WO2016138705A1

WO2016138705A1 - Method and device for managing link in load sharing

Info

Publication number: WO2016138705A1
Application number: PCT/CN2015/081354
Authority: WO
Inventors: 张树冲; 叶正鑫; 钟炜
Original assignee: 中兴通讯股份有限公司
Priority date: 2015-03-04
Filing date: 2015-06-12
Publication date: 2016-09-09
Also published as: CN105991426A

Abstract

A method and device for managing links in load sharing, the method comprising: when an ith load sharing link currently used by an output interface of a network device is abnormal, switching, by the network device, traffic to an (i+1)th load sharing link, and forwarding the traffic to a target network by using the (i+1)th load sharing link, wherein i is an positive integer.

Description

Method and device for managing link in load sharing

Technical field

This document relates to the field of communications, and in particular, to a method and apparatus for managing links in load sharing.

Background technique

In the field of data communication, especially data bearer network, load sharing is a strategy and method for load balancing on multiple links on network node devices. Today, this policy method is used more and more frequently in network data devices. With the frequent use, more and more problems are exposed, and data developers need to use reasonable methods to solve them.

If the link in the load sharing is abnormally interrupted, for example, one link in the load sharing is manually shut down, or the optical fiber module in the link is damaged, the demand for fast convergence of the link is becoming stronger. At present, there are two methods for achieving load sharing convergence at home and abroad: 1. One is that the link is found to be abnormal by the protocol, and the link load sharing is re-formed. For example, the load balancing is performed on the IP route that is formed by the route. When the link is interrupted, the update platform and the forwarding plane route form a route load sharing. The link aggregation is discovered by the Link Aggregation Control Protocol (LACP). Form and update the forwarding path in the link aggregation. 2. Link detection methods such as BFD (Bidirectional Forwarding Detection), OAM (Operation Administration and Maintenance), and other link detection methods are used to detect link connectivity. Switching.

In the above two modes, the load sharing convergence needs to be dependent on the protocol awareness. Then, the forwarding module is notified to reselect the effective path for forwarding. The load sharing convergence technology is implemented in the centralized device or the distributed device within 50 ms. Convergence is very difficult.

FIG. 1 is a schematic diagram of networking of a network in the related art. As shown in Figure 1, the IP of the network a: 100.1.1.2 communicates with the IP of the network b: 200.1.1.2. The network device a checks the routing table to find that there are n routes to the network b. According to the implementation process of the current forwarding plane, when the link to the network b on the network device a has an abnormality, the routing module of the platform will switch the path. After the path switching occurs, the route will not be updated immediately. This causes the route to reconverge to update this A large number of forwarded packets are discarded due to the incomplete routing update.

Summary of the invention

This document provides a method and apparatus for managing links in load sharing to reduce the number of lost packets during link switching.

A method for managing a link in load sharing includes: when an abnormality occurs in the i-th load sharing link currently used by the outbound interface of the network device, the network device switches the traffic to the i+1th load sharing link And forwarding the traffic to the destination network using the i+1th load sharing link, where i is a positive integer.

Optionally, before the network device switches the traffic to the i+1th load sharing link, the method further includes: the network device detecting whether the i+1th load sharing link works normally; If the i+1th load sharing link works normally, the network device triggers the operation of switching the traffic to the i+1th load sharing link.

Optionally, the detecting, by the network device, whether the operation of the (i)th load sharing link is normal: the network device acquiring the first random corresponding to the i-th load sharing link in the pre-stored routing policy a value, the value of the first random number is increased by one to obtain a second random value, and the second random value is used to perform a modulo operation on the number of load sharing links to obtain a pre-stored load sharing of the second random value. For the key value corresponding to the table, obtain the state of the outbound interface corresponding to the key value in the load sharing table; if the state of the outbound interface is enabled, determine that the i+1th load sharing link works normally. If the status of the outbound interface is off, it is determined that the i+1th load sharing link is not working properly.

Optionally, the method further includes: if the i+1th load sharing link is not working properly, the network device detects whether the i++th load sharing link works normally.

Optionally, the load sharing link includes at least one of a route load sharing link, a tunnel load sharing link, and a link aggregation load sharing link.

A device for managing a link in load sharing includes: a switching module, configured to: when an abnormality occurs in the i-th load sharing link currently used by the outbound interface of the network device, switch the traffic to the (i+1)th The load sharing module is configured to: forward the traffic to the destination network by using the i+1th load sharing link, where i is a positive integer.

Optionally, the device further includes: a detecting module, configured to: detect whether the i+1th load sharing link works normally; and the triggering module is configured to: if the i+1th load sharing link If the operation is normal, the operation of switching the traffic to the i+1th load sharing link is triggered.

Optionally, the detecting module includes: an acquiring unit, configured to: acquire a first random value corresponding to the i-th load sharing link in a pre-stored routing policy; and the first calculating unit is configured to: The value of the first random number is increased by one to obtain a second random value; the second calculating unit is configured to perform a modulo operation by using the second random value and the number of load sharing links, to obtain the second random value is stored in advance The key value corresponding to the load sharing table, the state of the outbound interface corresponding to the key value is obtained in the load sharing table; and the determining unit is configured to: if the state of the outgoing interface is open, determine the ith If the status of the outbound interface is off, the i+1 load balancing link is working properly.

Optionally, the detecting module is further configured to: if the i+1th load sharing link is not working properly, the network device detects whether the i++th load sharing link works normally.

A computer readable storage medium storing computer executable instructions for performing the method of any of the above.

The present invention provides an embodiment, in which the outbound interface of the network device is abnormal, the traffic is automatically switched to the next link to continue forwarding to the destination network, thereby realizing the rediscovering of the link without affecting the protocol or the detection mechanism. In this case, the purpose of link switching is to achieve the purpose of fast convergence of load sharing and link aggregation load sharing. This solves the problem of slow link convergence and serious packet loss when link abnormalities are common in data devices.

BRIEF abstract

FIG. 1 is a schematic diagram of networking of a network in related art;

2 is a schematic flowchart of a method for managing a link in load sharing according to an embodiment of the present invention;

3 is a schematic flowchart of a method for managing a link in load sharing according to an application example of the present invention;

FIG. 4 is a schematic diagram of networking of a network according to an embodiment of the present disclosure;

FIG. 5 is a schematic structural diagram of a device for managing a link in load sharing according to an embodiment of the present invention.

Embodiments of the invention

Embodiments of the present invention will be described below with reference to the accompanying drawings. It should be noted that, in the case of no conflict, the features in the embodiments and the embodiments herein may be arbitrarily combined with each other.

2 is a schematic flowchart of a method for managing a link in load sharing according to an embodiment of the present invention. The method shown in Figure 1 includes:

Step 201: When an abnormality occurs in the i-th load sharing link currently used by the outbound interface of the network device, the network device switches the traffic to the i+1th load sharing link.

Step 202: The network device forwards the traffic to the destination network by using an i+1th load sharing link, where i is a positive integer.

According to the method embodiment provided by the present invention, no matter which link is abnormal on the outbound interface of the network device, the traffic is automatically switched to the next link to continue forwarding to the destination network, so that the link is not affected or the detection mechanism is rediscovered. In the case of a link, the purpose of link switching is to achieve the purpose of fast convergence of load sharing and link aggregation load sharing, and the problem of slow link convergence and serious packet loss when the link abnormality is common in data devices is solved.

The method embodiments provided by the present invention are further described below:

If the network device switches the traffic to the i+1th load sharing link, the method further includes: the network device detecting whether the i+1th load sharing link works normally; If the i+1th load sharing link works normally, the network device triggers the operation of switching the traffic to the i+1th load sharing link.

Check whether the link is normal before switching, and ensure that the traffic can be transmitted after being switched to the i+1th link to improve the effectiveness of the handover.

The detecting, by the network device, whether the work of the i+1th load sharing link is normal includes:

The network device acquires a first random value corresponding to the ith load balancing link in the pre-stored routing policy, and increases the value of the first random number by one to obtain a second random value, and uses the second Performing a modulo operation on the number of the load sharing link, and obtaining a key value corresponding to the second random value in the pre-stored load sharing table, and acquiring the state of the outbound interface corresponding to the key value in the load sharing table;

If the status of the outbound interface is on, it is determined that the i+1th load sharing link is working normally; if the status of the outbound interface is off, it is determined that the i+1th load sharing link does not work. normal.

It should be noted that the i+1th link is determined by adding the first random value corresponding to the faulty link to the first random value, which conforms to the provisions of the management protocol in the prior art, and does not need to modify the management in the prior art. For the process, the method provided by the embodiment of the present invention can be used, and the compatibility is strong.

In addition, if the i+1th load sharing link is not working properly, the network device detects whether the i++th load sharing link works normally.

If the outbound interface is down, add 1 to the original random value to redirect to the next load sharing entry. Add the number of links obtained from the load balancing policy table, and set up a new load sharing table key value to check the load sharing table, and then loop through it until the interface status is up; if all interfaces are If it is down, the maximum number of cyclic links is n times.

The load sharing link includes at least one of a route load sharing link, a tunnel load sharing link, and a link aggregation load sharing link.

The foregoing load sharing link is only used as an example. Other types of load sharing links are also applicable to the embodiments of the present invention, and are not described herein again.

The method provided by the embodiment of the present invention is further described below by using an application example:

FIG. 3 is a schematic flowchart diagram of a method for managing a link in load sharing according to an application example of the present invention. The method shown in Figure 3 includes:

301: The data packet sent from the network a is in the forwarding module of the network device a, and the route is searched according to the destination (destination) IP in the packet.

302: Obtain an outbound interface, a next hop IP, a load balancing flag, and a load sharing index in the routing table.

303: Determine whether the destination IP route is load-sharing. If the load balancing is not performed, the next hop destination IP in the encapsulated packet is sent from the outbound interface given in the routing table to terminate the process. If the load is shared, the process is performed. Next step;

304: Check the load sharing routing policy table according to the load sharing index given in the routing table, obtain the number of links to the destination IP from the load sharing routing policy table, and select a routing algorithm based on the packet based on the flow;

305: Perform a modulo operation according to a random value of the routing policy of the policy table and the number of links, and calculate a key value of the load sharing table to check the load sharing table;

306: Obtain an outbound interface and a next hop IP in the load balancing table.

The above steps are the basic process of load sharing processing. When an abnormality occurs in the load sharing link, the processing flow of the related technology is in a waiting state, waiting for the platform to form a new routing load sharing, but in this process, the abnormal link occurs. The packet will be discarded by the link down. The embodiment of the present invention is improved on the basis of the original, and the interface is queried by the forwarding plane when the link is abnormal, and the link is not taken after the interface is down. The added exception is handled as follows:

307: Query the interface status of the egress port attribute table according to the outbound interface information obtained by the load sharing. The forwarding plane obtains the interface information from the load balancing table, and then uses the outbound interface information to find the interface attribute table, and obtains the interface status on/off (UP/DOWN) from the outbound interface attribute table.

308: If the outbound interface is up, the outbound interface and the next hop IP address in the current load sharing table are forwarded, and the process ends; if the outbound interface is down, the next step is performed;

309: Add 1 to the random value obtained originally, and the purpose is to point to the next load sharing entry. The number of links obtained from the load balancing policy table is decremented by one. If not equal to 0, step 305 is executed, and a new load sharing table key value is set to check the load sharing table; if equal to 0, the process ends.

The above steps 305-309 are sequentially looped until the interface status is up; if all interfaces are down, the maximum number of cyclic links is n times.

Data development technicians will know that there are two kinds of link abnormalities in general, namely: 1, artificially shut down the interface; 2, the physical components in the link suddenly damaged, including fiber breaks, optical module anomalies and so on. For the first case, the interface is actively shut down, and the platform will pass the information from the platform to the forwarding plane, and the physical component suddenly is detected by the forwarding plane and reported to the main control platform. For the second case, route load sharing will take more time to converge. However, for the embodiment of the present invention, whether in the first or the second case, minimal or no packet loss convergence is achieved.

The processing in the embodiment of the present invention is mainly processed when the link is abnormal and the new load sharing has not been formed. Referring to the processing flow of 307 to 309 in the flow shown in FIG. 3, the new steady state is restored and then restored. Normal load sharing processing, and just when the link does not reach steady state, packets with the outbound interface will be discarded. The amount of discarded packets is completely determined by the convergence of load sharing.

for example

Example 1 suitable for the present invention: As shown in FIG. 1, network a to network b have N route load sharing links on network device a. With the embodiment of the present invention, when any link of the outbound interface of the network device a is abnormal, the traffic is automatically switched to the next link to continue forwarding to the destination network. The network routing stability and security for the similar network 1 can be greatly improved by using the embodiment of the present invention.

Example 2 suitable for the embodiment of the present invention: As shown in FIG. 1, the network a to the network b have N tunnel load sharing links on the network device a. In the embodiment of the present invention, only the interface information obtained from the routing table is replaced with the interface information obtained from the label table. When the outbound interface of the network device a is abnormal, the traffic will be automatically switched to the next link to continue forwarding to the destination network. The embodiment of the present invention can greatly improve the stability and security of the network tunnel similar to the network 1.

Example 3 of the embodiment of the present invention: FIG. 4 is a schematic diagram of networking of a network according to an embodiment of the present invention. As shown in FIG. 4, the network a to the network b arrives at the network device b through the network device a, which is a smart group (link aggregation) link aggregated by N links, and the N physical links form a smart group load sharing. In the embodiment of the present invention, the interface is obtained from the routing table as a smartgroup interface, and then the forwarding side checks the smartgroup routing policy table and the load sharing table information to obtain the outbound interface information. When the link is abnormal, the embodiment of the present invention can also be connected. Port polling check.

FIG. 5 is a schematic structural diagram of a device for managing a link in load sharing according to an embodiment of the present invention. Referring to the method shown in FIG. 2 and FIG. 3, the apparatus shown in FIG. 5 includes:

The switching module 501 is configured to: when the i-th load sharing link currently used by the outbound interface of the network device is abnormal, switch the traffic to the i+1th load sharing link;

The transmission module 502 is configured to: forward the traffic to the destination network by using the i+1th load sharing link, where i is a positive integer.

Wherein, the device may further include:

The detecting module is configured to: detect whether the i+1th load sharing link works normally;

The triggering module is configured to: if the i+1th load sharing link works normally, trigger the operation of switching the traffic to the i+1th load sharing link.

The detection module includes:

The obtaining unit is configured to: obtain a first random value corresponding to the ith load balancing link in the pre-stored routing policy;

a first calculating unit, configured to: increase a value of the first random number by one to obtain a second random value;

The second calculating unit is configured to perform a modulo operation by using the second random value and the load sharing link number to obtain a key value corresponding to the second random value in a pre-stored load sharing table, where the load sharing table is Obtaining a state of the outbound interface corresponding to the key value;

The determining unit is configured to: if the status of the outbound interface is on, determining that the i+1th load sharing link works normally; if the status of the outbound interface is off, determining the i+1th The load sharing link is not working properly.

The detecting module is further configured to: if the i+1th load sharing link is not working properly, the network device detects whether the i++th load sharing link works normally.

According to the device embodiment provided by the present invention, no matter which link is abnormal on the outbound interface of the network device, the traffic is automatically switched to the next link to continue forwarding to the destination network, so that the link is not affected or the detection mechanism is rediscovered. In the case of a link, the purpose of link switching is to achieve the purpose of fast convergence of load sharing and link aggregation load sharing, and the problem of slow link convergence and serious packet loss when the link abnormality is common in data devices is solved.

One of ordinary skill in the art will appreciate that all or a portion of the steps of the above-described embodiments can be implemented using a computer program flow, which can be stored in a computer readable storage medium, such as on a corresponding hardware platform (eg, The system, device, device, device, etc. are executed, and when executed, include one or a combination of the steps of the method embodiments.

Alternatively, all or part of the steps of the above embodiments may also be implemented by using an integrated circuit. These steps may be separately fabricated into individual integrated circuit modules, or multiple modules or steps may be fabricated into a single integrated circuit module. achieve.

The devices/function modules/functional units in the above embodiments may be implemented by a general-purpose computing device, which may be centralized on a single computing device or distributed over a network of multiple computing devices.

When each device/function module/functional unit in the above embodiment is implemented in the form of a software function module and sold or used as a stand-alone product, it can be stored in a computer readable storage medium. The above mentioned computer readable storage medium may be a read only memory, a magnetic disk or an optical disk or the like.

Industrial applicability

Claims

A method for managing links in load sharing includes:

When the i-th load sharing link currently used by the outbound interface of the network device is abnormal, the network device switches the traffic to the i+1th load sharing link and uses the i+1th load sharing link. The traffic is forwarded to the destination network, where i is a positive integer.
The method of claim 1, wherein the method further comprises: before the network device switches the traffic to the i+1th load sharing link, the method further comprising:

The network device detects whether the work of the i+1th load sharing link is normal;

If the i+1th load sharing link works normally, the network device triggers an operation of switching traffic to the i+1th load sharing link.
The method of claim 2, wherein the detecting, by the network device, whether the i+1th load sharing link works normally comprises:

The network device acquires a first random value corresponding to the ith load balancing link in the pre-stored routing policy, and increases the value of the first random number by one to obtain a second random value, and uses the second Performing a modulo operation on the number of the load sharing link, and obtaining a key value corresponding to the second random value in the pre-stored load sharing table, and acquiring the state of the outbound interface corresponding to the key value in the load sharing table;

If the status of the outbound interface is on, it is determined that the i+1th load sharing link is working normally; if the status of the outbound interface is off, it is determined that the i+1th load sharing link does not work. normal.
The method of claim 2 or 3, further comprising:

If the i+1th load sharing link is not working properly, the network device detects whether the i++th load sharing link works normally.
The method of claim 1, wherein the load sharing link comprises at least one of a route load sharing link, a tunnel load sharing link, and a link aggregation load sharing link.
A device for managing a link in load sharing includes:

The switch module is configured to switch the traffic to the i+1th load sharing link when the i-th load sharing link currently used by the outbound interface of the network device is abnormal.

The transmission module is configured to: forward the traffic to the destination network by using the i+1th load sharing link, where i is a positive integer.
The apparatus of claim 6 further comprising:

The detecting module is configured to: detect whether the i+1th load sharing link works normally;

The triggering module is configured to: if the i+1th load sharing link works normally, trigger the operation of switching the traffic to the i+1th load sharing link.
The apparatus of claim 7 wherein said detecting module comprises:

The obtaining unit is configured to: obtain a first random value corresponding to the ith load balancing link in the pre-stored routing policy;

a first calculating unit, configured to: increase a value of the first random number by one to obtain a second random value;

The second calculating unit is configured to perform a modulo operation by using the second random value and the load sharing link number to obtain a key value corresponding to the second random value in a pre-stored load sharing table, where the load sharing table is Obtaining a state of the outbound interface corresponding to the key value;

The determining unit is configured to: if the status of the outbound interface is on, determining that the i+1th load sharing link works normally; if the status of the outbound interface is off, determining the i+1th The load sharing link is not working properly.
A device according to claim 7 or 8, wherein:

The detecting module is further configured to: if the i+1th load sharing link is not working properly, the network device detects whether the i+2 load sharing link works normally.
The apparatus of claim 6, wherein the load sharing link comprises at least one of a route load sharing link, a tunnel load sharing link, and a link aggregation load sharing link.
A computer readable storage medium storing computer executable instructions for performing the method of any of claims 1-5.