WO2017041577A1

WO2017041577A1 - Method and device for managing link aggregation member port packet loss

Info

Publication number: WO2017041577A1
Application number: PCT/CN2016/088709
Authority: WO
Inventors: 兰许昌
Original assignee: 中兴通讯股份有限公司
Priority date: 2015-09-09
Filing date: 2016-07-05
Publication date: 2017-03-16
Also published as: CN106533964A

Abstract

A method for managing link aggregation member port packet loss, comprising:　after enabling a bandwidth management function of dynamically managing a link aggregation member port, detecting a receiving or sending general packet loss rate of a member port in an aggregation link within a set time duration; if it is detected that the receiving general packet loss rate of the member port within the set time duration exceeds or equals to a set threshold, setting the state of a peer member port of the member port, the receiving general packet loss rate of which exceeds or equals to the set threshold, as an inactive state; and if it is detected that the sending general packet loss rate of the member port within the set time duration exceeds or equals to the set threshold, setting the state of the member port, the sending general packet loss rate of which exceeds or equals to the set threshold, as the inactive state. The technical solution significantly improves the quality of line transmission, and is of great significance to the business of large customers such as banks.

Description

Method and device for managing packet loss of link aggregation member port

Technical field

This document refers to, but is not limited to, the packet loss management technology in the field of data communication, and the packet loss management scheme involving the member ports of the link aggregation group.

Background technique

SMARTGROUP, also known as Link Aggregation, is a standard mechanism defined in IEEE 802.3ad. It refers to bundling multiple physical ports into one logical port to implement inbound/outbound traffic in each member port. Load balancing, the switch determines which member port the packet is sent to the peer switch based on the port load balancing policy configured by the user. When the switch detects that the link of one of the member ports is faulty, the switch stops sending packets on the port and recalculates the port that the packet is sent on the remaining link according to the load balancing policy. Calculate the packet sending port. Link aggregation is an important technology in terms of increasing link bandwidth, achieving link transmission flexibility and redundancy. SMARTGROUP is divided into manual, static mode and dynamic mode. Manual mode adds multiple physical ports to the SG (Smart Group) to form a logical port. The LACP protocol is not running. The static and dynamic modes are in compliance with the Link Aggregation Control Protocol (LACP). The application of the standard LAG (Link Aggregation Group) is shown in Figure 1.

Summary of the invention

The following is an overview of the topics detailed in this document. This Summary is not intended to limit the scope of the claims.

The embodiment of the invention provides a method and a device for managing packet loss on a link aggregation member port, which solves the problem that the port packet loss cannot be managed in the existing link aggregation.

An embodiment of the present invention provides a method for managing packet loss on a link aggregation member port, where the method includes:

After the dynamic management of the link aggregation member port is enabled, the device detects that the member ports in the aggregation link are in the set time. Rate or ordinary packet loss rate;

If it is detected that the normal packet loss rate of the member port is greater than or equal to the set threshold within the set duration, the state of the peer member port of the member port whose normal packet loss rate exceeds or equals the set threshold is set to Inactive state

If it is detected that the normal packet loss rate of the member port exceeds or is equal to the set threshold within the set duration, the state of the member port whose normal packet loss rate exceeds or equals the set threshold is set to an inactive state;

When the state of one or more member ports is set to an inactive state, according to the load sharing policy, the port for transmitting the message is recalculated in the remaining links except the active member ports;

When the inactive member port is in the set time period, the normal packet loss rate is lower than the set threshold, and the inactive member port of the member port in the inactive state is lower than the set packet loss rate. When the threshold is set, the status of the member port in the inactive state is updated to the active state, and the port on which the packet is sent is recalculated.

Optionally, if it is detected that the normal packet loss rate of the member port in the set duration is greater than or equal to the set threshold, the peer member port of the member port whose normal packet loss rate exceeds or equals the set threshold is received. The process of setting the status to inactive includes:

If the normal packet loss rate of the member port A is greater than or equal to the set threshold, the device corresponding to the member port B of the peer port A is notified by the link aggregation control protocol LACP. The device corresponding to the member port B sets the state of the member port B to an inactive state.

Optionally, if it is detected that the normal packet loss rate of the member port in the set duration is greater than or equal to the set threshold, the state of the member port with the normal packet loss rate exceeding or equal to the set threshold is set to a non- The process of active status includes:

If it is detected that the normal packet loss rate of the member port a within the set duration exceeds or is equal to the set threshold,

Sending information to the member port b of the local member port a and the member port b of the peer port a to notify the member port a that the normal packet loss rate exceeds or equals the set threshold.

The CPU of the member port a obtains the information that the member port a sends a normal packet loss rate that is greater than or equal to the set threshold, and sets the state of the member port a to the inactive state.

Optionally, when the inactive member port is in the set time period, the normal packet loss rate is lower than the set threshold, and the inactive member port of the inactive member port receives the normal packet loss within the set duration. When the rate is lower than the set threshold, the process of updating the status of the inactive member port to the active state includes:

When the member port A receives the normal packet loss rate within the set time period, the device corresponding to the member port B of the peer port of the member port A is notified to the member port A to receive the normal packet loss. Normal information;

After receiving the information that the member port A receives the normal packet loss, the CPU corresponding to the member port B detects that the normal packet loss rate of the member port B within the set duration is lower than the set threshold. The status of member port B is updated to the active state.

Optionally, when the member port in the inactive state has a normal packet loss rate within a set duration, the normal packet loss rate is lower than a set threshold, and the inactive member port has a lower packet loss rate within the set duration. When the threshold is reached, the process of updating the status of the inactive member port to the active state includes:

If the normal packet loss rate of the member port a is lower than the set threshold within the set duration, if it is determined that the peer member port b of the member port a is within the set duration, the normal packet loss rate is lower than the set. Threshold, the status of the member port a is updated to the active state.

Optionally, the method further includes:

Check whether there are member ports in the link aggregation port that receive bandwidth loss or excessive bandwidth loss.

If it is detected that the member port X receives the bandwidth loss packet, the information about the packet loss of the member port X is notified to the member port X of the local port and the member port Y of the member port X of the member port X. If the member port Y is the destination port of any service redirection, the service redirected to the member port Y is redirected to another port with low bandwidth utilization.

Optionally, after the service redirected to the member port Y is redirected to the other port with the lowest bandwidth utilization, if the member port X still receives the bandwidth loss packet, the method further includes:

The member port Y uses the hash key as a feature value to redirect the service access control list ACL redirected to the member port Y to the port with the lowest bandwidth utilization, and repeats the redirection action until the member port X no longer receives excess bandwidth loss.

Optionally, the method further includes:

If the member port x is sent out of the bandwidth loss, the information about the packet loss of the local end is reported to the CPU of the local member port x and the CPU of the member port y of the peer end; the CPU of the member port x is obtained. After the information about the bandwidth loss packet is sent to the local end, if the member port x is determined to be the destination port of any service redirection, the service redirected to the member port x is redirected to the other bandwidth utilization. port.

Optionally, after the service redirected to the member port x is redirected to the other port with the lowest bandwidth utilization, if the member port x still sends an excessive bandwidth packet loss, the method further includes:

The member port x is characterized by a hash key, and the service redirected to the member port x is redirected to the port with the lowest bandwidth utilization, and the redirection action is repeated until the member port x is no longer used. Send over bandwidth loss.

Optionally, the method further includes:

If the current mode is the manual link aggregation mode, if the member port of the link aggregation port is detected to be longer than or equal to the set threshold, the member port with the normal packet loss rate exceeding or equal to the set threshold will be received. The status is set to inactive, and the port for sending packets is recalculated in the remaining links except the member ports that have been set to inactive according to the load sharing policy until the detected member ports are within the set duration. If the packet loss rate is lower than the set threshold, the status of the member port is updated to the active state, and the packet sending port is recalculated again;

If it is detected that the member port in the link aggregation port is greater than or equal to the set threshold within the set duration, the state of the member port whose normal packet loss rate exceeds or equals the set threshold is set to an inactive state; According to the load sharing policy, the port for sending packets is recalculated in the remaining links except the member ports that have been set to inactive, until the detected member ports lose the packet loss rate within the set duration. Threshold, then update the status of the member port to the active state, and recalculate the message sending port again. The embodiment of the invention further provides an apparatus for managing packet loss of a link aggregation member port, including:

The first detecting unit is configured to detect the normal loss of the member ports in the aggregated link within the set duration when the current mode is static or dynamic link aggregation mode after the dynamic management link aggregation member port bandwidth management function is enabled. Packet rate or ordinary packet loss rate;

The state setting unit is configured to: when detecting that the normal packet loss rate of the member port within the set duration is greater than or equal to the set threshold, the peer packet with the normal packet loss rate exceeding or equal to the set threshold is received. The status is set to inactive; it is also set to detect that the member port is set for the duration When the normal packet loss rate is greater than or equal to the set threshold, the state of the member port whose normal packet loss rate exceeds or equals the set threshold is set to an inactive state; and the member that is also set to be inactive is also set. If the normal packet loss rate of the port is lower than the set threshold within the set duration, and the peer member port of the inactive member port is lower than the set threshold within the set duration, the The status of the inactive member port is updated to the active state;

a load sharing unit configured to recalculate a port for sending a message in a remaining link other than the active member port according to a load sharing policy when the state of one or more member ports is set to an inactive state; When the inactive member port is updated to the active state, the port on which the message is sent is recalculated.

Optionally, the state setting unit is configured to: when detecting that the normal packet loss rate of the member port exceeds or exceeds a set threshold within a set duration, the normal packet loss rate is greater than or equal to the Set the status of the peer end of the member port of the threshold to inactive:

If it is detected that the normal packet loss rate of the member port A is greater than or equal to the set threshold, the device corresponding to the member port B of the peer port A is notified by the link aggregation control protocol LACP to indicate The device corresponding to the member port B of the peer end sets the state of the peer member port B to an inactive state.

Optionally, the state setting unit is configured to: when detecting that the normal packet loss rate of the member port exceeds or exceeds a set threshold within a set duration, the normal packet loss rate is greater than or equal to the The status of the member port that sets the threshold is set to inactive:

If it is detected that the normal packet loss rate of the member port a is greater than or equal to the set threshold within the set duration, respectively, the CPU and the member port of the member port a that sends the normal packet loss rate exceeding or equal to the set threshold are detected respectively. The member port b of the peer end of a sends a message informing the member port a that the normal packet loss rate exceeds or equals the set threshold, and sets the state of the member port a to the inactive state.

Optionally, the state setting unit is configured to: when the member port in the inactive state is lower than the set threshold in the set duration, and the peer member of the member port in the inactive state is configured by: When the normal packet loss rate of the port within the set duration is lower than the set threshold, the status of the inactive member port is updated to the active state:

The device corresponding to the member port B of the peer end of the member port A is notified to the member port A when the normal packet loss rate of the member port A in the inactive state is lower than the set threshold. The information about the normal packet loss is normal. The device corresponding to the member port B receives the member port A. After receiving the normal packet loss information, if it is detected that the normal packet loss rate of the peer member port B within the set duration is also lower than the set threshold, the status of the member port B is updated to the active state.

When it is detected that the normal packet loss rate of the member port a within the set duration is lower than the set threshold, the number of the normal packet loss rate of the peer member port b that determines the member port a is within the set duration. When the threshold is set, the status of the member port a is updated to the active state. Optionally, the device further includes:

The second detecting unit is configured to detect whether a member port of the link aggregation port receives an excess bandwidth packet loss or sends an excessive bandwidth packet loss;

The ultra-bandwidth packet loss processing unit is configured to notify the member port Y of the member port X of the member port X that the member port is over-bandwidthed by the LACP when the member port X is detected to receive the bandwidth loss packet. The CPU is further configured to: when the CPU corresponding to the member port Y of the peer end receives the information of the packet loss of the member port X, if it is determined that the member port Y is the destination port of any service redirection, Traffic redirected to the member port Y is redirected to the other port with the lowest bandwidth utilization.

Optionally, the ultra-bandwidth packet loss processing unit is further configured to redirect the traffic redirected to the member port Y to another port with the lowest bandwidth utilization, and if the member port X still receives the bandwidth loss packet Then, the hash key of the member port Y is used as a feature value, and the service access control list ACL redirected to the member port Y is redirected to the port with the lowest bandwidth utilization, and the member is repeatedly redirected until the member is redirected. Port X no longer receives bandwidth loss.

Optionally, the ultra-bandwidth packet loss processing unit is further configured to: notify the CPU of the local member port x by using LACP to detect the packet loss of the local end by using the LACP when detecting that the member port x sends the ultra-bandwidth packet loss. And the CPU of the member port y of the peer port of the member port x, and the CPU of the local member port x obtain the information of the excessive bandwidth loss packet of the local end, and if it is determined that the member port x is any service weight at the same time The directed destination port redirects traffic redirected to the member port x to other ports with low bandwidth utilization.

Optionally, the ultra-bandwidth packet loss processing unit is further configured to redirect to the member port If the service on the member port is redirected to another port with low bandwidth utilization, if the member port x still sends over-bandwidth packet loss, the hash value of the member port x is used as the eigenvalue and will be redirected to the device. The service on member port x is redirected to the port with the lowest bandwidth utilization. The redirection action is repeated until the member port no longer sends excessive bandwidth loss.

Optionally, the first detecting unit is further configured to: when the current mode is the manual link aggregation mode, detect that the member port in the link aggregation port receives the normal packet loss rate and/or sends the normal lost packet within the set duration. Packet rate

The state setting unit is further configured to: when the current mode is the manual link aggregation mode, if the normal packet loss rate of the member port in the link aggregation port is longer than or equal to the set threshold, The status of the member port whose normal packet loss rate is greater than or equal to the set threshold is set to inactive. It is also set to detect the member port of the link aggregation port when the current mode is manual link aggregation mode. When the normal packet loss rate is greater than or equal to the set threshold, the state of the member port whose normal packet loss rate exceeds or equals the set threshold is set to an inactive state.

The embodiment of the invention is a computer storage medium, wherein the computer storage medium stores computer executable instructions, and when the computer executable instructions are executed, a method for managing packet loss of a link aggregation member port is implemented. The above technical solution has the following advantages:

1. For the case that the link has a small amount or even a large number of ordinary packet loss, the above technical solution provides a corresponding management method, which greatly improves the quality of the line transmission, and is of great significance to a large customer service such as a bank.

2. For the case of ultra-bandwidth traffic, the above technical solution improves the traffic redistribution capability and greatly improves the bandwidth utilization. If the number of smartgroup members is k, the bandwidth carried by member ports is g. According to the technical solution of the present application, the range of the bandwidth r to be carried is preferably g=<r<=(k-1)*g.

Other aspects will be apparent upon reading and understanding the drawings and detailed description.

DRAWINGS

FIG. 1 is a schematic diagram of an existing standard LAG application networking;

2 is a flowchart of a manual LAG packet loss limit processing in the first embodiment of the present invention;

3 is a flowchart of a process for processing a static or dynamic LAG to receive an ordinary packet loss in the first embodiment of the present invention;

4 is a flowchart of a static or dynamic LAG sending ordinary packet loss limit processing in the first embodiment of the present invention;

5 is a flowchart of processing a static or dynamic LAG over-bandwidth limit in the first embodiment of the present invention;

6 is a flowchart of processing a static or dynamic LAG transmission overband limit in the first embodiment of the present invention;

7 is a complete flowchart of managing packet loss processing of a smartgroup member port according to Embodiment 1 of the present invention;

FIG. 8 is a schematic structural diagram of an apparatus for managing packet loss of a link aggregation member port in the second embodiment;

FIG. 9 is another schematic structural diagram of an apparatus for managing packet loss of a link aggregation member port in the second embodiment of the present invention.

detailed description

The technical solutions of the embodiments of the present invention are further described in detail below with reference to the accompanying drawings and specific embodiments. It should be noted that the features in the embodiments and the embodiments herein may be combined with each other arbitrarily without conflict.

Embodiment 1

As can be seen from the standard LAG application shown in Figure 1, the standard LAG application method has its own drawbacks as follows:

1) When the switch detects that the link of one of the member ports is faulty, it stops sending packets on the port and recalculates the port for sending packets on the remaining link according to the load sharing policy. After the faulty port is restored, Recalculate the message sending port again. The link failure mentioned here generally refers to link down (Link Down) or other serious link failures that can cause Link Down; obviously, this is a strict requirement. If there is packet loss in the link, it is impossible to switch in time to ensure the link quality.

2) For standard LAG applications, you can require that the bandwidth carried should not exceed the bandwidth of one of the member ports. When the feature data is special, all traffic is calculated to the same member port. If the bandwidth of the bearer exceeds the bandwidth of the member port, traffic loss will obviously occur. This is obviously great. Wasting bandwidth.

In order to overcome the above-mentioned deficiencies, the present embodiment provides a method for managing packet loss on a smartgroup member port, which can be applied to a real network of a smartgroup to improve the transmission quality and bandwidth utilization of the smartgroup. The method mainly includes the following operations:

In the manual smartgroup mode, when the duration T1 of any port receiving/receiving normal packet loss exceeds the set packet loss threshold H1 (including CRC and other non-overbandwidth packet loss), the member port status is set to inactive. (inactive). Of course, for a link that strictly requires quality, once the packet loss is detected, the port member status can be set to inactive. When the member port is set to inactive, the device stops sending packets on the port and recalculates the port for sending packets based on the load sharing policy. When the member port is inactive, the port detection function is enabled. When the local port receives/transmits for a certain period of time, T2 continues to not drop packets or does not exceed the set packet loss threshold H1, and then resets the member port state to the active state (actvie). After re-calculating the packet sending port, the process is as shown in Figure 2, including the processing when the manual LAG is received, the normal packet loss is exceeded, and the manual LAG packet loss disappears.

For the static or dynamic smartgroup mode, the methods for managing packet loss on smartgroup member ports include the following operations:

Step 10: After the dynamic management of the link aggregation member port bandwidth management function is enabled, when the current mode is static or dynamic link aggregation mode, the normal packet loss rate of the member ports in the aggregation link is set or sent. Packet loss rate;

Step 11: If it is detected that the normal packet loss rate of the member port in the set duration is greater than or equal to the set threshold, the peer member port of the member port whose normal packet loss rate exceeds or equals the set threshold is received. The status is set to inactive;

Step 12: If it is detected that the normal packet loss rate of the member port exceeds or is equal to the set threshold within the set duration, the state of the member port with the normal packet loss rate exceeding or equal to the set threshold is set to be inactive. status;

Step 13: When the state of one or more member ports is set to an inactive state, according to the load sharing policy, the port for sending the message is recalculated in the remaining links except the member ports in the active state;

Step 14: When the inactive member port is within the set duration, the normal packet loss rate is lower than the set threshold, and the inactive member port of the inactive member port is lost within the set duration. When the packet rate is lower than the set threshold, the status of the member port in the inactive state is updated to the active state, and the port for sending the packet is recalculated.

It should be noted that, in the embodiment of the present invention, the execution sequence of step 14 of step 11 is not limited, and the corresponding steps are performed according to the result of the detection.

Optionally, on the basis of the foregoing method, the problem of transmitting and receiving ultra-bandwidth packet loss is also considered.

For static and dynamic smartgroup methods, the LACP protocol can be used. The LACP packet has three reserved bytes, Reserved[0]~Reserved[2]. Therefore, in the implementation process, ordinary packet loss alarms and ultra-bandwidth alarms can be transmitted by using bits 0 to 3 of Reserved[0]. as follows:

1) The local receiving normal packet loss rate exceeds the set threshold (including CRC and other non-overbandwidth packet loss)--coded as 0001;

2) The normal packet loss rate of the peer end exceeds the set threshold (including CRC and other non-overbandwidth packet loss)--coded as 0010;

3) The local packet loss rate exceeds the set threshold (including CRC and other non-overbandwidth packet loss)--coded as 0011;

4) The normal packet rate of the peer end exceeds the set threshold (including CRC and other non-super bandwidth packet loss)--coded to 0100;

5) The local end receives bandwidth loss packet -------- code is 0101;

6) The peer end receives more than the bandwidth packet loss ---- encoding is 0110;

7) The local end sends a bandwidth loss packet -------- coded to 0111;

8) The end-to-end transmission exceeds the bandwidth loss-----coded to 1000.

With the above definition, the next step is to set the behavior of the LACP protocol when receiving various encodings:

A device (such as a switch) detects each member port of the device. When the device detects that the packet loss rate of any member port of the device exceeds the threshold Rx1, the device generates the corresponding port A. The device (the LACP packet can be sent to the CPU of the device, and the LACP packet is described as being sent to the CPU) is in the LACP packet. The bits 0 to 3 of Reserved[0] are set to 0001 to the peer end of the member port. Set the bit 0 to 3 of Reserved[0] of the LACP packet sent by the CPU of member port B to 0010. After receiving the packet with Reserved[0]bit0~3 as 0010, the CPU of the peer member port B sets the state of the peer member port B to inactive. After the peer member port B is set to inactive, the device stops sending packets on the port B and is based on the load sharing policy. In the remaining link, the port for packet transmission is recalculated, and port detection is enabled. If the normal packet loss rate of the peer member port B continues to exceed the threshold Rx1, the inactive state of the peer member port B is maintained. If the peer member port B receives the packet loss during the T1 period. If the rate is lower than the set threshold Rx2, the bits 0 to 3 of the Reserved[0] in the LACP packet sent to the member port B are set to 0000, and the reserved LACP message sent to the CPU of the member port A is Reserved[0]. Bits 0 to 3 are set to 0000. After receiving the message of Reserved[0]bit0~3 to 0000, the CPU of member port A will update member port B if it is determined that the common rate of member port A within the set duration is also lower than the set threshold Tx2. In the active state, the packet sending port is recalculated again, and the port detection function is disabled. This process is shown in Figure 3. In this way, it has great significance for improving the quality of the link. When it is detected that the normal packet loss of any member port a exceeds the set threshold Tx1, the bits 0 to 3 of the Reserved[0] in the CPULACP packet corresponding to the member port a are set to 0011 to the peer member port. The corresponding CPU sends the bit 0 to 3 of Reserved[0] in the LACP packet to 0100. The CPU of member port a sets the member port a state to inactive after receiving the message with Reserved[0]bit0~3 being 0011. After the member port a is set to inactive, the device stops sending packets on the port a, and recalculates the port for sending packets on the remaining links according to the load sharing policy. At the same time, port detection is enabled. If the normal packet loss exceeds the set threshold Tx1 in the T2, the inactive state of the member port a is maintained. If the normal packet loss of the member port a is lower than the set threshold Tx2 in the T2 time, the member port is sent. Bits 0 to 3 of the LACP packet Reserved[0] of the CPU of a are set to 0000, and the bits 0 to 3 of the LACP packet Reserved[0] are set to 0000 to the CPU of the peer member port b. After receiving the message of Reserved[0]bit0~3 to 0000, the CPU of member port a determines that the normal rate of the peer member port b of the member port a is less than the set threshold Rx2. Then, the status of member port a is updated to the active state, and the packet sending port is recalculated again. At the same time, the port detection function is disabled. This process is shown in Figure 4. In this way, it has great significance for improving the quality of the link.

It should be noted that the device corresponding to the member port can be a switch, and each switch can have multiple ports. For example, switch 1 has ports A, B, and C, and switch 2 has ports a, b, and c. The peer port is b. When switch 1 detects that the normal packet loss rate of port A is exceeded, switch 1 reports LACP packets to the CPU of switch 1. The bits 0 to 3 of Reserved[0] are set to 0001, and switch 1 is in the opposite direction. The CPU of the switch 2 corresponding to the port b sends an LACP packet. The bits 0 to 3 of the Reserved[0] of the LACP packet are set to 0010. After receiving the packet, Switch 2 sets the state of port b to inactive.

When any member port X is detected to receive bandwidth loss, the bits 0 to 3 of Reserved[0] in the LACP packet sent to the member port X are set to 0101. Bits 0 to 3 of Reserved[0] in the LACP packet of the CPU corresponding to the peer member port Y of the member port X are set to 0110. After receiving the message with Reserved[0] of 0110, the CPU of the peer member port Y first redirects these services to other bandwidth utilizations if it determines that the peer member port Y is the destination port of any service redirection. The lowest port, if the peer member port Y still has the bandwidth loss packet, the hash key is used as the eigenvalue (classid), and the ACL (Acess Control List) is used to redirect the bandwidth. The port with the lowest utilization rate repeats the redirection action until the ultra-bandwidth packet loss disappears. This process is shown in Figure 5.

When any member port x is detected to send an excessive bandwidth packet, the packets of the CPU corresponding to the member port x are set to 0-11 of the reserved message[0], and are sent to the peer member port of the member port x. Bits 0 to 3 of the LACP message Reserved[0] of the CPU corresponding to y are set to 1000. If the member port 寇x receives the message with Reserved[0] of 0111, if it confirms that member port x is the destination port of any service redirection, it first redirects these services to other ports with the lowest bandwidth utilization. If the member port x still has a bandwidth loss packet, the hash key is used as the classid, and the ACL is directed to the port with the lowest bandwidth utilization, and the redirection action is repeated until the ultra-bandwidth packet loss disappears. This process is shown in Figure 6.

The complete implementation process of the above method is described in detail below. As shown in FIG. 7, the following operations are included:

Step 701: The version is started normally. After the smartgroup is configured, the dynamic management smartgroup bandwidth function is enabled. If it is manual smartgroup mode, go to step 702; if it is static or / dynamic smartgroup mode, go to step 705;

Step 702: Detecting the normal packet loss rate or the normal packet loss rate in the member port duration T1; when it is detected that the normal packet loss rate in any member port duration T1 exceeds the specified threshold H1, then step 703 is performed; When it is detected that the normal packet loss rate in any member port duration T1 exceeds the specified threshold H2, then step 704 is performed;

Step 703: Set the state of the member port to the inactive state, and enable the port packet loss detection function, and recalculate the port for sending the packet in the remaining link according to the load sharing policy. If the normal packet loss in the member port port duration T2 does not exceed the specified threshold H3, the The status update of the member port is set to the active state, and the packet sending port is recalculated again. Disable port packet loss detection. Go to step S712;

Step 704: Set the state of the member port to the inactive state, and enable the port packet loss detection function, and recalculate the port for sending the packet in the remaining link according to the load sharing policy. If the normal packet loss in the member port port duration T2 does not exceed the specified threshold H4, the status update of the member port is set to the active state, and the packet sending port is recalculated again. Disable port packet loss detection. Go to step S712;

In step 705, it is detected whether the normal packet loss rate and/or the normal packet loss rate are detected, or whether the over-bandwidth packet loss and/or the ultra-bandwidth packet loss are detected. If it is detected that the normal packet loss rate and/or the normal packet loss rate is further limited to step 706; if it is detected, the over bandwidth loss packet and/or the ultra bandwidth loss packet is transferred to step 709;

In step 706, the normal packet loss rate is exceeded or the normal packet loss rate is exceeded. If the normal packet loss rate is exceeded, the process proceeds to step 707; if the normal packet loss rate is exceeded, the process proceeds to step 708.

Step 707: If it is detected that the normal packet loss rate of the member port A in the duration T1 exceeds the specified threshold H1, the bits 0 to 3 of the Reserved[0] in the LACP packet sent to the member port A are set to 0001. Bits 0 to 3 of Reserved[0] in the LACP packets sent by the peer member port B of member port A are set to 0010. The CPU of the peer member port B sets the state of the peer member port B to inactive after receiving the message with Reserved[0]bit0~3 being 0010. After the peer member port B is set to inactive, the device stops sending packets on the peer member port B, and recalculates the port for sending packets on the remaining links according to the load balancing policy. . If the normal packet loss rate of the peer member port B exceeds the threshold in the T1, the inactive state of the peer member port B is maintained. If the peer member port B receives the normal packet loss rate during the T1 time. If the value is lower than the set threshold, the bits 0 to 3 of Reserved[0] in the LACP packets sent to the CPU of the peer port B are set to 0000, and the LACP packets sent by the CPU of the member port A are Reserved[0. Bits 0 to 3 of the ] are set to 0000. After receiving the message of Reserved[0]bit0~3 to 0000, the CPU of the peer member port B determines that the normal packet loss rate of the member port A within the set duration is lower than the set threshold. The status of the member port B is updated to the active state. The packet sending port is recalculated and the port detection function is disabled. Go to step 712;

Step 708: If it is detected that the normal packet loss rate of the member port a in the duration T2 exceeds the specified threshold H2, the bits 0 to 3 of the Reserved[0] in the LACP packet sent to the CPU of the member port a are set to 0011, bit 0 to 3 of Reserved[0] in the LACP packet sent to the peer member port b. Set to 0100. The CPU of the member port a sets the status of the member port a to inactive after receiving the message with Reserved[0]bit0~3 being 0011. After the member port a is set to inactive, the device stops sending packets on the member port a, and recalculates the port for sending packets on the remaining links according to the load balancing policy. If the normal packet loss of the member port a in the T2 time continues to exceed the threshold Tx1, the inactive state of the member port a is maintained; if the member port a is in the duration T2, the normal packet loss is lower than the setting. If the threshold is H2, the bits 0 to 3 of Reserved[0] in the LACP packet sent to the member port a are set to 0000, and the bits 0 to 3 of Reserved[0] in the LACP packet sent to the peer member port b are set. It is 0000. After receiving the message of Reserved[0]bit0~3 to 0000, the CPU of the member port a determines that the normal packet loss rate of the peer member port b within the set duration is also lower than the set threshold. The status of port a is updated to the active state. The packet sending port is recalculated again. At the same time, the port detection function is disabled. Go to step 712;

In step 709, the over-limit of the bandwidth loss packet or the over-bandwidth packet loss exceeds the limit. If the packet loss exceeds the limit, the process proceeds to step 710; if the packet loss exceeds the limit, the process proceeds to step 711;

Step 710: When it is detected that any member port X receives the bandwidth loss packet, the bit 0 to 3 of Reserved[0] in the LACP packet sent to the member port x is set to 0101, and the pair is sent to the member port X. Bits 0 to 3 of Reserved[0] in the LACP packet of the CPU of the member port Y are set to 0110. After receiving the message with Reserved[0] of 0110, the CPU of the peer member port Y first redirects these services to other bandwidths if it determines that the peer member port Y is the destination port of any service redirection. If the peer port member port Y still has ultra-bandwidth packet loss, the hash key is used as the classid, and the ACL is used to redirect to the port with the lowest bandwidth utilization. The redirection action is repeated until the ultra-bandwidth packet loss disappears. Go to step 712;

Step 711: When it is detected that any member port x sends an excessive bandwidth packet loss, the packets sent from the CPU of the member port x to the reserved port 0 of the reserved port 0 are set to 0111, and are sent to the opposite end of the member port x. Bits 0 to 3 of Reserved[0] in the LACP packet of member port y are set to 1000. If the member port x receives the message with the reserved [0] of 0111, if the member port x is determined to be the destination port of any service redirection, the CPU is redirected to the other port with the lowest bandwidth utilization. If the member port x still has an ultra-bandwidth packet loss, the hash key is used as the classid, and the ACL is redirected to the port with the lowest bandwidth utilization. The redirection action is repeated until the ultra-bandwidth packet loss disappears.

Step S712: End.

The above technical solution provides a corresponding management method for a case where a link has a small amount or even a large number of ordinary packet loss, which greatly improves the quality of the line transmission, and is of great significance to a large customer service such as a bank. At the same time, for the case of ultra-bandwidth traffic, the traffic redistribution capability is improved, and the bandwidth utilization is greatly improved. If the number of smartgroup members is k, the bandwidth carried by member ports is g. According to the technical solution of the embodiment of the present invention, the bandwidth r to be carried may be in the range of g=<r<=(k-1)*g, where k is a positive integer greater than or equal to 1.

Embodiment 2

As shown in FIG. 8, the embodiment provides a device for managing packet loss of a link aggregation member port, and the method of the first embodiment is implemented, and at least includes the following units.

The first detecting unit 11 is configured to detect that the member ports in the aggregation link are within the set duration when the current mode is static or dynamic link aggregation mode after the dynamic management link aggregation member port bandwidth management function is enabled. Packet loss rate or ordinary packet loss rate;

The state setting unit 12 is configured to: if it is detected that the normal packet loss rate of the member port within the set duration is greater than or equal to the set threshold, the state of the peer end of the member port whose normal packet loss rate exceeds or equals the threshold is received. Set to inactive state; if it is detected that the normal packet loss rate of the member port within the set duration is greater than or equal to the set threshold, then the member port with the normal packet loss rate exceeding or equal to the threshold is sent. The status is set to inactive; and the member port that is inactive is set to a lower than the set threshold within the set duration, and the peer member port of the inactive member port is set. When the normal packet loss rate in the duration is lower than the set threshold, the status of the member port in the inactive state is updated to the active state;

The load sharing unit 13 is configured to recalculate the port for sending the message in the remaining links except the active member port according to the load sharing policy when the state of the one or more member ports is set to the inactive state; Recalculates the port on which the message is sent when the inactive member port is updated to the active state.

Optionally, the state setting unit 12 implements, if it is detected that the normal packet loss rate of the member port within the set duration is greater than or equal to the set threshold, the normal packet loss rate is greater than or equal to the The status of the peer end of the member port of the threshold is set to inactive:

Optionally, the state setting unit 12 implements the following manner: if it is detected that the normal packet loss rate of the member port exceeds or is equal to the set threshold within the set duration, the normal packet loss rate is greater than or equal to the threshold. The status of the member port is set to inactive:

If it is detected that the normal packet loss rate of the member port a is greater than or equal to the set threshold within the set duration, respectively, the CPU and the member port of the member port a that sends the normal packet loss rate exceeding or equal to the set threshold are detected respectively. The CPU of the peer member port of a sends a message to the member port to notify the member port that the normal packet loss rate exceeds or equals the set threshold, and detects that the normal packet loss rate exceeds or equals the set threshold. The status is set to inactive.

Optionally, the state setting unit 12 implements, by using the following manner, that the normal packet loss rate of the member port in the inactive state is lower than the set threshold within the set duration, and the peer member port of the member port in the inactive state is configured. When the normal packet loss rate within the set duration is lower than the set threshold, the status of the inactive member port is updated to the active state:

When the normal packet loss rate of the member port A in the inactive state is lower than the set threshold, the CPU notification corresponding to the member port B of the peer port of the member port A of the member port is respectively notified. If the member port A sends the normal packet loss information, the CPU corresponding to the member port B receives the information that the member port A receives the normal packet loss. If the normal packet loss rate is lower than the set threshold, the status of the member port B is updated to the active state.

When it is detected that the normal packet loss rate of the member port a within the set duration is lower than the set threshold, the number of the normal packet loss rate of the peer member port b that determines the member port a is within the set duration. When the threshold is set, the status of the member port a is updated to the active state. Optionally, in addition to considering the problem of receiving and sending ordinary packet loss, the preferred solution of this embodiment also considers that the member port receives and sends. The problem of ultra-bandwidth packet loss.

As shown in FIG. 9, the device further includes:

The second detecting unit 14 is configured to detect whether a member port of the link aggregation port receives an excess bandwidth packet loss or sends an excessive bandwidth packet loss;

The ultra-bandwidth packet loss processing unit 15 is configured to notify the member port Y of the member port X of the member port X that the member port is over-bandwidthed by the LACP when the member port X is detected to receive the bandwidth loss packet. If the device corresponding to the member port Y receives the packet loss information of the member port X, if it is determined that the member port Y is the destination port of any service redirection at the same time, The traffic redirected to the member port Y is redirected to the other port with the lowest bandwidth utilization.

Optionally, the ultra-bandwidth packet loss processing unit 15 is further configured to redirect the traffic redirected to the member port Y to another port with the lowest bandwidth utilization, if the member port X still receives the bandwidth. If the packet is lost, the hash key of the member port Y, such as the VLAN, is used as the eigenvalue (class id), and the ACL redirected to the member port Y is redirected to the bandwidth. The port with the lowest utilization rate repeats the redirection action until the member port X no longer receives the excess bandwidth packet loss.

Optionally, the ultra-bandwidth packet loss processing unit 15 is further configured to: notify the local member port x of the packet loss of the local end by using the LACP when detecting that the member port x sends the ultra-bandwidth packet loss. The CPU and the CPU of the member port y of the peer port of the member port x and the CPU of the member port x obtain the information of the excessive bandwidth loss packet of the local end, and if the member port x is determined to be any service weight at the same time, The directed destination port redirects traffic redirected to the member port x to other ports with low bandwidth utilization.

Optionally, the ultra-bandwidth packet loss processing unit 15 is further configured to redirect the service redirected to the member port x member port to another port with low bandwidth utilization, if the member port x still exceeds If the bandwidth is lost, the hash value of the member port x is used as the eigenvalue, and the service redirected to the member port x is redirected to the port with the lowest bandwidth utilization, and the redirection action is repeated until the packet is used. Member ports no longer send excessive bandwidth drops.

Optionally, the first detecting unit 11 is further configured to: when the current mode is the manual link aggregation mode, detect that the member port in the link aggregation port receives an ordinary packet loss rate or sends a normal packet loss within a set time period. Rate

The state setting unit 12 is further configured to: when the current mode is the manual link aggregation mode, if the normal packet loss rate of the member port in the link aggregation port is greater than or equal to the set threshold within the set duration, The status of the member port whose normal packet loss rate is greater than or equal to the set threshold is set to inactive. If the current mode is manual link aggregation mode, the member port of the link aggregation port is set. When the normal packet loss rate in the duration exceeds or equals the set threshold, the state of the member port with the normal packet loss rate exceeding or equal to the set threshold is set to the inactive state.

The embodiment of the present invention further provides a computer storage medium, where the computer storage medium stores computer executable instructions, and when the computer executable instructions are executed, a method for managing packet loss of a link aggregation member port is implemented.

One of ordinary skill in the art will appreciate that all or a portion of the above steps may be performed by a program to instruct related hardware, such as a processor, which may be stored in a computer readable storage medium, such as a read only memory, disk or optical disk. Wait. Alternatively, all or part of the steps of the above embodiments may also be implemented using one or more integrated circuits. Correspondingly, each module/unit in the above embodiment may be implemented in the form of hardware, for example, by implementing an integrated circuit to implement its corresponding function, or may be implemented in the form of a software function module, for example, executing a program stored in the memory by a processor. / instruction to achieve its corresponding function. This application is not limited to any specific combination of hardware and software. A person skilled in the art should understand that the technical solutions of the present application can be modified or equivalent, without departing from the spirit and scope of the technical solutions of the present application, and should be included in the scope of the claims of the present application.

Industrial applicability

The above technical solution improves the quality of the line transmission and the traffic redistribution capability, and improves the bandwidth utilization.

Claims

A method for managing packet loss on a link aggregation member port includes:

After the dynamic management of the link aggregation member port is enabled, the device detects the normal packet loss rate or the normal packet loss rate of the member ports in the aggregation link in the static mode or the dynamic link aggregation mode. ;

If it is detected that the normal packet loss rate of the member port is greater than or equal to the set threshold within the set duration, the state of the peer member port of the member port whose normal packet loss rate exceeds or equals the set threshold is set to Inactive state

If it is detected that the normal packet loss rate of the member port exceeds or is equal to the set threshold within the set duration, the state of the member port whose normal packet loss rate exceeds or equals the set threshold is set to an inactive state;

When the state of one or more member ports is set to an inactive state, according to the load sharing policy, the port for transmitting the message is recalculated in the remaining links except the active member ports;

When the inactive member port is in the set time period, the normal packet loss rate is lower than the set threshold, and the inactive member port of the member port in the inactive state is lower than the set packet loss rate. When the threshold is set, the status of the member port in the inactive state is updated to the active state, and the port on which the packet is sent is recalculated.
The method of claim 1, wherein if the normal packet loss rate of the member port within the set duration is greater than or equal to the set threshold, the normal packet loss rate is greater than or equal to the set threshold. The process of setting the status of the peer member port of a member port to inactive includes:

If the normal packet loss rate of the member port A is greater than or equal to the set threshold, the member port A of the peer port A is notified by the link aggregation control protocol LACP. The device corresponding to the member port B sets the state of the member port B to an inactive state.
The method of claim 1, wherein if the normal packet loss rate of the member port within the set duration is greater than or equal to the set threshold, the normal packet loss rate is greater than or equal to the set threshold. The process of setting the status of a member port to inactive includes:

If it is detected that the normal packet loss rate of the member port a within the set duration exceeds or is equal to the set threshold,

Sending information to the member port b of the local member port a and the member port b of the peer port a to notify the member port a that the normal packet loss rate exceeds or equals the set threshold.

The CPU of the member port a obtains the information that the member port a sends a normal packet loss rate that is greater than or equal to the set threshold, and sets the state of the member port a to the inactive state.
The method according to claim 2, wherein the normal packet loss rate of the member port in the inactive state is lower than the set threshold within the set duration, and the peer member port of the member port in the inactive state When the normal packet loss rate within the set duration is lower than the set threshold, the process of updating the status of the inactive member port to the active state includes:

When the member port A receives the normal packet loss rate within the set time period, the CPU corresponding to the member port B of the peer port of the member port A is notified to the member port A to receive the normal packet loss. Normal information;

After receiving the information that the member port A receives the normal packet loss, the CPU corresponding to the member port B detects that the normal packet loss rate of the member port B within the set duration is lower than the set threshold. The status of member port B is updated to the active state.
The method according to claim 3, wherein the normal packet loss rate of the member port in the inactive state is lower than the set threshold within the set duration, and the peer member port of the member port in the inactive state When the normal packet loss rate within the set duration is lower than the set threshold, the process of updating the status of the inactive member port to the active state includes:

If the normal packet loss rate of the member port a is lower than the set threshold within the set duration, if it is determined that the peer member port b of the member port a is within the set duration, the normal packet loss rate is lower than the set. Threshold, the status of the member port a is updated to the active state.
The method of any one of claims 1 to 5, further comprising:

Check whether there are member ports in the link aggregation port that receive bandwidth loss or excessive bandwidth loss.

If it is detected that the member port X receives the bandwidth loss packet, the information about the member port X receiving the bandwidth loss packet is notified to the member port X CPU and the member port Y of the member port X corresponding to the member port Y of the member port X. If the member port Y is determined to be the destination port of any service redirection, the traffic redirected to the member port Y is redirected to the other port with the lowest bandwidth utilization.
The method of claim 6 redirecting traffic to said member port Y After the port port X is still in excess of the bandwidth loss, the method further includes:

The member port Y uses the hash key as a feature value to redirect the service access control list ACL redirected to the member port Y to the port with the lowest bandwidth utilization, and repeats the redirection action until the member port X no longer receives excess bandwidth loss.
The method of claim 6 further comprising:

If the member port x is sent out of the bandwidth loss packet, the information about the packet loss of the local end is reported to the CPU of the local member port x and the CPU of the member port y of the peer end through the LACP; the CPU of the member port x is obtained. After the local end sends the information of the bandwidth loss packet, if the member port x is the destination port of any service redirection, the service redirected to the member port x is redirected to the other port with the lowest bandwidth utilization. .
The method of claim 8, after redirecting the traffic redirected to the member port x to another port with the lowest bandwidth utilization, if the member port x still sends an excessive bandwidth packet loss, the method further includes :

The member port x is characterized by a hash key, and the service redirected to the member port x is redirected to the port with the lowest bandwidth utilization, and the redirection action is repeated until the member port x is no longer used. Send over bandwidth loss.
The method of any one of claims 1 to 5, further comprising:

If the current mode is the manual link aggregation mode, if the member port of the link aggregation port is detected to be longer than or equal to the set threshold, the member port with the normal packet loss rate exceeding or equal to the set threshold will be received. The status is set to inactive, and the port for sending packets is recalculated in the remaining links except the member ports that have been set to inactive according to the load sharing policy until the detected member ports are within the set duration. If the packet loss rate is lower than the set threshold, the status of the member port is updated to the active state, and the packet sending port is recalculated again;

If it is detected that the member port in the link aggregation port is greater than or equal to the set threshold within the set duration, the state of the member port whose normal packet loss rate exceeds or equals the set threshold is set to an inactive state; According to the load sharing policy, the port for sending packets is recalculated in the remaining links except the member ports that have been set to inactive, until the detected member ports lose the packet loss rate within the set duration. Threshold, then update the status of the member port to the active state, and recalculate the message sending port again.
A device for managing packet loss on a link aggregation member port includes:

The first detecting unit is configured to detect the normal loss of the member ports in the aggregated link within the set duration when the current mode is static or dynamic link aggregation mode after the dynamic management link aggregation member port bandwidth management function is enabled. Packet rate or ordinary packet loss rate;

The state setting unit is configured to: when detecting that the normal packet loss rate of the member port within the set duration is greater than or equal to the set threshold, the peer packet with the normal packet loss rate exceeding or equal to the set threshold is received. The status is set to inactive. It is also set to send a member whose normal packet loss rate exceeds or equals the set threshold when it detects that the normal packet loss rate of the member port is longer than or equal to the set threshold within the set duration. The status of the port is set to inactive; and the normal packet loss rate of the inactive member port is less than the set threshold within the set duration, and the inactive member port of the inactive member port is When the normal packet loss rate within the set duration is lower than the set threshold, the status of the member port in the inactive state is updated to the active state;

a load sharing unit configured to recalculate a port for sending a message in a remaining link other than the active member port according to a load sharing policy when the state of one or more member ports is set to an inactive state; When the inactive member port is updated to the active state, the port on which the message is sent is recalculated.
The device of claim 11, wherein the state setting unit is configured to: when detecting that the normal packet loss rate of the member port within a set duration is greater than or equal to a set threshold, The state of the peer end of the member port whose packet rate exceeds or equals the set threshold is set to the inactive state:

If it is detected that the normal packet loss rate of the member port A is greater than or equal to the set threshold, the member corresponding to the member port B of the peer port A of the member port A is notified by the link aggregation control protocol LACP to indicate The CPU corresponding to the member port B of the peer end sets the state of the peer member port B to an inactive state.
The device of claim 11, wherein the state setting unit is configured to: when detecting that the normal packet loss rate of the member port within a set duration is greater than or equal to a set threshold, The status of the member port whose packet rate exceeds or equals the set threshold is set to inactive:

If it is detected that the normal packet loss rate of the member port a within the set duration exceeds or equals the set threshold, respectively, the CPU that detects the member port a whose normal packet loss rate exceeds or equals the set threshold is detected. The member port of the peer end of the member port a sends a message notifying that the member port a sends a normal packet loss rate exceeding or equal to a set threshold, and sets the state of the member port a to an inactive state.
The device according to claim 12, wherein the state setting unit is configured to realize that when a member port in an inactive state has a normal packet loss rate within a set duration, which is lower than a set threshold, and is inactive, When the peer member port of the member port of the status is lower than the set threshold within the set duration, the status of the member port in the inactive state is updated to the active state:

When the normal packet loss rate of the member port A in the inactive state is lower than the set threshold, the CPU corresponding to the member port and the CPU corresponding to the member port B of the peer port A of the member port A respectively. The member port A is notified that the normal packet loss is normal. After the member port B receives the information that the member port A receives the normal packet loss, the device detects that the peer member port B is in the set duration. If the internal normal packet loss rate is also lower than the set threshold, the status of the member port B is updated to the active state.
The apparatus according to claim 13, wherein the state setting unit is configured to realize that the normal packet loss rate in the inactive state is lower than a set threshold within a set duration, and is inactive by: When the peer member port of the member port of the status is lower than the set threshold within the set duration, the status of the member port in the inactive state is updated to the active state:

When it is detected that the normal packet loss rate of the member port a within the set duration is lower than the set threshold, the number of the normal packet loss rate of the peer member port b that determines the member port a is within the set duration. When the threshold is set, the status of the member port a is updated to the active state.
The apparatus according to any one of claims 11 to 15, the apparatus further comprising:

The second detecting unit is configured to detect whether a member port of the link aggregation port receives an excess bandwidth packet loss or sends an excessive bandwidth packet loss;

The ultra-bandwidth packet loss processing unit is configured to notify the CPU of the local member port X and the member that the member port X receives the packet loss over the bandwidth when the member port X receives the bandwidth loss packet. The CPU corresponding to the member port Y of the peer X of the port X is configured to: when the device corresponding to the member port Y of the peer end receives the information of the packet loss of the member port X, if it is determined that the member port Y is any The destination port of the service redirection redirects the traffic redirected to the member port Y to the other port with the lowest bandwidth utilization.
The apparatus of claim 16, wherein the ultra-bandwidth packet loss processing unit is further configured to redirect traffic redirected to the member port Y to another terminal with the lowest bandwidth utilization If the member port X still receives the bandwidth loss packet, the hash key of the member port Y is eigenvalued, and the service access control list ACL redirected to the member port Y is redirected to the bandwidth. The port with the lowest utilization rate repeats the redirection action until the member port X no longer receives the excess bandwidth packet loss.
The device of claim 16 wherein

The ultra-bandwidth packet loss processing unit is further configured to: notify the local member port of the CPU that wants to use the LA and the said After the CPU of the member port y of the member port x and the CPU of the local member port x obtain the information about the excessive bandwidth loss of the local end, the local member port x is determined to be any service redirection. The destination port redirects traffic redirected to the member port x to other ports with low bandwidth utilization.
The apparatus of claim 18, wherein the ultra-bandwidth packet loss processing unit is further configured to redirect traffic redirected to the member port x member port to another port with low bandwidth utilization, if If the member port x still sends the bandwidth loss packet, the hash key of the member port x is used as the eigenvalue, and the service ACL redirected to the member port x is redirected to the port with the lowest bandwidth utilization. The redirection action is repeated until the member port no longer sends excessive bandwidth loss.
The apparatus according to any one of claims 11 to 15, wherein

The first detecting unit is further configured to: when the current mode is the manual link aggregation mode, detect an ordinary packet loss rate or a normal packet loss rate of the member ports in the link aggregation port within a set duration;

The state setting unit is further configured to: when the current mode is the manual link aggregation mode, if the normal packet loss rate of the member port in the link aggregation port is longer than or equal to the set threshold, The status of the member port whose normal packet loss rate is greater than or equal to the set threshold is set to inactive. It is also set to detect the member port of the link aggregation port when the current mode is manual link aggregation mode. When the normal packet loss rate is greater than or equal to the set threshold, the state of the member port whose normal packet loss rate exceeds or equals the set threshold is set to an inactive state.