WO2016177190A1

WO2016177190A1 - Method and device for auto-negotiating lacp reference system and reference port

Info

Publication number: WO2016177190A1
Application number: PCT/CN2016/077579
Authority: WO
Inventors: 郭宇
Original assignee: 中兴通讯股份有限公司
Priority date: 2015-07-27
Filing date: 2016-03-28
Publication date: 2016-11-10
Also published as: CN106411731B; CN106411731A

Abstract

The present invention relates to the technical field of link aggregations. Disclosed are a method and device for auto-negotiating a link aggregation control protocol (LACP) reference system and a reference port. The method comprises the following steps: determining, by a first system and a second system of an LACP link aggregation group (LAG), a system stability thereof respectively via a real-time detection on a real-time resource usage condition of the respective system; selecting, by the first system and the second system via a negotiation, a system having the best stability therebetween as a reference system; selecting, by the reference system, a port from ports of the reference system as a reference port, according to condition history information of each of port pairs constituted by each of the ports and a corresponding directly-connected port of an opposite system and a resource usage condition of a single board to which the port belongs. The present invention introduces records of the system, the resource of the single board, and the port condition history information to auto-negotiate and adjust the reference system and reference port, thus improving the stability of a link bearing service flows.

Description

Method and device for self-negotiating LACP protocol reference system and reference port

Technical field

The present invention relates to the field of link aggregation technologies, and in particular, to a method and an apparatus for a self-negotiating LACP protocol (Link Aggregation Control Protocol) reference system and a reference port.

Background technique

LA (Link Aggregation) technology is characterized in that multiple physical links are aggregated into a higher-traffic logical link whose bandwidth is equal to the bandwidth of multiple physical links that are aggregated together. with. The LACP protocol, defined by the IEEE 802.3ad standard, is a standard Layer 2 protocol for dynamically forming link aggregation groups between devices. The basic principle of LACP is to dynamically detect the status and information of the peer port through the periodic exchange of packets between the two device ports, and determine whether the port joins or leaves a LAG (Link Aggregation Group). For the convenience of description, in LACP, for the devices at both ends of the aggregation link, each device port is called an Actor, and the peer device port is a Partner, as shown in Figure 1. In an aggregation group, the possible states of the port are selected, standby, and unselected. The member port in the selected state is the port that can finally carry the service (the port that actually delivers the hardware). The port in the standby state is the port that meets the aggregation condition but cannot be aggregated due to the maximum number of aggregated ports. The ports in the unselected state are those that are not at all. A port that satisfies the aggregation condition (such as link down).

The LACP protocol exchanges information with the peer through the LACPDU (Link Aggregation Control Protocol Data Unit) message. The format of the LACPDU packet is shown in Figure 2. The priority of the local and remote systems, the priority of the local and remote ports, and the key value of the local and remote ports. Based on this information, both parties of the aggregation select the appropriate link according to a certain selection algorithm to control the state of the aggregation.

The following explains the meaning of the main information of the LACPDU according to the order of importance:

1. System priority: The LACP protocol needs to distinguish the system priorities of the devices at both ends. The device with the highest priority (system) serves as the reference system, and the other device selects the selected active port according to the reference system configuration, as shown in Figure 3. In addition, if the LACP priorities of the two systems are the same, the MAC address (Media Access Control) address of the devices at both ends is compared (only 802.3 MAC is supported). The smaller the value, the higher the priority.

2. Port priority: It is used to define the priority of the port selected as the reference port. Other ports need to be the same as the reference port's KEY value to become the selected active port to participate in data forwarding. Otherwise, it is an unselected port and cannot forward data. As shown in Figure 4, if the LACP priorities of the ports are the same, the port numbers of the two ports are compared. The smaller the value, the higher the priority.

3. The KEY value of the port is obtained by combining the aggregation group number, the port rate, and the duplex mode. The LACP protocol determines whether the port is consistent by comparing the KEY values of the port and the reference port. Can become a selected active port;

The workflow of the LACP protocol is as follows:

The device at both ends completes the negotiation and configuration of the link aggregation group by sending LACP packets. The working process is as follows:

1, determine the reference system

Among the two devices, whoever has higher system priority will become the reference system. After being used as a reference system, the selected active port selected by it becomes the reference standard, and the peer device takes it as the standard.

2, determine the reference port

Among the ports of the reference device, whoever has the higher port priority and the current port and peer port status are normal, who will become the reference port. The reference port is selected as the selected active port, and the other ports require the same KEY value as the reference port to become the selected active port.

3. Identify other selected active ports

The remaining non-reference ports, if they also become selected active ports, the KEY value of the port should be the same as the reference port.

The above is the approximate workflow of the LACP protocol. The final protocol will complete the dynamic formation of the aggregation port, and then the data can be sent directly on the aggregation port. From the perspective of how to distribute data flows, the LACP protocol can be divided into two modes: load balancing mode and protection mode. In load balancing mode, all selected active ports will carry service flows. In protected mode, The port is responsible for carrying all the traffic, and the other selected ports are negative for the link-level protection of n:1, so the selection of the reference port is significant.

The LACP protocol is widely used in the industry, but there are several shortcomings in the application of the standard protocol directly: 1. The selection of a device with high system priority (reference system) is crucial because its configuration directly determines which selected activities are adopted by the peer device. The port (the selected active port of the peer is the direct port of the port configured corresponding to the reference system), and once the system priority is set, the stable operation of LACP will rely more on the device with higher priority. For example, in the MC-LAG (Multi-Chassis Link Aggregation Group) scenario, the user prefers to determine the reference system based on the real-time resource status of the system in real time. 2. The current priority value of each port is also If the local port is a member of a different board, the local LACPDUs are not related to the real-time resource information of the board. It is not limited to cpu utilization and memory usage. The ability of different boards to process LACPDU packets is different, but once the reference port is selected. The board is faulty on the board where the peer port is directly connected (for example, the CPU usage is high, or there is a memory leak, etc.). Especially in the protection mode, the LACP protocol may be reported only after the resources are really exhausted. If the negotiation fails, the switch can perform the switchover of the reference port. The switchover cannot be performed in a timely manner. If the protocol packet negotiation fails, the service flow may have been lost. In the LACP standard protocol, the historical abnormal state information of each port does not participate in the decision of the port priority (the setting of the port priority does not consider this factor), assuming that a port has a short-term probability Link down, but because its (or its peer port) meets the reference port condition, each port is selected as the reference port by default after the port is normal. This choice will affect the stability of the system link bearer service flow.

Summary of the invention

The purpose of the embodiments of the present invention is to provide a method and a device for auto-negoing the LACP protocol reference system and the reference port, which solves the problem that the stability of the reference link cannot be timely switched to affect the stability of the system link bearer service flow in the prior art.

According to an embodiment of the present invention, a method for auto-negotiating a LACP protocol reference system and a reference port is provided, including the following steps:

The first system and the second system of the LACP protocol link aggregation group determine the stability of the respective systems by performing real-time detection on the real-time resource usage status of the respective systems;

The first system and the second system select, by negotiation, a system with the best stability in the first system and the second system as a reference system;

The reference system selects a port from the reference port of the reference system as a reference according to the state history information of each port pair formed by each port and the corresponding peer system direct port and the resource usage status of the board. port;

The LACP protocol refers to a link aggregation control protocol.

In the embodiment of the present invention, the status history information of each port pair and the resource usage status of the associated board include:

Port state history information of each port and the resource usage status of the card to which the port belongs.

Port state history information of the peer system port directly connected to each port and the resource usage status of the board to which the port belongs.

In the embodiment of the present invention, the resource usage status includes a board CPU utilization rate and a board memory usage rate; and the port status history information includes the number of times the port link is disconnected in the current time period.

In the embodiment of the present invention, the first system constructs a resource usage metric value of each board of the first system according to a resource usage status of a card to which the port of each port of the system belongs; The resource usage status of each board of the second system is constructed, and the resource usage metric of each board in the second system is constructed.

The resource usage metric of each board in the first system includes: a board CPU utilization rate and a board memory usage rate of the board to which the port of each port of the first system belongs; the second system The resource usage metrics of the boards include: the CPU usage of the board and the memory usage of the board of the board to which the port of each port of the second system belongs.

In the embodiment of the present invention, the determining respective system stability includes:

Selecting a maximum resource usage metric value as the stability of the first system in a resource usage metric value of each board of the first system;

Selecting a maximum resource usage metric value as the stability of the second system in a resource usage metric value of each board of the second system;

The greater the resource usage metric, the worse the stability.

In the embodiment of the present invention, according to the resource usage status of the board to which the port of each port belongs, and each port The resource usage status of the board to which the port of the peer system port is directly connected, and the resource usage metric of each board in the reference system.

The CPU usage and board memory usage of the board of the card to which the port of each port belongs.

The CPU usage and board memory usage of the board of the board to which the port of the peer system port is directly connected to each port.

In the embodiment of the present invention, the reference system is based on the state history information of each port pair formed by each port and the corresponding peer system direct port, and the resource usage status of the card to which the card belongs, from the port to which the reference system belongs. Selecting a port as a reference port includes:

You can sort the resource usage metrics of the card to the port to be referenced.

Detecting the number of times the port link is disconnected in the current time period of the port to be referenced;

When the number of times the port link is disconnected in the current time period of the port to be referenced is 0, the port to be referenced is used as a reference port in the reference system.

According to another embodiment of the present invention, an apparatus for auto-negotiating a LACP protocol reference system and a reference port is provided, including:

Determining the stability module, the first system and the second system set to the LACP protocol link aggregation group determine the stability of the respective systems by performing real-time detection on the real-time resource usage status of the respective systems;

Selecting a reference system module, configured to select, by the first system and the second system, a system with the best stability in the first system and the second system as a reference system;

Selecting a reference port module, and setting the status history information of each port pair formed by the reference system according to each port and the corresponding peer system direct port, and the resource usage status of the board to which the reference system belongs, from the port to which the reference system belongs. Select a port as the reference port;

The LACP protocol refers to a link aggregation control protocol.

Compared with the prior art, the beneficial effects of the embodiments of the present invention are:

The embodiment of the present invention introduces the recording of the resource and the port state history information of the system board, and combines the transmission of the LACPDUs between the two systems, and automatically adjusts the system priority and the port priority to help improve the bearer service flow chain. Road stability.

DRAWINGS

1 is a schematic diagram of a network running the LACP protocol provided by the prior art;

2 is a schematic diagram of a packet format of an LACP protocol provided by the prior art;

3 is a schematic diagram of system priority of an LACP protocol provided by the prior art;

4 is a schematic diagram of port preference levels of the LACP protocol provided by the prior art;

FIG. 5 is a flowchart of a method for a self-negotiating LACP protocol reference system and a reference port according to an embodiment of the present invention;

6 is a schematic diagram of an apparatus for a self-negotiating LACP protocol reference system and a reference port according to an embodiment of the present invention;

7 is a schematic diagram of an LACP protocol single device environment according to an embodiment of the present invention;

8 is a schematic diagram of a multi-device environment of an LACP protocol according to an embodiment of the present invention;

9 is a flowchart of a method for calculating a reference port of an LACP protocol according to an embodiment of the present invention;

FIG. 10 is a flowchart of a method for obtaining a reference port during operation of an LACP protocol according to an embodiment of the present invention.

Detailed ways

The preferred embodiments of the present invention are described in detail below with reference to the accompanying drawings.

FIG. 5 is a flowchart of a method for a self-negotiating LACP protocol reference system and a reference port according to an embodiment of the present invention. As shown in FIG. 5, the method includes the following steps:

Step S501: The first system and the second system of the link aggregation group of the LACP protocol perform real-time detection on the real-time resource usage status of the respective systems to determine the stability of the respective systems;

Step S502: The first system and the second system select, by negotiation, a system with the best stability in the first system and the second system as a reference system;

Step S503: The reference system selects one of the ports of the reference system according to the state history information of each port pair formed by each port and the corresponding peer system direct port and the resource usage status of the board. The port acts as a reference port.

The status history information of each port pair and the resource usage status of the port are as follows: the port state history information of each port and the resource usage status of the board to which the port belongs; the peer directly connected to each port System port port Status history information and resource usage status of the board to which the port belongs. Specifically, the resource usage status includes a board CPU utilization and a board memory usage rate; and the port status history information includes the number of times the port link is disconnected in the current time period.

The first system is configured to use a resource usage metric of each board of the first system according to a resource usage status of a card to which the port of each port of the system belongs; the second system is configured according to each port of the system. The resource usage metric of each board of the second system is configured, and the resource usage metric of each board of the first system includes: the first system The CPU usage of the board to which the port of the port belongs and the memory usage of the board. The resource usage metric of each board in the second system includes: the board of the port to which the port of the second system belongs. Board CPU utilization and board memory usage.

Determining the stability of the respective systems according to the present invention includes: selecting a maximum resource usage metric value as the stability of the first system among the resource usage metric values of each board of the first system; The resource usage metric of each board selects the largest resource usage metric as the stability of the second system; wherein the greater the resource usage metric, the worse the stability.

The port of each reference port of the reference system is constructed according to the resource usage status of the board to which the port of each port belongs and the resource usage status of the board to which the port of the peer system port directly connected to each port belongs. The metric of the resource usage of the board, including the CPU usage of the board to which the port of each port belongs and the memory usage of the board. The port of the peer system port directly connected to each port belongs to the board. Board CPU utilization and board memory usage. Specifically, the reference system selects the state history information of each port pair formed by each port of the port and the corresponding peer system directly connected port, and the resource usage status of the board to which the board belongs. A port as a reference port includes: sorting the resource usage metrics of the card to be used in a descending order, and sequentially using the port of the card as the port to be referenced; The number of times the port link is disconnected is detected. When the number of times the port link is disconnected in the current time period of the port to be referenced is 0, the port to be referenced is used as a reference port in the reference system.

FIG. 6 is a schematic diagram of a device for a self-negotiating LACP protocol reference system and a reference port according to an embodiment of the present invention. As shown in FIG. 6, the method includes: determining a stability module 601, selecting a reference system module 602, and selecting a reference port module 603. . Specifically, the determining stability module 601, the first system and the second system configured as the LACP protocol link aggregation group determine the stability of the respective systems by performing real-time detection on the real-time resource usage status of the respective systems; The reference system module 602 is configured to select, by the first system and the second system, a system with the best stability in the first system and the second system as a reference system; the selection reference port module 603 And the reference system is configured to select one of the ports of the reference system according to the state history information of each port pair formed by each port and the corresponding peer system direct port and the resource usage state of the board to which the reference system belongs. The port acts as a reference port.

The state history information of each port pair and the resource usage status of the card to be used in the present invention include: the port state history information of each port and the resource usage status of the board to which the port belongs; the peer directly connected to each port Port state history information of the system port and the resource usage status of the card to which the port belongs. The resource usage status includes a board CPU utilization rate and a board memory usage rate; the port status history information includes the number of times the port link is disconnected in the current time period.

The innovations of the present invention are mainly as follows: 1. The system priority is self-coordinated according to the real-time resource usage of the local and the peer system. 2. The port priority is adjusted according to the real-time resource usage and port state history information of the board where the local end and the peer port reside. The main functions of these two innovations are: the introduction of the system, the resources of the board, and the record of the port state history information, combined with the transmission of LACPDU packets between the two systems, and the auto-negotiation to adjust the system priority and port priority. To help improve the stability of the bearer service flow link. In addition to the above two points, the present invention also considers how the device of the present invention supports the peer device in the case where the device using the solution of the present invention is connected with the device not using the LACP protocol.

The technical solution of the present invention includes the following steps:

Step 1: The LACP protocol is run between system A and system B. The two ends are classified by ports, and the system resource status table and port abnormal status record table are established. The system priority and port priority are calculated according to the contents of the table. Level, and notify the peer through the LACPDU message;

Step 2: A and B decide which of the two is the reference system, and which port is used as the reference port;

Step 3: During the whole process of starting and running the LACP protocol between A and B, the above steps are performed periodically.

The present invention will be further described below with reference to Figs. 7 to 10.

The current system A and system B run the LACP protocol. From the system level, A and B can be a single device (as shown in Figure 7) or a multi-device aggregation group (Figure 8). From the port distribution level, run LACP. The ports of the protocol can be on different boards (such as port 3 and port 4 in Figure 8) or on the same board (such as port 1 and port 2 in Figure 8). The process is as follows: Take system A in Figure 8 as an example. Each board in system A obtains real-time resource information of the board (including but not limited to cpu utilization and memory usage), and establishes port-based system resources. The status table is shown in Table 1 below.

Table 1: Port-based system resource status table

If A is a single device (as shown in Figure 7) and the link aggregation is formed by n ports, the n ports are distributed on m boards (if m=1, as shown in Figure 7 board b; if m>1, such as In Figure 7, the board a, the m board vpu each a timing task (the timing interval can be manually set), calculate the real-time use of their respective resources, the following only cpu utilization C (m) and memory usage M (m) As an example, the resource usage metric R(m) of the board is defined as a weighted sum of C(m) and M(m) (the weighting parameter is set by the user); then the system resource state metric R of A ( A)=MAX(R(1), R(2),...,R(m)), that is, take the largest resource usage metric in several boards as the system resource status metric. The larger the value, the larger The system has a high probability of instability and is not suitable as a high priority system for LACP. If the default system priority of LACP is K, the system priority value of system A is P(A)=K*R(A). If P(B) calculated on System B is greater than P(A), the LACPDU packet is between A and B. Transfer (fill in the Actor_System_Priority field according to the system priority and forward it, as shown in Figure 2). The system of the LACP protocol stack at both ends of the system finds that the system priority of A is higher than the system priority of B. According to the LACP protocol, the priority value is smaller. The higher the system priority, Decision A is used as the reference system.

If A is a multi-device aggregation group (Figure 8) and consists of x devices, then A's system resource status metric R(A)=MAX(R(1), R(2),...,R(x) ), where R(x) is calculated in the same way as system A is a single device (in the case of Figure 7, R(A) = 70%), then system A's system priority value P(A) = K*R (A). If the P(B) calculated on System B is greater than P(A), and the LACPDU packet is transmitted between A and B, the LACP protocol stack of the system at both ends finds that the system priority of A is higher than the system priority of B. Decision A is used as a reference system.

Take system A in Figure 8 as an example. Each port in system A periodically collects port status exception information, and establishes an abnormal status record table for the port and the directly connected peer port, as shown in Table 2 below.

Table 2: Port and direct peer port abnormal status record table

The number of link down times of each port in the near ΔT time period (which can be set by the user) is set, and the value is sent to the system B through the first Reserved field of the LACPDU packet (see Table 3 for field definition). System B will also send its associated value to System A. In this way, even if any device restarts reset, the peer device has a link down count record of its near ΔT period. Generally, the number of link downs at both ends should be the same. The reason for this design is that the status of the ports on both ends is not synchronized.

Table 3: Byte information of the Reserved field

Among them, 0-3 bytes in the table are the device identification bits of the company, 4-7 bytes are cpu utilization, 8-11 bytes are memory usage, and 20-23 bytes are ports near ΔT time period. 12-19 bytes are other reserved fields, mainly for other bit resource information.

Assume that A has been used as the reference system. A establishes the system resource status table of each selected port and directly connected peer port, as shown in Table 4 below. Define the resource usage metric r(n) of the board to which the selected port and the directly connected peer port are equal to C(n), The weighted sum of M(n), C'(n) and M'(n) (the weighting parameters are set by the user), where C'(n) and M'(n) represent the direct connection of the peer port Board CPU utilization and memory usage. System B passes the Reserved field of the LACPDU packet. The location of the packet in the Reserved field is shown in Figure 2. The first Reserved field is reserved for the packet sending device. System B can fill in its own timing query here. The resource information, the value of the second Reserved field is filled with the first Reserved field information of the message sent by A to B, and the reserved 3-byte field information definition is as shown in FIG. 10, wherein the first 4 bytes define the present The unique byte information of the invented device is used to distinguish whether the peer device is a device adopting the solution of the present invention) to transmit real-time resource usage information to A. The r(n) value reflects the potential risk of the port processing traffic flow. The larger the r(n) value, the higher the potential risk.

Table 4: The system resource status table of the port and the directly connected peer port

The first 4 bytes of the first Reserved field of the LACPDU message sent by the A in the present invention are not the unique byte information defined by the method of the present invention, which means that the system B enables the standard LACP protocol or at least the system. If the A does not belong to the same manufacturer, the priority of the reference system and the reference port of the system A will not be dynamically adjusted, and will only be consistent with the initial value. This ensures that the system of the present invention is successfully connected to the LACP protocol of other vendors.

FIG. 9 is a flowchart of a method for calculating a reference port by using the LACP protocol according to an embodiment of the present invention. As shown in FIG. 9, in the selection of a reference port, the system periodically detects and records r(1), r(2), ..., r(n) is sorted from small to large, the port corresponding to the smallest value is used as the first candidate for the reference port, the port corresponding to the second smallest value is used as the second candidate for the reference port, and so on, and according to the port and The final ΔT time period of the direct connection peer port abnormal state record table is determined by the total number of link down times: 1. If r(n) is the first candidate of the reference port, and the number of link down times is N (N>0), then The first candidate is culled, the second candidate is selected, and the number of link down times N is determined. If N is equal to 0, it is used as a reference port. If N is greater than 0, the reference port third candidate is selected by analogy.

FIG. 10 is a flowchart of a method for obtaining a reference port during operation of an LACP protocol according to an embodiment of the present invention, as shown in FIG. 10, during the operation of the LACP protocol, considering that the selected reference port is close to the values of other selected ports, for example, port 1 is the reference port, but r(1) and r(2) are similar, and they are The number of link down times of the port is 0, which is easy for the two to take the reference port in turn, which is unfavorable for maintaining the stability of the traffic flow. The resource usage metric threshold Q of the port and the directly connected peer port of the present invention is introduced. The upper limit of the sum of the CPU usage and the memory usage of the local end and the peer end is 4, and the value of Q ranges from 0 to 4. If the port 1 is used as the reference port, it is judged whether r(1) is greater than Q in the next timing detection. If yes, step 4 is started to recalculate the reference port; if not, port 1 is still used as the reference port. .

In summary, the present invention has the following technical effects:

The present invention automatically adjusts the port priority according to the real-time resource usage of the local and the peer system, and adjusts the system priority according to the real-time resource usage and port state history information of the board where the local end and the peer port are located. Improve the stability of the bearer service flow link.

Although the invention has been described in detail above, the invention is not limited thereto, and various modifications may be made by those skilled in the art in accordance with the principles of the invention. Therefore, modifications made in accordance with the principles of the invention are to be understood as falling within the scope of the invention.

Industrial applicability

The technical solution of the embodiment of the present invention can be applied to the process of the self-negotiating LACP reference system and the reference port, and the recording of the resource and the port state history information of the system board is combined with the LACPDU packet between the two systems. Transmission, auto-negotiation adjusts system priority and port priority to help improve the stability of bearer service flow links.

Claims

A method for self-negotiating the LACP protocol reference system and reference port includes the following steps:

The first system and the second system of the LACP protocol link aggregation group determine the stability of the respective systems by performing real-time detection on the real-time resource usage status of the respective systems;

The first system and the second system select, by negotiation, a system with the best stability in the first system and the second system as a reference system;

The reference system selects a port from the reference port of the reference system as a reference according to the state history information of each port pair formed by each port and the corresponding peer system direct port and the resource usage status of the board. port;

The LACP protocol refers to a link aggregation control protocol.
The method according to claim 1, wherein the status history information of each port pair and the resource usage status of the associated board include:

Port state history information of each port and the resource usage status of the card to which the port belongs.

Port state history information of the peer system port directly connected to each port and the resource usage status of the board to which the port belongs.
The method according to claim 2, wherein the resource usage status comprises a board CPU utilization and a board memory usage rate; and the port status history information includes a number of times the port link is disconnected in the current time period.
The method according to claim 3, wherein the first system constructs a resource usage metric value of each board of the first system according to a resource usage state of a board to which a port of each port of the system belongs; The second system constructs a resource usage metric value of each board of the second system according to a resource usage status of the board to which the port of each port of the system belongs;

The resource usage metric of each board in the first system includes: a board CPU utilization rate and a board memory usage rate of the board to which the port of each port of the first system belongs; the second system The resource usage metrics of the boards include: the CPU usage of the board and the memory usage of the board of the board to which the port of each port of the second system belongs.
The method of claim 4 wherein said determining respective system stability comprises:

Selecting a maximum resource usage metric value as the stability of the first system in a resource usage metric value of each board of the first system;

Selecting a maximum resource usage metric value as the stability of the second system in a resource usage metric value of each board of the second system;

The greater the resource usage metric, the worse the stability.
The method of claim 3, wherein the resource usage status of the board to which the port of each port belongs and the resource usage status of the board to which the port of the peer system port directly connected to each port belongs is constructed. Reference system The resource usage metrics of each port on the card to which it belongs, including:

The CPU usage and board memory usage of the board of the card to which the port of each port belongs.

The CPU usage and board memory usage of the board of the board to which the port of the peer system port is directly connected to each port.
The method according to claim 6, wherein the reference system according to the status history information of each port pair formed by each port and the corresponding peer system direct port and the resource usage status of the belonging board are referenced from the reference. One of the ports of the system selected as the reference port includes:

You can sort the resource usage metrics of the card to the port to be referenced.

Detecting the number of times the port link is disconnected in the current time period of the port to be referenced;

When the number of times the port link is disconnected in the current time period of the port to be referenced is 0, the port to be referenced is used as a reference port in the reference system.
A self-negotiating LACP protocol reference system and reference port device, including:

Determining the stability module, the first system and the second system set to the LACP protocol link aggregation group determine the stability of the respective systems by performing real-time detection on the real-time resource usage status of the respective systems;

Selecting a reference system module, configured to select, by the first system and the second system, a system with the best stability in the first system and the second system as a reference system;

Selecting a reference port module, and setting the status history information of each port pair formed by the reference system according to each port and the corresponding peer system direct port, and the resource usage status of the board to which the reference system belongs, from the port to which the reference system belongs. Select a port as the reference port;

The LACP protocol refers to a link aggregation control protocol.
The device according to claim 8, wherein the status history information of each port pair and the resource usage status of the associated board include:

Port state history information of each port and the resource usage status of the card to which the port belongs.

Port state history information of the peer system port directly connected to each port and the resource usage status of the board to which the port belongs.
The device according to claim 9, wherein the resource usage status comprises a board CPU utilization and a board memory usage rate; and the port status history information includes a number of times the port link is disconnected in the current time period.