluster switch, different types of nodes are divided into different roles. These roles specifically include manager and worker. Each role has different responsibilities: 1) Manager: Responsible for cloud cluster management. The manager role has two identities: leader and follower. All API processing in the cloud cluster switch is handled by the leader manager, and the cloud cluster status is also managed by the leader manager; 2) Worker: Responsible for local node management, reporting node resources to the leader manager, and receiving container description files from the leader manager to deploy containers locally.

[0025] Nodes with different roles include the following four types of components, such as Figure 2 As shown, Figure 2 This is a schematic diagram of the internal components of a node provided in an embodiment of this application. Figure 2 Within a node, there may be manager, worker, admin, and agent components. The manager component runs within the main management node of the cloud cluster, typically the NCC node; 2) the worker component runs within the main physical node, typically the NCF or NCP; 3) the admin component runs within each physical node, receiving and processing container description files; 4) the agent component runs inside the container, but may not run within third-party containers.

[0026] The functions of the above components are as follows:

[0027] The admin component is responsible for deploying cloud clusters and application scenarios; 1) Mode management: supports independent mode and cluster mode; 2) Cloud cluster configuration management: configuration refresh, cloud cluster management network construction, and starting and stopping the manager / worker services of the cloud cluster; 3) Application definition management: basic containers and extended business containers.

[0028] The manager component includes: 1) API service: providing interfaces for executing application definition files, cluster management and information query, container management and information query; 2) Cloud cluster management: responsible for cloud cluster HA function, establishing cloud clusters, and managing cloud cluster members; 3) Resource management: responsible for the unified management and allocation of node physical resources, and statistics on resource usage; 4) Service registration and discovery: providing application registration service function to facilitate access to the service by other applications; 5) Scheduler: responsible for application deployment and scheduling.

[0029] Worker components: 1) API service: provides container management and query interfaces to the outside world; 2) Local node management: responsible for collecting and reporting the physical resources of the local node; 3) Container lifecycle management: the implementation of container runtime, responsible for the lifecycle management of local containers; 4) Container monitoring: periodically monitors and reports container status changes.

[0030] Agent component: 1) Runs inside the container and reports changes in the business status inside the container; 2) Responds to command-line plugin messages sent by the manager with the leader identity, converts them into cluster configuration and writes them to the container configuration file, and notifies the admin component of cluster configuration changes; 3) Obtains node status changes and notifies the node of identity changes, node joining and leaving the cloud cluster; 4) Provides internal businesses with programming interfaces for querying cloud cluster and container information, as well as subscribing to and notifying cloud cluster and container events.

[0031] It should be noted that if no container is created within a node, then the node does not include the agent component.

[0032] The API service provided by the aforementioned manager component can specifically be to receive requests from the requester to query files, containers, content, etc., and then provide the matching content to the recipient after retrieving it locally based on the request.

[0033] Of course, the manager component can also provide alarm functions. For example, when there are a large number of failures in the node's internal containers or when the load rate is too high (above 85%), the manager component will display alarm information.

[0034] The following explanation uses a cloud cluster switch including the first node as an example. The first node includes the admin component and role components.

[0035] The admin component is used to obtain the cluster configuration of the cloud cluster switch; determine the role of the first node in the cloud cluster switch from the cluster configuration; and start the corresponding role component according to the role of the first node in the cloud cluster switch.

[0036] The role component is used to receive a container description file; create and start a container locally according to the container description file; the container is used to implement the business processing of the cloud cluster switch.

[0037] The container description file is sent by an NCC node with the role of manager and the identity of leader, or the container description file is sent by the admin component included in the NCC node with the role of manager and the identity of leader.

[0038] In this embodiment, regardless of the first node's role in the cloud cluster switch, the first node includes an admin component. The role component varies depending on the first node's role in the cloud cluster switch. For example, if the first node's role in the cloud cluster switch is manager, the role component is manager; if the first node's role in the cloud cluster switch is worker, the role component is worker.

[0039] Furthermore, each first node also includes a base container, which contains an agent component. Administrators pre-write the cluster configuration into the base container via the agent component. After the first node powers on, the admin component accesses the base container and retrieves the cluster configuration. Based on the cluster configuration, it determines its role within the cloud cluster switch. After determining its role, the admin component starts the corresponding role component. Once the role component starts, it waits to receive container description files. Upon receiving the container description files, it creates and starts containers locally (e.g., business containers). It should be noted that subsequent nodes create and start business containers locally.

[0040] The cluster configuration includes, but is not limited to, the following: member ID, member IP, interface bound to the physical cluster, member role, virtual IP of the physical cluster, and container cluster port. The member ID identifies the physical node; the member IP is used to exchange cluster management messages; the interface bound to the physical cluster specifies the physical interface for exchanging cluster management messages; the member role is manager or worker; the virtual IP of the physical cluster is used by other components to access the cluster, implemented through interface aliases or auxiliary IPs; and the container cluster port is used to build clusters between containers.

[0041] If the first node's role in the cloud cluster switch is manager and its identity is follower, or if the first node's role in the cloud cluster switch is worker, then the container description file is sent by the NCC node with the role of manager and its identity is leader; if the first node's role in the cloud cluster switch is manager and its identity is leader, then the container description file is sent by the admin component included in the NCC node with the role of manager and its identity is leader.

[0042] The nodes for different roles will be explained separately below.

[0043] Optionally, in this embodiment of the application, the first node is a first NCC node (e.g., Figure 1 When the first NCC node is the NCC1 node, the role component included in the first NCC node is the manager component.

[0044] Of course, the first NCC node also includes an admin component. Specifically, the admin component is used to start the manager component if the first NCC node's role in the cloud cluster switch is manager.

[0045] The admin component is also used to determine the identity of the manager; if the manager's identity is leader, then send the first container description file to the manager component;

[0046] The manager component is used to receive the first container description file; create and start a container locally according to the first container description file; allocate resources and schedule containers for the cloud cluster switch; and send the second container description file to the NCC node with the identity of follower or the NCF node or NCP node with the role of worker through the resource allocation result and the container scheduling result.

[0047] The NCF node, the NCP node, and the NCC node whose identity is a follower are all located in the cloud cluster switch.

[0048] In this embodiment of the application, after the admin component determines that the node role is manager, it can notify the program launcher (initd), which can start the manager component.

[0049] It should be noted that after receiving the first container description file, the manager component will also receive resource attributes sent by other nodes in the cloud cluster switch, such as resource attributes sent by worker nodes and follower nodes. Based on the acquired resource attributes (representing the resource attributes of the entire cloud cluster switch) and the first container description file, the manager component can perform resource allocation and container scheduling. For example, the number and services of containers deployed within worker nodes; the number and services of containers deployed within follower nodes; and the number and services of containers deployed within leader nodes. Based on the resource allocation and container scheduling results, the manager component generates and sends a second container description file to other nodes.

[0050] After the admin component starts the manager component, the manager's identity can be verified by accessing the manager component. This identity is also pre-configured by the administrator in the manager component.

[0051] The first container description file may include the business tasks that the cloud cluster switch needs to undertake, the business type, the required resources, the total number of containers, the number of containers corresponding to each business, etc.

[0052] Optionally, in this embodiment of the application, the admin component is further configured to, if the manager's identity is a follower, receive the second container description file sent by the admin component included in the second NCC node, and send the second container description file to the manager component;

[0053] The manager component is further configured to receive the second container description file, and create and start the container locally according to the second container description file;

[0054] In this context, the second NCC node plays the role of manager and is identified as leader in the cloud cluster switch.

[0055] In the above process, the first NCC node acts as the manager and is identified as a follower in the cloud cluster switch. At this time, the admin component included in the first NCC node will receive the second container description file sent by the leader.

[0056] Optionally, the manager component is further configured to send the second container description file to the third NCC node, so that the manager component included in the third NCC node creates and starts a container locally according to the second container description file;

[0057] The manager component is also used to receive a first container creation success message sent by the third NCC node, wherein the first container creation success message is sent by the manager component included in the third NCC node after it has created and started the container locally;

[0058] In this context, the third NCC node plays the role of manager and is identified as follower in the cloud cluster switch, and the content device of the first NCC node and the content device of the third NCC node form a master role and a slave role.

[0059] In this embodiment of the application, the number of NCC nodes within the cloud cluster switch can be multiple, for example, Figure 2 The NCC1 and NCC2 nodes are configured in the system. After an NCC node acting as a follower creates a container locally, it will send a container creation success message to the NCC node acting as the leader. When the NCC node acting as the leader also creates a container locally, the container role of the leader NCC node becomes the master, while the container role of the follower NCC node becomes the slave.

[0060] It should be noted that the container role and service configuration will be sent by the manager component of the NCC node acting as the leader to the agent component of the third NCC node via command-line plugin messages. The container role and service configuration of the containers in each node are determined by the manager component of the NCC node acting as the leader, based on the first container description file sent by the admin component of its own NCC node and the obtained resource attributes, after resource allocation and container scheduling.

[0061] Optionally, the manager component is further configured to send a first notification message via broadcast in the cloud cluster switch, the first notification message including the leader identity of the manager, so as to announce the leader identity of the first NCC node in the cloud cluster switch.

[0062] The admin component is also used to receive feedback messages sent by other nodes in the cloud cluster switch other than the first NCC node, the feedback messages including the roles or identities of the other nodes; start a cluster joining timer for each other node that sends a feedback message; within a preset time, determine whether the other nodes that have started the cluster joining timer have joined the cloud cluster switch; if so, delete the cluster joining timer.

[0063] In a cloud cluster switch, once an NCC node confirms its leader status, its manager component will broadcast this announcement throughout the switch. Other nodes in the switch (those that are powered on, have acquired cluster configuration, have determined their roles, and have started their role software) will then identify the leader NCC node upon receiving the notification and send a feedback message to that node, informing it of their own identity or role (follower or worker).

[0064] After receiving a feedback message, the admin component within the NCC node starts a cluster join timer for the other nodes that sent the feedback message. The timer duration is 1 minute. Within 1 minute, the admin component checks whether the other nodes with started cluster join timers have joined the cloud cluster switch. If other nodes join the cloud cluster switch within 1 minute, the admin component deletes the cluster join timer; if other nodes do not join the cloud cluster switch within 1 minute, the admin component waits for the timer duration to expire, then can re-establish the cluster join timer, and delete it after other nodes join the cloud cluster switch.

[0065] Understandably, if the admin component creates a cluster and adds a timer multiple times, but the other node still hasn't joined the cloud cluster switch, the admin component will delete the feedback messages sent by that other node.

[0066] In the process of determining whether other nodes that have started the cluster join timer have joined the cloud cluster switch, the admin component can do so by polling the status of other nodes. For example, it can probe whether other nodes already exist on the cloud cluster switch by sending keep-alive messages within the cloud cluster switch.

[0067] Optionally, the manager component is further configured to: send a second notification message to the second NCC node, the second notification message including a first resource attribute, so that the manager component included in the second NCC node generates the second container description file according to the first resource attribute; and send a second container creation success message to the second NCC node.

[0068] In a cloud cluster switch, the first NCC node, acting as a follower, after identifying the second NCC node as the leader, sends its first resource attribute along with a second notification message to the second NCC node (e.g., the manager component). The second NCC node can then generate a second container description file based on the first resource attribute and distribute it to the first NCC node. After the first NCC node creates the container locally based on the second container description file, it sends a container creation success message to the second NCC node again. Specifically, the first resource attribute can include the second NCC node's total hardware and software resources, current remaining hardware and software resources, current load rate, etc.

[0069] Optionally, when the first node is an NCF node or an NCP node, the role component is a worker component;

[0070] The admin component is specifically used to start the worker component if the NCF node or the NCP node plays the role of a worker in the cloud cluster switch.

[0071] The worker component is configured to send a third notification message to the first NCC node, the third notification message including a second resource attribute, so that the manager component included in the first NCC node generates the second container description file according to the second resource attribute; receive the second container description file sent by the first NCC node; create a container locally according to the second container description file; and send a third container creation success message to the first NCC node.

[0072] In this embodiment of the application, the first node is Figure 1When the first node is an NCF1 or NCP1 node, the role component included in the first node is the worker component.

[0073] Of course, the first node also includes the admin component. After the admin component determines that the node's role in the cloud cluster switch is worker, it can notify the program initiator (initd), which can then start the worker component.

[0074] As described above, after identifying the leader node as a first NCC node, the NCF or NCP node with the role of worker sends its second resource attribute along with a third notification message to the first NCC node (e.g., the manager component). The first NCC node can then generate a second container description file based on the second resource attribute and distribute it to the NCF or NCP node. After creating the container locally based on the second container description file, the NCF or NCP node sends a container creation success message to the first NCC node again.

[0075] The second resource attribute can specifically be the total hardware and software resources of the NCF node or NCP node, the current remaining hardware and software resources, the current load rate, etc.

[0076] It should be noted that during the process of generating the second container description file, the NCC node acting as the leader will allocate resources and schedule containers based on the resource attributes of all nodes in the entire cloud cluster switch and the first container description file issued by the admin component in the NCC node.

[0077] Optionally, the admin component is further configured to obtain the cluster mode of the cloud cluster switch and configure the cluster management network of the cloud cluster switch according to the cluster mode and the cluster configuration.

[0078] The cluster management network uses the LIPC protocol, and all nodes in the cloud cluster switch establish communication connections through the LIPC protocol.

[0079] In this embodiment, the cluster mode of the cloud cluster switch is also pre-written into the basic container by the administrator through the agent component included in the basic container. Regardless of the role of the first node in the cloud cluster switch, after the first node is powered on, its admin component accesses the basic container and obtains the cluster mode in addition to the cluster configuration. This cluster mode is specifically a cluster networking mode, such as the communication protocol used between nodes, the networking mode used between NCF nodes and NCP nodes, the networking mode used between NCP nodes and servers / switch, etc.

[0080] In cluster mode, each node configures its own cluster management network in the cloud cluster switch.

[0081] Optionally, when a container is created within the first node, the first node further includes an agent component, which is deployed inside the container;

[0082] The agent component is used to receive command-line plugin messages; convert the command-line plugin messages into cluster configuration and write them to the container configuration file; and send a fourth notification message to the NCC node with the role of manager and the identity of leader when the business state inside the container changes.

[0083] The command-line plugin message is sent by the NCC node whose role is manager and whose identity is leader.

[0084] In this embodiment, the aforementioned conversion refers to format conversion of the content included in the command-line plugin message, converting the message format (XML format) into a file format (JSON format). The command-line plugin message can be sent by the manager component of an NCC node with the role of manager and the identity of leader. The fourth notification message can be sent by the agent component to the manager component of an NCC node with the identity of leader.

[0085] Optionally, if containers have been created and started within the nodes included in the cloud cluster switch, then containers belonging to the same type of nodes include master roles and slave roles.

[0086] In this embodiment of the application, master and slave roles are formed between nodes of the same type. For example, Figure 1 In this context, the NCC1 node has the role of master, and the NCC2 node has the role of slave; the NCF1 node has the role of master, and the NCF2 node has the role of slave; the NCP1 node has the role of master, and the NCP2 node has the role of slave.

[0087] The container role can be carried in command-line plugin messages.

[0088] Therefore, the communication system provided in this application is applied in a cloud cluster switch. The cloud cluster switch includes a first node, which includes an admin component and a role component. The admin component is used to obtain the cluster configuration of the cloud cluster switch; determine the role of the first node in the cloud cluster switch from the cluster configuration; and start the corresponding role component according to the role of the first node in the cloud cluster switch. The role component is used to receive a container description file; and create and start a container locally according to the container description file. The container description file is sent by an NCC node with the role of manager and the identity of leader, or the container description file is sent by the admin component included in the NCC node with the role of manager and the identity of leader.

[0089] Therefore, this application proposes a communication system implemented using a cloud cluster switch, where various types of nodes are aggregated into a single cloud cluster switch. Based on container description files, different types of nodes create containers within themselves, and these containers handle the business processing of the cloud cluster switch. Thus, through container technology, the cloud cluster switch can be dynamically managed, satisfying the switch's scalability and robustness requirements, and also adapting well to the current development of AIGC (AI Generic Communication). The aforementioned cloud cluster switch can be deployed within existing DDC (Distributed Data Center) and DCI (Distributed Communication Interface) networks, solving the problems of traditional switches being unable to adapt to the development of AIGC, meet the demands of high-speed data transmission, and address the increasing storage capacity requirements.

[0090] Based on the same inventive concept, embodiments of this application also provide a network device, such as... Figure 3 As shown, the system includes a processor 310, a transceiver 320, and a machine-readable storage medium 330. The machine-readable storage medium 330 stores machine-executable instructions that can be executed by the processor 310. The processor 310 is prompted by the machine-executable instructions to implement the components included in the node of the aforementioned communication system. The node in the aforementioned embodiment may employ, for example... Figure 3 The hardware structure of the network device shown is implemented. The network device provided in this embodiment may specifically be a switch.

[0091] The aforementioned computer-readable storage medium 330 may include random access memory (RAM) or non-volatile memory (NVM), such as at least one disk storage device. Optionally, the computer-readable storage medium 330 may also be at least one storage device located remotely from the aforementioned processor 310.

[0092] The processor 310 mentioned above can be a general-purpose processor, including a central processing unit (CPU), a network processor (NP), etc.; it can also be a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components.

[0093] In addition, this application provides a machine-readable storage medium 330 that stores machine-executable instructions. When called and executed by the processor 310, the machine-executable instructions cause the processor 310 itself and the transceiver 320 to implement the functions of the components included in the nodes of the aforementioned communication system.

[0094] For the device embodiments, since they basically correspond to the system embodiments, relevant details can be found in the descriptions of the system embodiments. The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate, and the components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this application according to actual needs. Those skilled in the art can understand and implement this without any creative effort.

[0095] For network devices and machine-readable storage media embodiments, since the methods involved are basically similar to those described in the foregoing method embodiments, the description is relatively simple, and relevant details can be found in the descriptions of the method embodiments.

[0096] The above description is merely a preferred embodiment of this application and is not intended to limit this application. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of protection of this application.

Claims

1. A communication system, characterized by The communication system is applied in a cloud cluster switch, the cloud cluster switch includes a first node, the first node includes an admin component and a role component; The admin component is used to obtain the cluster configuration of the cloud cluster switch; From the cluster configuration, determine the role of the first node in the cloud cluster switch; based on the role of the first node in the cloud cluster switch, start the corresponding role component; The role component is used to receive a container description file; create and start a container locally according to the container description file; the container is used to implement the business processing of the cloud cluster switch. The container description file is sent by an NCC node with the role of manager and the identity of leader, or the container description file is sent by the admin component included in the NCC node with the role of manager and the identity of leader.

2. The system of claim 1, wherein, When the first node is the first NCC node, the role component is the manager component; The admin component is specifically used to start the manager component if the first NCC node plays the role of manager in the cloud cluster switch. The admin component is also used to determine the identity of the manager; if the manager's identity is leader, then send the first container description file to the manager component; The manager component is used to receive the first container description file; and to create and start the container locally according to the first container description file. The cloud cluster switch performs resource allocation and container scheduling, and sends the second container description file to the NCC node with the identity of follower or the NCF node and NCP node with the role of worker based on the resource allocation results and container scheduling results. The NCF node, the NCP node, and the NCC node whose identity is a follower are all located in the cloud cluster switch.

3. The system according to claim 2, characterized in that, The admin component is further configured to, if the manager's identity is a follower, receive the second container description file sent by the admin component included in the second NCC node, and send the second container description file to the manager component; The manager component is further configured to receive the second container description file, and create and start the container locally according to the second container description file; In this context, the second NCC node plays the role of manager and is identified as leader in the cloud cluster switch.

4. The system according to claim 2, characterized in that, The manager component is also configured to send the second container description file to the third NCC node, so that the manager component included in the third NCC node can create and start the container locally according to the second container description file; The manager component is also used to receive a first container creation success message sent by the third NCC node, wherein the first container creation success message is sent by the manager component included in the third NCC node after it has created and started the container locally; In this context, the third NCC node plays the role of manager and is identified as follower in the cloud cluster switch, and the content device of the first NCC node and the content device of the third NCC node form a master role and a slave role.

5. The system according to claim 2, characterized in that, The manager component is also used to send a first notification message via broadcast in the cloud cluster switch, the first notification message including the leader identity of the manager; The admin component is also used to receive feedback messages sent by other nodes in the cloud cluster switch other than the first NCC node, the feedback messages including the roles or identities of the other nodes; start a cluster joining timer for each other node that sends a feedback message; and determine within a preset time whether the other nodes that have started the cluster joining timer have joined the cloud cluster switch. If so, then remove the timer added to the cluster.

6. The system according to claim 3, characterized in that, The manager component is further configured to: send a second notification message to the second NCC node, the second notification message including a first resource attribute, so that the manager component included in the second NCC node generates the second container description file according to the first resource attribute; and send a second container creation success message to the second NCC node.

7. The system according to claim 2, characterized in that, When the first node is an NCF node or an NCP node, the role component is a worker component; The admin component is specifically used to start the worker component if the NCF node or the NCP node plays the role of a worker in the cloud cluster switch. The worker component is configured to send a third notification message to the first NCC node, the third notification message including a second resource attribute, so that the manager component included in the first NCC node generates the second container description file according to the second resource attribute; receive the second container description file sent by the first NCC node; create a container locally according to the second container description file; and send a third container creation success message to the first NCC node.

8. The system according to claim 1, characterized in that, The admin component is also used to obtain the cluster mode of the cloud cluster switch and configure the cluster management network of the cloud cluster switch according to the cluster mode and the cluster configuration. The cluster management network uses the LIPC protocol, and all nodes in the cloud cluster switch establish communication connections through the LIPC protocol.

9. The system according to claim 1, characterized in that, When a container is created within the first node, the first node also includes an agent component, which is deployed inside the container; The agent component is used to receive command-line plugin messages; convert the command-line plugin messages into cluster configuration and write them to the container configuration file; and send a fourth notification message to the NCC node with the role of manager and the identity of leader when the business state inside the container changes. The command-line plugin message is sent by the NCC node whose role is manager and whose identity is leader.

10. The system according to any one of claims 1-9, characterized in that, If containers have been created and started within the nodes included in the cloud cluster switch, then containers belonging to the same type of nodes include both master and slave roles.

Images