CN116346634A - State sensing information processing method and device of network management and control system and electronic equipment - Google Patents

Abstract

The embodiment of the application discloses a state sensing information processing method and device of a network management and control system and electronic equipment. The method comprises the following steps: establishing a state sensing model under a service scene; real-time monitoring is carried out on the target network system based on the state sensing model, so that state sensing information of the target network system at the current moment is obtained; and performing data cleaning processing on the state sensing information of the target network system at the current moment to obtain cleaned state sensing information of the target network system, and storing the cleaned state sensing information into a database. By the method and the device, the technical problem that the existing network sensing mechanism is difficult to sense the complex network state comprehensively is solved, the complex network state is sensed comprehensively, and the technical effect of keeping high sensitivity and adaptability to the current network topology condition is improved.

Description

State sensing information processing method and device of network management and control system and electronic equipment

Technical Field

The present disclosure relates to the field of computer networks, and in particular, to a method and an apparatus for processing state-aware information of a network management and control system, and an electronic device.

Background

With the widespread use of information and network technology, how to maximize network service benefits and quality of service using limited network resources is a fundamental requirement of communication networks. Network services in certain specific scenes have traffic burstiness, a large amount of network resources are required to be allocated in a short time, and the network resources are efficiently adapted. The state sensing is a process of sensing resources, network topology and network states by fusing different technologies, and provides data support for further network topology modeling and traffic load modeling. However, due to the adverse effects of the factors such as the stiffness of the existing network structure, the complex and changeable network states, the dynamic heterogeneous service function requirements, and the like, the existing network sensing mechanism is difficult to sense the complex network states comprehensively.

For the above-described problems, no effective solution has been proposed.

Disclosure of Invention

The embodiment of the application provides a state sensing information processing method and device of a network management and control system and electronic equipment, and aims to at least solve the technical problem that an existing network sensing mechanism is difficult to sense the state of a complex network comprehensively.

According to an aspect of the embodiments of the present application, there is provided a state-aware information processing method of a network management and control system, including: establishing a state sensing model under a service scene, wherein the state sensing model is used for acquiring state sensing information of a target network system, the target network system is a network management and control system constructed for a plurality of network devices under the service scene, and the state sensing information comprises at least one of the following: network topology information, network state information, and resource information; the target network system is monitored in real time based on the state sensing model, so that state sensing information of the target network system at the current moment is obtained; and performing data cleaning processing on the state sensing information of the target network system at the current moment to obtain cleaned state sensing information of the target network system, and storing the cleaned state sensing information into a database.

Optionally, when the state sensing information is the network topology information, monitoring the target network system in real time based on the state sensing model to obtain state sensing information of the target network system at the current moment, including: the state awareness model comprises a network topology awareness module, wherein the network topology awareness module monitors the target network system in real time based on a software defined network (Software Defined Network, SDN) controller and a link layer discovery protocol (Link Layer Discovery Protocol, LLDP) to obtain current network topology information of the target network system.

Optionally, the monitoring the target network system in real time based on the SDN controller and the LLDP to obtain current network topology information of the target network system includes: configuring ports of SDN switches; establishing connection between the SDN controller and the SDN switch, and judging whether the connection between the SDN controller and the SDN switch is successfully established or not; if the connection between the SDN controller and the SDN switch is successfully established, the SDN controller selects a port of the SDN switch and sends an LLDP data packet to the selected port of the SDN switch; or if the connection between the SDN controller and the SDN switch is not successfully established, updating current network topology information of the target network system, wherein the current network topology information of the target network system includes that the SDN switch connected with the SDN controller does not exist; judging whether the SDN controller receives an LLDP data packet replied by the SDN switch or not; and if the SDN controller receives the LLDP data packet replied by the SDN switch, updating the current network topology information of the target network system, wherein the current network topology information of the target network system comprises the SDN switch connected with the SDN controller.

Optionally, after determining whether the SDN controller receives the LLDP packet replied to by the SDN switch, the method further includes: the SDN controller sending broadcast data discovery protocol (Broadcast DATA Discovery Protocol, BDDP) packets to ports of the SDN switch that have been selected and forwarding the BDDP packets to a non-SDN switch connected to the SDN switch; judging whether the SDN controller receives BDDP data packets replied by the non-SDN switch or not; and if the SDN controller receives the BDDP data packet replied by the non-SDN switch, updating the current network topology information of the target network system, wherein the current network topology information of the target network system comprises the non-SDN switch connected with the SDN switch.

Optionally, after determining whether the SDN controller receives the BDDP packet replied to by the non-SDN switch, the method further includes: judging whether the SDN controller receives an address resolution protocol (Address Resolution Protocol, ARP) data packet, wherein the ARP data packet is sent by a host and forwarded to the SDN controller through a port of the SDN switch; if the SDN controller receives the ARP data packet, updating current network topology information of the target network system, wherein the current network topology information of the target network system comprises the position of the host; and if the SDN controller does not receive the ARP data packet, updating the current network topology information of the target network system, wherein the current network topology information of the target network system comprises the absence of an SDN switch connected with the SDN controller, the absence of a non-SDN switch connected with the SDN switch and the absence of a host connected with the SDN switch.

Optionally, when the state sensing information is the resource information, monitoring the target network system in real time based on the state sensing model to obtain state sensing information of the target network system at the current moment, including: the state sensing model comprises a resource sensing module, wherein the resource sensing module monitors the target network system in real time based on an open source monitoring tool of a Linux operating system and a simple network management protocol to obtain current resource information of the target network system; wherein the resource information is at least one of: CPU usage, disk I/O, queue information, and progress information.

Optionally, the open source monitoring tool includes at least one of: command line based monitoring tools and patterning based monitoring tools.

Optionally, when the state sensing information is the network state information, monitoring the target network system in real time based on the state sensing model to obtain state sensing information of the target network system at the current moment, including: the state sensing model comprises a network state sensing module, wherein the network state sensing module respectively monitors the target network system in real time based on in-band network telemetry, wireless bandwidth estimation and data stream random sampling of a programmable protocol independent message processing language to obtain the current network state information of the target network system; wherein the network status information includes at least one of: link latency, link queue depth, link swallowing amount, port information, latency, jitter, and traffic.

According to another aspect of the embodiments of the present application, there is also provided a state-aware information processing apparatus of a network management and control system, including: the system comprises a building unit, a state sensing module and a control unit, wherein the building unit is used for building a state sensing module in a service scene, the state sensing module is used for obtaining state sensing information of a target network system, the target network system is a network control system built for a plurality of network devices in the service scene, and the state sensing information comprises at least one of the following: network topology information, network state information, and resource information; the monitoring unit is used for monitoring the target network system in real time based on the state sensing model to obtain state sensing information of the target network system at the current moment; the cleaning and storing unit is used for carrying out data cleaning processing on the state sensing information of the target network system at the current moment to obtain the cleaned state sensing information of the target network system, and storing the cleaned state sensing information into the database.

According to another aspect of the embodiments of the present application, there is also provided an electronic device, including: a processor; and a memory storing a program comprising instructions that when executed by the processor cause the processor to perform a method according to any one of the preceding claims.

In the embodiment of the application, a state sensing model under a service scene is established, wherein the state sensing model is used for acquiring state sensing information of a target network system, and the target network system is a network management and control system constructed by a plurality of network devices under the service scene, and the state sensing information comprises at least one of the following components: network topology information, network state information, and resource information; real-time monitoring is carried out on the target network system based on the state sensing model, so that state sensing information of the target network system at the current moment is obtained; and performing data cleaning processing on the state sensing information of the target network system at the current moment to obtain cleaned state sensing information of the target network system, and storing the cleaned state sensing information into a database. That is, the embodiment of the application monitors the target network system in real time based on the state sensing model in the service scene, obtains the state sensing information of the target network system at the current moment, and stores the state sensing information of the cleaned target network system at the current moment into the database, thereby realizing multi-dimensional joint sensing of network node resources, link resources and flow among nodes, further solving the technical problem that the prior network sensing mechanism is difficult to fully sense the complex network state, achieving the technical effects of fully sensing the complex network state, and improving the sensitivity and adaptability to the current network topology state to be high.

Drawings

The accompanying drawings, which are included to provide a further understanding of the application and are incorporated in and constitute a part of this application, illustrate embodiments of the application and together with the description serve to explain the application and do not constitute an undue limitation to the application. In the drawings:

fig. 1 is a flowchart of a method for processing state-aware information of a network management and control system according to an embodiment of the present application;

fig. 2 is a system architecture diagram of a method for processing state-aware information of a network management and control system according to an embodiment of the present application;

fig. 3 is a schematic diagram of an SDN controller using LLDP to discover a specific port link principle according to an embodiment of the present application;

fig. 4 is a schematic diagram of a communication flow of a non-SDN switch existing between switches according to an embodiment of the present application;

fig. 5 is a schematic diagram of an ARP request parsing principle according to an embodiment of the present application;

fig. 6 is a schematic diagram of a network topology sensing flow of an SND controller according to an embodiment of the present application;

FIG. 7 is a schematic diagram of an active measurement process according to an embodiment of the present application;

FIG. 8 is a schematic diagram of a passive measurement process according to an embodiment of the present application;

fig. 9 is a schematic diagram of an sFlow system framework according to an embodiment of the present application;

Fig. 10 is a schematic diagram of a state-aware information processing apparatus of a network management and control system according to an embodiment of the present application.

Detailed Description

In order to make the present application solution better understood by those skilled in the art, the following description will be made in detail and with reference to the accompanying drawings in the embodiments of the present application, it is apparent that the described embodiments are only some embodiments of the present application, not all embodiments. All other embodiments, which can be made by one of ordinary skill in the art based on the embodiments herein without making any inventive effort, shall fall within the scope of the present application.

It should be noted that the terms "first," "second," and the like in the description and claims of the present application and in the drawings are used for distinguishing between different objects and not for defining a particular order.

According to one aspect of the embodiments of the present application, there is provided a state-aware information handling method of a network management and control system, it should be noted that the steps illustrated in the flowchart of the drawings may be performed in a computer system such as a set of computer executable instructions, and that although a logical order is illustrated in the flowchart, in some cases the steps illustrated or described may be performed in an order different from that herein.

Aiming at stronger network state and network resource instantaneity in certain service scenes, in order to reduce the loss of service quality to the greatest extent, a multidimensional resource depth self-perception mechanism needs to be established, and high sensitivity and adaptability are maintained for the current network topology condition. The method and the device establish a state sensing model aiming at the service scene demand, and realize multidimensional joint sensing of network node resources, link resources and flow among nodes.

Fig. 1 is a flowchart of a method for processing state-aware information of a network management and control system according to an embodiment of the present application, as shown in fig. 1, where the method includes the following steps:

step S102, a state sensing model under a service scene is established, wherein the state sensing model is used for acquiring state sensing information of a target network system, the target network system is a network management and control system constructed by a plurality of network devices under the service scene, and the state sensing information comprises at least one of the following steps: network topology information, network state information, and resource information;

the plurality of network devices include, but are not limited to, servers, switches, routers, hosts, and the like. It should be noted that, for the network management and control system, the number of network devices included in the network management and control system is determined according to the requirements of the service scenario.

Step S104, monitoring the target network system in real time based on the state sensing model to obtain state sensing information of the target network system at the current moment;

step S106, data cleaning processing is carried out on the state sensing information of the target network system at the current moment, the cleaned state sensing information of the target network system is obtained, and the cleaned state sensing information is stored in a database.

The data cleansing process described above includes, but is not limited to, defining fields, selecting data ranges, deleting duplicates, processing exception data, missing values, aligning formats, and the like.

Through the steps, the state sensing model under the service scene can be used for monitoring the target network system in real time to obtain the state sensing information of the target network system at the current moment, and the state sensing information of the cleaned target network system at the current moment is stored in the database, so that the multidimensional joint sensing of network node resources, link resources and flow among nodes is realized, the technical problem that the prior network sensing mechanism is difficult to fully sense the state of a complex network is solved, the technical effect that the state of the complex network is fully sensed, and the sensitivity and the adaptability to the current network topology condition are kept high is achieved.

Based on hardware such as a server, a switch, a router, wireless transmission equipment and the like, a Linux environment is built, virtual network elements, controllers and network topology are deployed, and the controllers referred to in the application are SND controllers. Fig. 2 is a system architecture diagram of a state sensing information processing method of a network management and control system, where, as shown in fig. 2, a core of the state sensing system is three modules including network topology sensing, network state sensing and resource sensing, and the network topology sensing is implemented mainly by using an LLDP of an SND controller, and stores sensed network topology information in a MySQL database; network state sensing is mainly realized through in-band telemetry, and sensed real-time network state information including time stamp, queue depth, throughput, link utilization, packet loss, jitter, port ID, switch ID and the like is stored in a database; the resource sensing is realized by combining a Linux operating system open source tool and a simple network management protocol (Simple Network Management Protocol, SNMP), and monitored resource information including CPU utilization rate, disk I/O, queue information, memory, processes and the like is stored in a database. Network state awareness is a module of the network management and control system, and the output data mainly provides input for the route planning and control of the traffic. Information parsing and processing involves the cleansing of some data, including defining some fields, selecting data ranges, deleting duplicates, handling exception data, missing values, etc., and format alignment, etc. The status information in the database further provides support for other modules.

Note that the open network operating system (Open Network Operating System, ONOS) is an SDN controller platform.

In an exemplary embodiment, when the state sensing information is network topology information, the real-time monitoring is performed on the target network system based on the state sensing model to obtain state sensing information of the target network system at the current moment, including: the state sensing model comprises a network topology sensing module, wherein the network topology sensing module monitors the target network system in real time based on a software defined network SDN controller and a link layer discovery protocol LLDP to obtain the current network topology information of the target network system.

Optionally, the network topology aware module performs dynamic topology acquisition through a Secure Shell (SSH) based SND controller application program interface (Application Program Interface, API) based on LLDP, BDDP and ARP of the SND controller. Given a particular type of device, link, and host, the SND controller can discover the network topology with little intervention by an administrator. For example, an SDN switch automatically connects to a designated SND controller using a protocol (e.g., P4 run). When the SND switch establishes a connection with the SND controller, the SND controller will perceive the SND switch and add it to the network topology. Depending on the protocol used to establish the connection, information such as the SND switch processing power and port number may be further communicated to the SND controller through a series of messages in the handshake. Therefore, before the SND controller performs network topology automatic sensing, a communication protocol used by the SDN switch and basic information required for the SDN switch to establish a connection with the SND controller need to be configured.

In an exemplary embodiment, the method for monitoring the target network system in real time based on the SDN controller and the LLDP to obtain current network topology information of the target network system includes: configuring ports of SDN switches; establishing connection between the SDN controller and the SDN switch, and judging whether the connection between the SDN controller and the SDN switch is successfully established or not; if the connection between the SDN controller and the SDN switch is successfully established, the SDN controller selects a port of the SDN switch and sends an LLDP data packet to the port of the selected SDN switch; or if the connection between the SDN controller and the SDN switch is not successfully established, updating the current network topology information of the target network system, wherein the current network topology information of the target network system comprises the SDN switch which is not connected with the SDN controller; judging whether the SDN controller receives LLDP data packets replied by the SDN switch or not; if the SDN controller receives the LLDP data packet replied by the SDN switch, updating the current network topology information of the target network system, wherein the current network topology information of the target network system comprises the SDN switch connected with the SDN controller.

It should be noted that LLDP is a data link layer protocol. The SND controller sends LLDP data packets to all SDN switch ports according to port information acquired when connection is established with the SDN switches, performs hop-by-hop link discovery, and detects whether the two SDN switch ports are directly connected. In the control domain of the SND controller, the LLDP data packet is sent to a designated port of the switch by the SND controller, the designated switch port forwards the LLDP data packet, the LLDP data packet can only be forwarded for one hop, and any switch receives the LLDP data packet sent by the non-SND controller and needs to directly send the LLDP data packet to the SND controller. The SDN controller discovers a particular port link principle using LLDP as shown in fig. 3.

The SND controller first issues a packet_out message, i.e., an LLDP packet, to port 2 of switch1, where the LLDP packet includes port 2 information of the switch. switch1 reads port 2 information of the switch in the LLDP packet and forwards the packet out of the port designated by the SND controller.

When the SND switch2 receives the LLDP packet, the LLDP packet read is sent by port 2 of switch1 instead of by the SND controller, so the switch must send the LLDP packet to the SND controller without forwarding.

Switch2 builds a PACKET _ IN message and adds the port that received the PACKET (Switch port 4) to the data payload, which is reported to the SND controller. The SND controller can know that the port 2 of the switch1 and the port 4 of the switch2 are connected according to the information (the port 2 of the switch1 sends and the port 4 of the switch2 receives) contained in the LLDP packet, thereby completing one-time link discovery detection.

Because LLDP packets have the property of only one-hop survival time, i.e. are only present in two adjacent switches. The SND controller can only perform link discovery on a piece-by-piece basis and one link discovery can only prove that the link is one-way communication. The SND controller also needs to perform the link discovery from switch2 port 4 to switch1 port 2 according to the above procedure to display the bidirectional link communication of the two SDN switch ports in the network topology.

In an exemplary embodiment, after determining whether the SDN controller receives the LLDP packet replied to by the SDN switch, the method further includes: the SDN controller sends a broadcast data discovery protocol BDDP data packet to a port of the selected SDN switch, and forwards the BDDP data packet to a non-SDN switch connected with the SDN switch; judging whether the SDN controller receives BDDP data packets replied by the non-SDN switch or not; if the SDN controller receives the BDDP data packet replied by the non-SDN switch, updating the current network topology information of the target network system, wherein the current network topology information of the target network system comprises the non-SDN switch connected with the SDN switch.

Unlike LLDP, which can only forward one hop, BBDP packets are forwarded by non-SDN switches without limiting the number of hops. The SND controller detects whether a non-SDN switch exists in the SDN switch interconnection using the BDDP protocol, and the specific principle is shown in fig. 4.

Firstly, IN the SND controller, the LLDP PACKET is preferentially issued to the port 2 of the switch1 through the packet_out message, the switch1 forwards the PACKET according to the port specified IN the LLDP PACKET, if the PACKET passes through the non-SDN switch, the LLDP PACKET is discarded by the non-SDN switch, and the SND controller cannot receive the packet_in message of the PACKET.

Then, the SND controller will issue a BDDP PACKET to switch1 through a packet_out message, forward the BDDP PACKET through a port corresponding to switch1, and upload the BDDP PACKET to the controller through a packet_in message after forwarding the BDDP PACKET by the non-SDN switch, where the SND controller may determine that the non-SDN switch exists between switch1 and switch 2.

IN addition, there is a case that the LLDP PACKET is lost, and the BDDP PACKET is not reported to the SND controller by packet_in, it can be determined that the 2 port of switch1 is connected to a host, and is not connected to the SND switch. Regardless of how many non-SDN switches the BDDP packet passes through, the SND controller will look to it as a one-hop forwarding.

In an exemplary embodiment, after determining whether the SDN controller receives the BDDP packet replied to by the non-SDN switch, the method further includes: judging whether the SDN controller receives an Address Resolution Protocol (ARP) data packet, wherein the ARP data packet is sent by a host and forwarded to the SDN controller through a port of an SDN switch; if the SDN controller receives the ARP data packet, updating the current network topology information of the target network system, wherein the current network topology information of the target network system comprises the position of the host; if the SDN controller does not receive the ARP data packet, updating current network topology information of the target network system, wherein the current network topology information of the target network system comprises the absence of an SDN switch connected with the SDN controller, the absence of a non-SDN switch connected with the SDN switch and the absence of a host connected with the SDN switch.

It should be noted that ARP is a TCP/IP protocol that obtains a physical address from an IP address. When the host sends the information, the ARP request containing the target IP address is broadcast to all hosts on the local area network, and a return message is received, so that the physical address of the target is determined. The SND controller detects a host connected with the SDN switch through an ARP protocol. When a host is connected to the network topology, ARP requests are broadcast to the whole network and the SND controller updates the host location in the network topology by analyzing the receiving ports of the ARP requests.

When the host 1 pings the host 2, i.e. 10.0.0.1ping 10.0.0.2, there is no MAC address of the host 2 at this time, so ARP parsing needs to be performed once, and at this time, the ARP request of the host 1 will be reported to the controller by the switch1 in the form of a packet_in message; after receiving the ARP request, the SND controller records the position of the host 1, requests the position of the host 2 from all SND switches in the same network segment, and sends packet_out to all switches in a flooding mode; after all SND switches receive the request, checking whether the host exists or not, and only switch2 will reply an ARP to the SND controller in the form of packet_in; when the SND controller receives the reply of switch2, it records the position of host 2, and sends the ARP reply to switch1 in packet_out message, at this time, switch1 sends to host 1 again, and the principle of the request analysis of the process is shown in FIG. 5.

In summary, the network topology aware flow of the SND controller is shown in fig. 6.

In an exemplary embodiment, when the state sensing information is resource information, monitoring the target network system in real time based on the state sensing model to obtain state sensing information of the target network system at the current moment, including: the state sensing model comprises a resource sensing module, wherein the resource sensing module monitors the target network system in real time based on an open source monitoring tool of the Linux operating system and a simple network management protocol to obtain the current resource information of the target network system; wherein the resource information is at least one of: CPU usage, disk I/O, queue information, and progress information.

The distributed network management and control system operates in a Linux environment, so that the overall resource monitoring of the Linux system is necessary. The resource perception module is based on the open source monitoring tool technology of Linux, and has the advantages of online monitoring, high cost performance, visualization and strong compatibility.

It should be noted that, the software environment of the resource sensing module of the state sensing model is a Linux operating system, and in this environment, the monitoring of the current resource usage situation can be realized based on a Linux operating system open source monitoring tool, such as nomon, dstat, top, vmstat, iostat, and the SNMP protocol, and compared with the implementation of other plug-ins, the resource sensing function can be realized, and the monitoring tool is utilized to have the advantage of consuming less system resources.

In an exemplary embodiment, the open source monitoring tool comprises at least one of: command line based monitoring tools and patterning based monitoring tools.

Optionally, the command line based monitoring tool includes the following:

1) dstat: three commands of vmstat, iostat and ifstat are integrated, and new characteristics and functions are added, so that various resource use conditions can be observed in time. The information can be acquired more clearly and easily through the help of interfaces with different colors and block layouts. In addition, the information data is exported to a cvs format file and imported to a database, and the command can monitor the change of CPU, memory and network state with time;

2) atop: all process activities can be displayed, daily system logs are displayed for long-term process activity analysis, and overloaded system use resources are highlighted, so that measurement indexes of a CPU (central processing unit), a memory, a switching space, a disk and a network layer can be monitored;

3) sar: the selected accumulated activity counter content information on the operating system may be output onto a standard output. The auditing system based on the count value and the time interval parameter outputs monitoring information of a designated number of times according to a designated time interval. If the time interval parameter is set to 0, the sar command displays average statistical information of the system from starting up to the current moment;

4) top: unix-like operating system task manager. The list of the currently running processes can be displayed, the list is ordered according to different conditions, and the use conditions of the system processes on the CPU and the memory are displayed;

5) Sysdig: information about the unified ordering and granularity of storage, processes, networks, and memory subsystems can be provided, and system activity record files can be created for analysis;

6) iftop: top-like programs based on network information. The network connection status ordered according to the bandwidth usage amount or the uploading or downloading amount at the current moment can be displayed, and the estimated completion time of the downloaded file is provided;

7) Aperf: TCP and UDP data connections can be created and their transmission performance measured over the network. Support for adjusting different parameters with respect to time, protocols, buffering, etc. For each test, bandwidth, packet loss, and some other parameters will be reported.

In addition, the system resource related information can be perceived and stored in the database in real time by using a monitoring tool based on a command line.

Optionally, the graphical-based monitoring tool includes the following: the monitoring tools such as Linux process explorer and Collectl, MRTG, monit are more beneficial to visualizing system resources, and are simple, convenient and easy to understand.

In an exemplary embodiment, when the state sensing information is network state information, monitoring the target network system in real time based on the state sensing model to obtain state sensing information of the target network system at the current moment, including: the state sensing model comprises a network state sensing module, wherein the network state sensing module respectively monitors the target network system in real time based on in-band network telemetry, wireless bandwidth estimation and data stream random sampling of a programmable protocol independent message processing language to obtain the current network state information of the target network system; wherein the network status information includes at least one of: link latency, link queue depth, link swallowing amount, port information, latency, jitter, and traffic.

For network state awareness of the data plane, in-band network telemetry (In-band network telemetry, INT) and sFlow schemes based on programmable protocol independent message processing language (P4, programming protocol-independent packet processors) will be employed, and the INT can collect statistical network state information including, but not limited to, switch ID, timestamp, egress/ingress port number, queue length, link utilization, etc. The sFlow mainly realizes the monitoring of flow fine granularity, adopts a data flow random sampling technology, can provide complete flow information from a second layer to a fourth layer and even within a whole network range, can adapt to flow analysis under the environment of ultra-large network flow (such as more than 10 Gbit/s), and can analyze the performance, trend and existing problems of network transmission flow in real time. INT and sFlow are jointly perceived, and the requirements of the current background can be met. In addition, the wireless bandwidth measurement adopts a bandwidth estimation scheme, and the estimated bandwidth is adjusted through indexes such as the round trip delay RTT, the packet loss rate and the like of the acquired data packet until the estimated bandwidth approaches to a real bandwidth value.

INT is realized based on P4 language, intervention of control plane is not needed, the data plane software exchanger is programmed, the INT collects network information according to the remote measuring instruction on the carried message head, each node passing through the message, and the collected information is added to the message, after the message reaches the destination, the INT destination node can extract flow information of each node on the message passing through path and upload the flow information to the remote data collecting controller, and the remote measuring flexibility is improved.

Typically, an INT domain contains 3 major functional nodes: INT-source, INT-transit, and INT-sink. FIG. 7 is a schematic diagram of an active measurement process provided by an embodiment of the present application, as shown in FIG. 7, the INT-source switch encapsulates the telemetry probe into an INT header and inserts telemetry instructions; the INT-transit exchanger collects and inserts network state data and forwards the data to the next node according to the remote flow table; the INT-sink extracts and integrates all telemetry information and reports to a remote INT data collection controller. Fig. 8 is a schematic diagram of a passive measurement process provided in an embodiment of the present application, as shown in fig. 8, in the passive probing method, a last hop switch acts as an INT-sink role, and the actions performed include:

1) Extracting and integrating all telemetry information;

2) Reporting to a remote INT data collection controller;

3) And restoring the original user data packet according to the destination IP or the source IP and forwarding the original user data packet to the terminal.

Further, the clustering method is based on clustering processing of the network topology, switches which are governed by the clustering method are distributed to each cluster, each switch can act as a cluster head, the switches in the cluster upload the generated telemetry report to the cluster head, and the cluster head uploads the telemetry report to a collector; the collector counts the network state information in the collected telemetry report, selects a switch which does not collect the network state information for a long time, and determines the switch which does not collect the network state information for a long time as a designated switch; the detection data packet is sent to the appointed exchanger, and the collector can confirm the state of the appointed exchanger according to the feedback information of the received detection data packet.

When network state sensing is carried out, each exchanger needs to judge the state of an incoming data packet, and if the current data packet state meets the requirement of inserting network state information, the network state information is inserted; if the current data packet state does not meet the requirement of inserting network state information, extracting the collected network state information in the data packet to generate a telemetry report, and sending the telemetry report to a cluster head, wherein the cluster head uploads the data to a corresponding collector; deleting the INT head inserted by the data packet, and restoring the data packet; when the user data packet arrives at the last hop switch, the last hop switch generates a telemetry report, and cuts off the rest invalid loads except the INT head and the INT data part in the telemetry report according to the total length of the INT information. Each collector re-uploads the collected network state information to a remote data collector, which reduces controller level overhead to a greater extent than if a centralized collector were employed.

The INT-source needs to encapsulate the INT header and push the INT data, and finally, the INT-sink de-encapsulates the INT header and the INT data based on the source and target IP, and restores the initial data packet.

Fig. 9 is a schematic diagram of an sFlow system framework provided in an embodiment of the present application, where, as shown in fig. 9, the sFlow system includes an sFlow Agent (embedded in a forwarding device such as a switch or a router) and an sFlow Collector of a core. The sFlow Agent acquires flow forwarding statistics on the network equipment through a specific sampling technology, and sends the flow forwarding statistics to the Collector in real time through sFlow data messages for analysis by the Collector, and a network manager is helped to manage network flow of the whole station more effectively through a form of generating a flow view or report.

Further, sFlow is a sampling-based technique, where packet flow sampling (Packet flow sampling) samples network packets through a device port at a sampling rate N; counter sampling (Counter sampling) obtains statistics for each port from a port Counter (interface Counter) at sampling interval timing. Two sampling mechanisms produce corresponding samples: a flow sample (flow sample) and a counter sample (counter sample). The stream samples describe the network packet characteristics and contain not only header information of the original packet but also information related to the original packet transfer process. The counter sample describes port traffic characteristics and also contains a variety of rich count information. The information in the samples is organized in a record (record) format, corresponding to a stream record (flow record) and a counter record (counter record). Through INT+sFlow joint perception, fine granularity monitoring of network states in the scene can be realized, and a large amount of accurate data support is provided for other decision-making modules.

Further, the link bandwidth is the number of bits a link can transmit per second. When the number of bits transmitted per second is greater than the link bandwidth, link overload can occur, resulting in a loss of a large number of packets, which can degrade network performance. When the number of bits transmitted per second is much smaller than the link bandwidth, the bandwidth of the link is underutilized resulting in lower network throughput. Therefore, the link bandwidth needs to be accurately estimated, so that the bit number transmitted per second is slightly smaller than the link bandwidth, the link bandwidth can be fully utilized, and the network performance is improved.

Since the bandwidth of the wireless link cannot be directly obtained through detection like Time delay and packet loss rate, the bandwidth estimation is performed by considering parameters such as Round-Trip Time (RTT) and packet loss rate of the obtained link. According to the theory of the earliest delivery path first algorithm (Earliest Deadline and Processing First, EDPF), it can maximize throughput with minimal in-order transmission delay, i.e., the total delay per link during transmission is approximately equal. Based on this conclusion, the bandwidth estimation of the link may be adjusted by the maximum delay difference of each link.

In order to allow the estimated link bandwidth to converge faster to the actual link bandwidth, the initialization link bandwidth setting cannot be too large nor too small. Considering that the link bandwidth of a wireless environment typically ranges between 1Mbps and 10Mbps, and that the conditions of the links are not the same, some of the link bandwidths are higher, and some of the link bandwidths may be lower. Therefore, the initialization bandwidth of each link is set to about 5 Mbps.

When the maximum delay delta RTT of each link is smaller than a certain time threshold, the bandwidth estimation of each link is accurate, and the estimated link bandwidth does not need to be adjusted. If the link bandwidth estimation is not right, the delay estimation of each link is inaccurate, and the delay of each link measured by the link environment detection module has larger difference.

When the maximum delay delta RTT of each link is greater than a certain time Threshold value Threshold, the estimated bandwidth of each link is adjusted, and the adjustment expression is as follows:

B _i,cur ＝B _i,last +B _i,last *(T _i,last -SRTT _i,cur /2)*e ^-β

wherein T is _i,last Represents the last adjusted link estimated delay, B _i,last Indicating last adjusted link estimated bandwidth, SRTT _i,cur Indicating the link smooth RTT just updated and β indicating the adjustment factor. When the last estimated link bandwidth is smaller than the actual link bandwidth, the last estimated delay is greater than the delay currently detected, and therefore the estimated link bandwidth needs to be enlarged. Otherwise, when the last estimated link bandwidth is larger than the actual link bandwidth, the last estimated delay is smaller than the delay obtained by the current detection, and the estimated link bandwidth needs to be reduced.

According to another aspect of the embodiments of the present application, there is further provided a status awareness information processing device of a network management and control system, and fig. 10 is a schematic diagram of the status awareness information processing device of the network management and control system provided in the embodiments of the present application, as shown in fig. 10, where the status awareness information processing device of the network management and control system includes: a setup unit 1002, a monitoring unit 1004 and a cleaning and storage unit 1006. The state-aware information processing apparatus of the network management and control system will be described in detail.

The establishing unit 1002 is configured to establish a state sensing model in a service scenario, where the state sensing model is configured to obtain state sensing information of a target network system, and the target network system is a network management and control system constructed by a plurality of network devices in the service scenario, and the state sensing information includes at least one of the following: network topology information, network state information, and resource information;

the monitoring unit 1004 is connected to the establishing unit 1002, and is configured to monitor the target network system in real time based on the state sensing model, so as to obtain state sensing information of the target network system at the current moment;

the cleaning and storing unit 1006 is connected to the monitoring unit 1004, and is configured to perform data cleaning processing on the state sensing information of the target network system at the current moment, obtain cleaned state sensing information of the target network system, and store the cleaned state sensing information in the database.

The device of the embodiment of the application monitors the target network system in real time based on the state sensing model in the service scene, obtains the state sensing information of the target network system at the current moment, and stores the state sensing information of the cleaned target network system at the current moment into the database, thereby realizing multi-dimensional joint sensing of network node resources, link resources and flow among nodes, further solving the technical problem that the prior network sensing mechanism is difficult to fully sense the complex network state, achieving the technical effects of fully sensing the complex network state, and improving the sensitivity and adaptability to the current network topology state to be high.

It should be noted that the above-mentioned establishing unit 1002, monitoring unit 1004, and cleaning and storing unit 1006 correspond to steps S102 to S108 in the method embodiment, and the above-mentioned units are the same as examples and application scenarios implemented by the corresponding steps, but are not limited to those disclosed in the above-mentioned method embodiment.

According to another aspect of the embodiments of the present application, there is also provided an electronic device, including: a processor; and a memory storing a program comprising instructions which, when executed by the processor, cause the processor to perform a method according to any one of the preceding claims.

The foregoing embodiment numbers of the present application are merely for describing, and do not represent advantages or disadvantages of the embodiments.

In the foregoing embodiments of the present application, the descriptions of the embodiments are emphasized, and for a portion of this disclosure that is not described in detail in this embodiment, reference is made to the related descriptions of other embodiments.

In the several embodiments provided in the present application, it should be understood that the disclosed technology content may be implemented in other manners. The above-described embodiments of the apparatus are merely exemplary, and the division of the units, for example, may be a logic function division, and may be implemented in another manner, for example, a plurality of units or components may be combined or may be integrated into another system, or some features may be omitted, or not performed. Alternatively, the coupling or direct coupling or communication connection shown or discussed with each other may be through some interfaces, units or modules, or may be in electrical or other forms.

The units described as separate parts may or may not be physically separate, and parts displayed as units may or may not be physical units, may be located in one place, or may be distributed on a plurality of units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

In addition, each functional unit in each embodiment of the present application may be integrated in one processing unit, or each unit may exist alone physically, or two or more units may be integrated in one unit. The integrated units may be implemented in hardware or in software functional units.

The integrated units, if implemented in the form of software functional units and sold or used as stand-alone products, may be stored in a computer readable storage medium. Based on such understanding, the technical solution of the present application may be embodied in essence or a part contributing to the prior art or all or part of the technical solution in the form of a software product stored in a storage medium, including several instructions to cause a computer device (which may be a personal computer, a server or a network device, etc.) to perform all or part of the steps of the methods described in the embodiments of the present application. And the aforementioned storage medium includes: a U-disk, a Read-Only Memory (ROM), a random access Memory (RAM, random Access Memory), a removable hard disk, a magnetic disk, or an optical disk, or other various media capable of storing program codes.

The foregoing is merely a preferred embodiment of the present application and it should be noted that modifications and adaptations to those skilled in the art may be made without departing from the principles of the present application and are intended to be comprehended within the scope of the present application.

Claims (10)

1. The state sensing information processing method of the network management and control system is characterized by comprising the following steps of:

establishing a state sensing model under a service scene, wherein the state sensing model is used for acquiring state sensing information of a target network system, the target network system is a network management and control system constructed for a plurality of network devices under the service scene, and the state sensing information comprises at least one of the following: network topology information, network state information, and resource information;

the target network system is monitored in real time based on the state sensing model, so that state sensing information of the target network system at the current moment is obtained;

and performing data cleaning processing on the state sensing information of the target network system at the current moment to obtain cleaned state sensing information of the target network system, and storing the cleaned state sensing information into a database.

2. The method according to claim 1, wherein when the state sensing information is the network topology information, monitoring the target network system in real time based on the state sensing model to obtain state sensing information of the target network system at a current time, includes:

the state sensing model comprises a network topology sensing module, wherein the network topology sensing module monitors the target network system in real time based on a software defined network SDN controller and a link layer discovery protocol LLDP to obtain current network topology information of the target network system.

3. The method of claim 2, wherein the real-time monitoring of the target network system based on the SDN controller and the LLDP to obtain current network topology information of the target network system comprises:

configuring ports of SDN switches;

establishing connection between the SDN controller and the SDN switch, and judging whether the connection between the SDN controller and the SDN switch is successfully established or not;

if the connection between the SDN controller and the SDN switch is successfully established, the SDN controller selects a port of the SDN switch and sends an LLDP data packet to the selected port of the SDN switch; or if the connection between the SDN controller and the SDN switch is not successfully established, updating current network topology information of the target network system, wherein the current network topology information of the target network system includes that the SDN switch connected with the SDN controller does not exist;

Judging whether the SDN controller receives an LLDP data packet replied by the SDN switch or not;

and if the SDN controller receives the LLDP data packet replied by the SDN switch, updating the current network topology information of the target network system, wherein the current network topology information of the target network system comprises the SDN switch connected with the SDN controller.

4. The method of claim 3, wherein after determining whether the SDN controller receives an LLDP packet replied to by the SDN switch, the method further comprises:

the SDN controller sends a broadcast data discovery protocol BDDP to the port of the SDN switch which has been selected

The BDDP data packet is forwarded to a non-SDN switch connected with the SDN switch;

judging whether the SDN controller receives BDDP data packets replied by the non-SDN switch or not;

and if the SDN controller receives the BDDP data packet replied by the non-SDN switch, updating the current network topology information of the target network system, wherein the current network topology information of the target network system comprises the non-SDN switch connected with the SDN switch.

5. The method of claim 4, wherein after determining whether the SDN controller receives BDDP packets replied to by the non-SDN switch, the method further comprises:

judging whether the SDN controller receives an Address Resolution Protocol (ARP) data packet, wherein the ARP data packet is sent by a host and forwarded to the SDN controller through a port of the SDN switch;

if the SDN controller receives the ARP data packet, updating current network topology information of the target network system, wherein the current network topology information of the target network system comprises the position of the host;

and if the SDN controller does not receive the ARP data packet, updating the current network topology information of the target network system, wherein the current network topology information of the target network system comprises the absence of an SDN switch connected with the SDN controller, the absence of a non-SDN switch connected with the SDN switch and the absence of a host connected with the SDN switch.

6. The method according to claim 1, wherein when the state sensing information is the resource information, monitoring the target network system in real time based on the state sensing model to obtain state sensing information of the target network system at a current time, includes:

The state sensing model comprises a resource sensing module, wherein the resource sensing module monitors the target network system in real time based on an open source monitoring tool of a Linux operating system and a simple network management protocol to obtain current resource information of the target network system; wherein the resource information is at least one of: CPU usage, disk I/O, queue information, and progress information.

7. The method of claim 6, wherein the open source monitoring tool comprises at least one of: command line based monitoring tools and patterning based monitoring tools.

8. The method according to claim 1, wherein when the state sensing information is the network state information, monitoring the target network system in real time based on the state sensing model to obtain state sensing information of the target network system at a current time, including:

the state sensing model comprises a network state sensing module, wherein the network state sensing module respectively monitors the target network system in real time based on in-band network telemetry, wireless bandwidth estimation and data stream random sampling of a programmable protocol independent message processing language to obtain the current network state information of the target network system; wherein the network status information includes at least one of: link latency, link queue depth, link swallowing amount, port information, latency, jitter, and traffic.

9. A state-aware information processing apparatus of a network management and control system, comprising:

the system comprises a building unit, a state sensing module and a control unit, wherein the building unit is used for building a state sensing module in a service scene, the state sensing module is used for obtaining state sensing information of a target network system, the target network system is a network control system built for a plurality of network devices in the service scene, and the state sensing information comprises at least one of the following: network topology information, network state information, and resource information;

the monitoring unit is used for monitoring the target network system in real time based on the state sensing model to obtain state sensing information of the target network system at the current moment;

the cleaning and storing unit is used for carrying out data cleaning processing on the state sensing information of the target network system at the current moment to obtain the cleaned state sensing information of the target network system, and storing the cleaned state sensing information into the database.

10. An electronic device, comprising: a processor; and a memory storing a program, characterized in that the program comprises instructions which, when executed by the processor, cause the processor to perform the method according to any one of claims 1 to 8.

Images