CN102594696A

CN102594696A - Method and device for managing link aggregation groups (LAGs) of network device

Info

Publication number: CN102594696A
Application number: CN2012100598425A
Authority: CN
Inventors: 范汇华
Original assignee: ZTE Corp
Current assignee: ZTE Corp
Priority date: 2012-03-08
Filing date: 2012-03-08
Publication date: 2012-07-18
Anticipated expiration: 2032-03-08
Also published as: CN102594696B; WO2013131431A1

Abstract

The invention discloses a method for managing link aggregation groups (LAGs) of a network device. The method comprises the following steps: judging whether the priority of a current standby LAG is higher than that of a main LAG according to a priority strategy; if so, setting the standby LAG as a current main LAG and setting the main LAG as the current standby LAG; and if not, continuing tojudge whether the priority of the current standby LAG is higher than that of the main LAG according to the priority strategy. The invention further discloses a device for managing LAGs of the network device. By using the method and the device for managing LAGs of the network device, when the network device is used as a Customer Edge (CE) network device, the situation of the LAGs managed by the network device is analyzed and an optimal LAG is selected and used as the main LAG, so that the performance of network is ensured.

Description

Method and device for managing link aggregation group of network equipment

Technical Field

The present invention relates to Link Aggregation technology in ethernet communication, and in particular, to a method and an apparatus for managing a Link Aggregation Group (LAG) of a network device.

Background

Currently, in a communication network, link aggregation technology is generally used to solve the bandwidth bottleneck problem in the network, so-called link aggregation is a method of binding two or more ports with the same property of a network device together as a logical interface to increase the bandwidth. The LAG is established between the two network devices to aggregate the N ports in the network devices, so that the bandwidth between the network devices is increased to be N times of the bandwidth of the original single port, and the bandwidth of the network devices can be increased without hardware upgrading through link aggregation. Moreover, all links in the LAG are mutually backup, and as long as one available link exists in the LAG, the intercommunication between the network devices at two ends can be ensured, thereby ensuring the reliability of the network. However, when all links in the LAG fail or the opposite-end network device fails, the service of the network device is lost, and the reliability of the network is affected, thereby affecting the customer experience.

To solve the above problem, the industry proposes a cross-device link aggregation (MC-LAG, MultiChassis-LAG), as shown in fig. 1, comprising: customer Edge (CE) network equipment, primary Provider Edge (PE) network equipment, and backup PE network equipment; the CE network device is a user side network device, and the PE network device is a network device of a service provider. CE network equipment and main PE network equipment establish a main LAG, and CE network equipment and standby PE network equipment establish a standby LAG; when the active PE network device or the active LAG fails, the CE network device starts the standby LAG to connect with the standby PE network device, and transmits the service to the standby PE network device. However, this method requires an operator to set the main/standby PE network device and the main/standby LAG according to the actual network condition, and the CE network device cannot automatically evaluate the main/standby LAG condition corresponding to the main/standby PE network device, and select the current optimal LAG as the main LAG, which may cause a time delay and affect the network performance due to the need of manual intervention.

It can be seen that, when the existing network device is used as a CE network device, the situation of the LAG managed by the existing network device cannot be automatically analyzed, and the optimal primary LAG is selected, thereby affecting the network performance.

Disclosure of Invention

In view of this, an object of the present invention is to provide a method and an apparatus for managing a network device LAG, which can analyze the situation of the LAG managed by the network device when the network device is used as a CE network device, and select an optimal LAG as an active LAG, thereby ensuring network performance.

In order to achieve the purpose, the technical scheme of the invention is realized as follows:

the invention provides a method for managing a network equipment LAG, which comprises the following steps:

judging whether the priority of the current standby LAG is higher than that of the main LAG or not according to the priority strategy, if so, setting the standby LAG as the current main LAG, and setting the main LAG as the current standby LAG; otherwise, whether the priority of the current standby LAG is higher than that of the main LAG is continuously judged according to the priority strategy.

In the foregoing solution, the determining, according to the priority policy, whether the priority of the current standby LAG is higher than that of the active LAG, further includes: after the network equipment is started, the main LAG is established, and the standby LAG is established.

In the foregoing scheme, the establishing the main LAG includes: judging whether the network equipment has the LAG which enters the received message state, if so, setting the current LAG which enters the received message state as an activated state and using the activated state as a main LAG; otherwise, whether the LAG entering the received message state exists is continuously judged.

In the foregoing scheme, the establishing a standby LAG includes: judging whether LAGs appear in other LAGs except an activated LAG in the network equipment and enter a received message state, and if so, taking the LAG as a current standby LAG; otherwise, continuously judging whether the LAG appears in other LAGs except the activated LAG in the network equipment and entering the received message state.

In the foregoing solution, before setting the LAG currently entering the received packet state to the active state and serving as the active LAG, the method further includes: after the network equipment is started, adding each port into the corresponding LAG one by one;

after receiving a Link Aggregation Control Protocol Data Unit (LACPDU) message, a port in the network equipment sets the port to be in a message received state and moves out of an original LAG; and adding the port into the corresponding LAG, and setting the state of the LAG as the received message.

In the foregoing solution, the priority policy includes:

judging whether the main/standby LAGs are full-duplex LAGs, if not, the priority of the full-duplex LAG is higher than that of the non-full-duplex LAG in the main/standby LAGs;

otherwise, judging whether the main/standby LAGs both meet the minimum number of activated links, if not, the priority of the LAG meeting the minimum number of activated links in the main/standby LAGs is higher than that of the LAG not meeting the minimum number of activated links;

otherwise, comparing the system priorities of the main/standby LAG opposite-end network equipment, and if the system priorities are not equal, the LAG with the low system priority of the opposite-end network equipment has a higher priority than the LAG with the high system priority of the opposite-end network equipment in the main/standby LAG;

otherwise, comparing the sizes of the MAC of the opposite-end network equipment of the main/standby LAG, and if the sizes are not equal, the priority of the LAG with the smaller MAC of the opposite-end network equipment in the main/standby LAG is higher than the priority of the LAG with the larger MAC of the opposite-end network equipment;

otherwise, comparing the number of the ports in the state of receiving the message in the main/standby LAG, if the number of the ports is not equal, the priority of the LAG with more ports in the state of receiving the message in the main/standby LAG is higher than the priority of the LAG with less ports in the state of receiving the message; otherwise, the priority of the primary/backup LAGs is equal.

The invention also provides a device for managing the network equipment LAG, which comprises the following components: the system comprises a priority strategy module and a link aggregation control protocol LACP processing module; wherein,

the priority strategy module is used for judging whether the priority of the current standby LAG is higher than that of the main LAG according to the priority strategy, if so, informing the LACP processing module that the priority of the current standby LAG is higher than that of the main LAG, and if not, continuously judging whether the priority of the current standby LAG is higher than that of the main LAG according to the priority strategy;

and the LACP processing module is configured to set the standby LAG to the current active LAG and set the active LAG to the current standby LAG after receiving the notification that the priority of the current standby LAG is higher than that of the active LAG, where the notification is sent by the priority policy module.

In the above scheme, the LACP processing module is further configured to establish a primary LAG and a standby LAG after the network device is started.

In the above scheme, the LACP processing module is specifically configured to determine whether there is an LAG entering a received packet state in the network device, and if yes, set the current LAG entering the received packet state to an active state and serve as a primary LAG; otherwise, whether the LAG entering the received message state exists is continuously judged.

In the above scheme, the LACP processing module is specifically configured to determine whether LAG occurs in other LAGs except an activated LAG in the network device, and enter a received packet state, and if yes, use the LAG as a current standby LAG; otherwise, continuously judging whether the LAG appears in other LAGs except the activated LAG in the network equipment and entering the received message state.

In the above scheme, the LACP processing module is specifically configured to add each port to a corresponding LAG one by one after the network device is turned on; after receiving a Link Aggregation Control Protocol Data Unit (LACPDU) message, a port in the network equipment sets the port to be in a message received state and moves out of an original LAG; and adding the port into the corresponding LAG, and setting the state of the LAG as the received message.

In the above scheme, the priority policy module is further configured to store a priority policy;

the priority policy includes:

According to the method and the device for managing the network equipment LAG, the priority of the main and standby LAGs established in the current network equipment is compared by using the priority strategy, once the priority of the standby LAG is higher than that of the main LAG, the current standby LAG is the optimal LAG, and the current standby LAG is used as the main LAG, so that the influence caused by manually selecting the main/standby LAG can be avoided, the situation of the self-managed LAG is analyzed when the network equipment is used as CE network equipment, and the optimal LAG is automatically selected as the main LAG, so that the network performance is ensured.

Drawings

FIG. 1 is a schematic diagram of MCLAG;

FIG. 2 is a flowchart illustrating a method for managing a network device LAG according to the present invention;

fig. 3 is a schematic structural diagram of an apparatus for managing a network device LAG according to the present invention.

Detailed Description

The basic idea of the invention is: judging whether the priority of the current standby LAG is higher than that of the main LAG or not according to the priority strategy, if so, setting the standby LAG as the current main LAG, and setting the main LAG as the current standby LAG; otherwise, whether the priority of the current standby LAG is higher than that of the main LAG is continuously judged according to the priority strategy.

The present invention will be described in further detail with reference to the accompanying drawings and specific embodiments.

As shown in fig. 2, a method for managing a network device LAG according to the present invention includes the following steps:

step 101: judging whether the priority of the current standby LAG is higher than that of the main LAG or not according to the priority strategy, if so, executing the step 102; otherwise, manually setting the main standby LAG, and ending the processing flow.

Here, the judgment may be a periodic judgment, and the period may be arbitrarily set as needed; the priority policy includes:

Wherein, the number of the activated links is the total number of links contained in the LAG; the system priority of the opposite-end network equipment and the size of the MAC of the opposite-end network equipment are parameters set for the network equipment in link aggregation according to the LACP requirement in the prior art; the number of the ports in the received message state is the recorded port number which is set as the port number in the received message state.

The manual setting of the primary and standby LAGs is prior art and is not described herein.

Step 102: setting a standby LAG as an active LAG and using the standby LAG as a current main LAG; and setting the main LAG as an inactive LAG and using the main LAG as a current standby LAG.

Before the step 101, it is further required to establish a primary LAG and a standby LAG, including the following steps:

step a: after the network equipment is started, judging whether any LAG enters the received message state in real time, and if so, executing the step b; otherwise, repeating the step a.

Here, the LAG entering the received packet state includes: after the network equipment is started, each port in the network equipment does not receive network information, and each port is added into a corresponding LAG one by one; when any port in the network equipment receives a Link Aggregation Control Protocol Data Unit (LACPDU), setting the port to be in a state of receiving a message and moving out of an original LAG; adding the port receiving the LACPDU message into a corresponding LAG, and setting the state of the LAG as the received message;

wherein the joining a corresponding LAG comprises: searching whether a corresponding LAG exists according to the Key value of the port and the system ID, and if so, adding the LAG; otherwise, establishing a new LAG by taking the Key value of the port and the system ID as the LAG ID; the adding of the LAG and the establishment of the new LAG are technologies specified by the existing LACP protocol, which are not described herein again; the generation and acquisition methods of the Key values are all the technologies specified by the existing LACP, and are not described herein;

the LACPDU message is a message that is received and sent by a network device according to a rule of a Link Aggregation Control Protocol (LACP), and generation and receiving methods thereof are all the prior art, and details are not described here.

Step b: and setting the current LAG entering the received message state as an activated state and taking the LAG as the current main LAG.

Step c: judging whether any LAG in other LAGs except an activated LAG in the network equipment enters a received message state in real time, if so, taking the LAG as a current standby LAG, and executing step 101; otherwise, repeating the step c.

In addition, after the step c is finished, the network equipment can also judge whether the idle time of any port which is managed by the network equipment and is in the state of receiving the message is overtime or not in real time, and if the idle time is overtime, the state of the port is set to be the state of not receiving the message; and (c) removing the port from the current LAG, searching the LAG with the same Key value, the same system ID and the same state, adding the corresponding LAG, and returning to the step (c).

To implement the foregoing method, the present invention further provides an apparatus for managing a network device LAG, as shown in fig. 3, including: a priority policy module 21 and an LACP processing module 22; wherein,

the priority policy module 21 is configured to determine, according to the priority policy, whether the priority of the current standby LAG is higher than that of the main LAG, notify the LACP processing module 22 that the priority of the current standby LAG is higher than that of the main LAG if the priority of the current standby LAG is higher than that of the main LAG, and otherwise, continue to determine, according to the priority policy, whether the priority of the current standby LAG is higher than that of the main LAG;

and the LACP processing module 22 is configured to set the standby LAG to the current active LAG and set the active LAG to the current standby LAG after receiving the notification that the priority of the current standby LAG sent by the priority policy module 21 is higher than that of the active LAG.

The LACP processing module 22 is further configured to, after the network device is turned on, determine whether any LAG in the received packet state exists, and if yes, set the current LAG in the received packet state as an active LAG, and use the active LAG as the current active LAG; otherwise, continuously judging whether any LAG enters the received message state.

The LACP processing module 22 is specifically configured to add each port to a corresponding LAG one by one after the network device in which the LACP processing module is located is turned on, set the port to a state of having received a packet and move out of an original LAG after any one port receives network information, add the port that has received the network information to the corresponding LAG, and set the state of the LAG to be the received packet.

The LACP processing module 22 is specifically configured to search whether a corresponding LAG exists according to a Key value of a port and a system ID, and if yes, add the LAG; otherwise, establishing a new LAG by taking the Key value of the port and the system ID as the LAG ID; the adding of the LAG and the establishment of the new LAG are techniques specified by the existing LACP protocol, and are not described herein again.

The LACP processing module 22 is further configured to determine in real time whether any LAG occurs in LAGs other than the activated LAG and enters a received packet state, if yes, use the LAG as a current standby LAG, notify the priority policy module 21 to compare priorities of the standby LAG and the main LAG, and otherwise, continue to determine whether any LAG occurs in LAGs other than the activated LAG and enters a received packet state; correspondingly, the priority policy module 21 is further configured to receive a notification sent by the LACP processing module 22 for comparing the priorities of the standby LAG and the active LAG.

The priority policy module 21 is further configured to store a priority policy, where the priority policy includes:

The LACP processing module 22 is further configured to determine in real time whether an idle time of any one of the ports managed by the LACP processing module, which is in a state of having received a message, is overtime, and if the idle time is overtime, set the state of the port as a state of not receiving the message; and removing the port from the current LAG, searching the LAG which has the same Key value, the same system ID and the same state as the system ID, adding the corresponding LAG, and then judging whether any LAG appears in the LAGs except the activated LAG in the network equipment or not in real time to enter the received message state.

The LACP processing module further has all functions of processing LACP related messages specified in the network device in the prior art, which are not described herein again.

The device can be used as a logic module to be installed in network equipment such as an Ethernet switch.

The above description is only a preferred embodiment of the present invention, and is not intended to limit the scope of the present invention.

Claims

1. A method of managing a network device link aggregation group, LAG, the method comprising:

2. The method according to claim 1, wherein the determining whether the priority of the current standby LAG is higher than that of the active LAG according to the priority policy further comprises: after the network equipment is started, the main LAG is established, and the standby LAG is established.

3. The method of claim 2, wherein establishing the primary LAG comprises: judging whether the network equipment has the LAG which enters the received message state, if so, setting the current LAG which enters the received message state as an activated state and using the activated state as a main LAG; otherwise, whether the LAG entering the received message state exists is continuously judged.

4. The method of claim 2, wherein establishing the standby LAG comprises: judging whether LAGs appear in other LAGs except an activated LAG in the network equipment and enter a received message state, and if so, taking the LAG as a current standby LAG; otherwise, continuously judging whether the LAG appears in other LAGs except the activated LAG in the network equipment and entering the received message state.

5. The method according to claim 3, wherein before setting the LAG currently entering the received packet state to the active state and serving as the active LAG, the method further comprises: after the network equipment is started, adding each port into the corresponding LAG one by one;

6. The method according to any of claims 1 to 5, wherein the priority policy comprises:

7. An apparatus for managing a network device, LAG, the apparatus comprising: the system comprises a priority strategy module and a link aggregation control protocol LACP processing module; wherein,

8. The apparatus of claim 7,

the LACP processing module is further configured to establish a primary LAG and a standby LAG after the network device is started.

9. The apparatus of claim 8,

the LACP processing module is specifically configured to determine whether there is an LAG entering a received message state in the network device, and if so, set the current LAG entering the received message state to an active state and serve as a primary LAG; otherwise, whether the LAG entering the received message state exists is continuously judged.

10. The apparatus of claim 9,

the LACP processing module is specifically configured to determine whether LAG occurs in other LAGs except an activated LAG in the network device and enters a received message state, and if yes, use the LAG as a current standby LAG; otherwise, continuously judging whether the LAG appears in other LAGs except the activated LAG in the network equipment and entering the received message state.

11. The apparatus of claim 10,

the LACP processing module is specifically configured to add each port to a corresponding LAG one by one after the network device in which the LACP processing module is located is started; after receiving a Link Aggregation Control Protocol Data Unit (LACPDU) message, a port in the network equipment sets the port to be in a message received state and moves out of an original LAG; and adding the port into the corresponding LAG, and setting the state of the LAG as the received message.

12. The apparatus of claim 7,

the priority strategy module is also used for storing a priority strategy;

the priority policy includes: