CN112468391B

CN112468391B - Network fault delivery method and related product

Info

Publication number: CN112468391B
Application number: CN201910857678.4A
Authority: CN
Inventors: 路小刚; 高红亮; 王临春; 党娟娜
Original assignee: Huawei Technologies Co Ltd
Current assignee: Huawei Technologies Co Ltd
Priority date: 2019-09-09
Filing date: 2019-09-09
Publication date: 2024-04-26
Anticipated expiration: 2039-09-09
Also published as: CN112468391A

Abstract

The embodiment of the application discloses a network fault transmission method and a related product, wherein the method can comprise the following steps: receiving first fault information from a second network node; the first fault information is used for indicating the fault condition of a first path, and the first path is a path from the second network node to the target network node; analyzing the first fault information to obtain the target network node and a first address; the first address is the address of an interface where the second network node communicates with the destination network node; in a route path from the first network node to the destination network node, when the next hop of the first network node comprises the first address, the first network node sends second fault information to a third network node; the second fault information is used for indicating the fault condition of a second path, and the second path is a path from the first network node to the destination network node; the network node can be enabled to timely stop forwarding data through the disconnected path.

Description

Network fault delivery method and related product

Technical Field

The present application relates to the field of communications technologies, and in particular, to a network failure transmission method and related products.

Background

A Router (Router), which is a device connected to each local area network or wide area network in the internet, automatically selects and sets a route according to the channel conditions, and transmits signals in order from front to back on the optimal path. The router is the hub of the internet. At present, the router is widely applied to various industries, and products with different grades become the main force for realizing various interconnection of backbone networks, interconnection between backbone networks and interconnection and intercommunication services between backbone networks and the Internet. Routers are the primary node devices of the internet. The router decides the forwarding of the data through the route. The forwarding policy is called routing. As a hub for interconnection between different networks, routers form the backbone of the internet, and its reliability directly affects the quality of the network interconnection.

In recent years, the development speed of the internet has far exceeded expectations, and the application range of the internet support is expanding. With the advent of new applications such as vehicle networking, industrial control, telemedicine, etc., these applications place more stringent demands on network reliability and latency. In practice, routers in the network (i.e. network nodes) may inevitably experience some failures, such as a break in one or more communication links between two routers. When a router downstream of the network fails, the router upstream continues to forward data on the communication link before the failure occurs before the route of the entire network converges. This can result in the network losing much data during route convergence. That is, routers in the network affected by the failure cannot timely stop forwarding data through the broken path until the routes of the entire network converge. Therefore, there is a need to study a network failure delivery scheme in which routers affected by a failure can timely stop forwarding data through a broken path before the routing of the entire network converges.

Disclosure of Invention

The embodiment of the application discloses a network fault transmission method and a related product, which can enable a network node to stop forwarding data through a disconnected path in time, thereby avoiding losing data in the route convergence process.

In a first aspect, an embodiment of the present application provides a network failure delivery method, where the method may include: the first network node receives first fault information from the second network node; the second network node is a neighbor node of the first network node, the first fault information is used for indicating a fault condition of a first path, and the first path is a path from the second network node to a destination network node; analyzing the first fault information to obtain the target network node and a first address; the first address is the address of an interface where the second network node communicates with the destination network node; in a route path from the first network node to the destination network node, when a next hop of the first network node comprises the first address, the first network node sends second fault information to a third network node; the third network node is a neighboring node of the first network node, the second fault information is used to indicate a fault condition of a second path, and the second path is a path from the first network node to the destination network node.

The next hop in the routing path from the first network node to the destination network node comprises the first address indicating that at least one path of the first network node sending data to the destination network node passes through the second network node. Since the second network node fails to the first path of the destination network node and at least one path of the first network node sending data to the destination network node passes through the second network node, the first network node can determine that the second path thereof to the destination network node fails. It can be seen that after the first network node parses the first failure information, whether the second path fails can be quickly determined according to whether the next hop in the routing path from the first network node to the destination network node includes the first address. In case the next hop in the routing path of the first network node to the destination network node comprises the first address, the first network node sends second failure information to a third network node (a neighboring node of the first network node). That is, after the first network node determines that the second path fails according to the first failure information, the second failure information is sent to the third network node. It can be seen that the first network node can quickly send second failure information indicating that the second path failed to its neighboring nodes. Optionally, the first network node only sends the second fault information to its neighboring nodes, so that the scope of spread of the second fault information can be effectively reduced. In the embodiment of the application, the fault information is just a clue, each network node only needs to announce the clue to the neighbor nodes, and the neighbor nodes determine whether to continue to transmit the clue. According to the transmission of the clues, each network node can quickly determine each path with faults so as to quickly converge the route. It can be understood that after the first network node determines that the second path fails according to the first failure information, the first network node generates the second failure information to its neighboring nodes, so that the failure information can be quickly transferred out, so that other network nodes can timely stop forwarding data through the disconnected path.

In the embodiment of the application, a first network node determines whether a second path fails according to the condition that whether the next hop in a routing path from the first network node to a destination network node comprises a first address, and sends failure information to a neighbor node after determining that the second path fails; the fault information can be quickly transmitted out, so that each network node stops forwarding data through the disconnected path in time.

In an optional implementation manner, after the parsing the first fault information to obtain the destination network node and the first address, the method further includes: and discarding the first fault information when the next hop of the first network node does not comprise the first address in a routing path from the first network node to the destination network node.

The next hop in the route path from the first network node to the destination network node does not include the first address, indicating that none of the paths through which the first network node sends data to the destination network node passes through the second network node. That is, the first path fails and does not affect the first network node. Since the failure of the first path does not affect the first network node, each path of the first network node is unchanged. So that the first network node does not have to send failure information to its neighboring nodes. The first network node discards the first failure information, which may reduce the spread of the first failure information.

In this implementation, by discarding the first fault information, the spread of fault information may be effectively reduced.

In an alternative implementation, the first failure information includes a bandwidth field, where the bandwidth field is used to indicate a remaining bandwidth of the first path. The bandwidth field may be the remaining bandwidth of the first path or parsing the bandwidth field may result in the remaining bandwidth of the first path.

In this implementation, the first network node receives the first failure information including the bandwidth field, and may quickly and accurately obtain the remaining bandwidth of the first path, so as to subsequently adjust the traffic of the first network node forwarding data through each path.

In an alternative implementation, the first fault information further includes an interface field for indicating the first address and a prefix field for indicating the destination network node.

In this implementation, the first network node may accurately and quickly determine the path that fails through the interface field and the prefix field.

In an alternative implementation, before the sending the second fault information to the third network node, the method further includes: the first network node determining a remaining bandwidth of the second path; generating the second fault information according to the second path and the residual bandwidth of the second path; the second fault information is also used for analyzing and obtaining the second bandwidth.

In this implementation manner, the first network node generates the second fault information carrying the residual bandwidth of the second path, so that other network nodes adjust the traffic forwarded through each path according to the residual bandwidth of the second path, thereby reducing the packet loss rate.

In an alternative implementation, the first network node determining the remaining bandwidth of the second path includes: the first network node analyzes the first fault information to obtain the residual bandwidth of the first path; the first network node takes the residual bandwidth of the first path as the bandwidth of a third path in a routing table; the third path is included in the second path, and a next hop of the third path is the first address; and taking the sum of bandwidths of all paths included in the second path in the routing table as the residual bandwidth of the second path.

In this implementation manner, the first network node uses the sum of bandwidths of the paths included in the second path in the routing table as the residual bandwidth of the second path, so that the residual bandwidth of the second path can be obtained quickly and accurately.

In an optional implementation manner, after the parsing the first fault information to obtain the destination network node and the first address, the method further includes: the first network node queries a route path from the first network node to the destination network node in a route table, wherein a next hop of the first network node is a path of the first address to obtain a fourth path, and the fourth path is included in the second path; updating the state of the fourth path in the routing table and/or updating the bandwidth of the fourth path to the residual bandwidth of the first path according to the residual bandwidth of the first path; the state of the fourth path is used to determine the bandwidth of the fourth path.

In the implementation manner, according to the first residual bandwidth, the state and/or the residual bandwidth of the fourth path in the routing table can be updated in time, so that the implementation is simple.

In an optional implementation manner, the updating the state of the fourth path in the routing table according to the remaining bandwidth of the first path includes: updating the state of the fourth path in the routing table to be a fault under the condition that the residual bandwidth of the first path is zero; determining a bandwidth reduction condition of the fourth path if the remaining bandwidth of the first path is greater than zero; updating the state of the fourth path in the routing table to be the bandwidth reduction condition.

In this implementation, the state of the fourth path may be accurately and quickly updated according to the first residual bandwidth.

In an optional implementation manner, after updating the state of the fourth path in the routing table according to the residual bandwidth of the first path and/or updating the bandwidth of the fourth path to the residual bandwidth of the first path, the method further includes: the first network node determines the bandwidth of a fifth path in the routing table, wherein the fifth path is contained in the second path; and the first network node adopts a non-equivalent load sharing UCMP technology to distribute flow for the fourth path and the fifth path according to the bandwidth of the fourth path and the bandwidth of the fifth path.

In the implementation mode, UCMP technology is adopted to distribute flow for a fourth path and a fifth path according to the bandwidth of the fourth path and the bandwidth of the fifth path; the bandwidth utilization rate can be effectively improved, and the congestion of the link can be reduced.

In an alternative implementation, the method further includes: the first network node generates third fault information under the condition that a sixth path fails; the sixth path is a path from the first network node to a reference network node, and the third fault information is used for analyzing and obtaining the residual bandwidth of the sixth path; the first network node sends the third fault information to the third network node.

In the implementation manner, the first network node sends third fault information to the third network node under the condition that the sixth path is determined to be faulty; the third network node may be timely notified of the remaining bandwidth of the sixth path so that the third network node timely stops the disconnected path from transmitting data.

In a second aspect, an embodiment of the present application provides another network failure delivery method, where the method may include: generating fault information by the second network node under the condition that the first path fails; transmitting the fault information to a first network node; the fault information is used for indicating the fault condition of the first path, the first network node is a neighbor node of the second network node, and the fault information comprises a bandwidth field, and the bandwidth field is used for indicating the residual bandwidth of the first path.

The bandwidth field may be a remaining bandwidth of the first path or the first network node may parse the bandwidth field to obtain the remaining bandwidth of the first path. The first network node analyzes the fault information to obtain the residual bandwidth of the first path, and the first network node can update the state of each path in the routing table according to the residual bandwidth of the first path and adjust the flow of at least one path.

In the embodiment of the application, the second network node sends the fault information comprising the bandwidth field to the first network node, so that the first network node can adjust the state of each path in the routing table according to the residual bandwidth of the first path, and the traffic of each path is further planned more reasonably.

In an alternative implementation, the first path is a path from the second network node to a destination network node; the fault information further comprises an interface field for indicating an interface address of the second network node and a prefix field for indicating the destination network node.

In the implementation manner, the fault path can be accurately indicated through the interface field and the prefix field, the occupied bytes are less, and the signaling overhead can be reduced.

In an alternative implementation manner, the method further includes, before the second network node generates the fault information in a case that the first path fails: the second network node detects that the reference path has a fault; the reference path is included in the first path; the second network node sets the state of the reference path in the routing table as a fault; and under the condition that the state of the reference path in the routing table is a fault, determining that the first path is faulty.

After the second network node detects that the reference path has a fault, the state of the reference path in the routing table is set as the fault. The second network node may detect the state of the reference path in the routing table before transmitting data over the reference path, such that the second network node may determine whether to transmit data over the reference path based on the state of the reference path. The reference path is included in the first path. The second network node may detect whether each path included in the first path in the routing table has failed before sending data over the first path. If the second network node detects that at least one path included in the first path fails, the second network node may stop sending data through the failed path and send failure information indicating that the first path fails to its neighbor nodes.

In this implementation manner, the second network node sets the state of the reference path in the routing table as a fault, so that the situation that the first path fails can be timely detected.

In an alternative implementation, the generating, by the second network node, the fault information includes: the second network node takes the sum of bandwidths of all paths included in the first path in a routing table as the residual bandwidth of the first path under the condition that the first path fails; and generating the fault information according to the first path and the residual bandwidth of the first path.

In the implementation manner, the residual bandwidth of the first path can be obtained quickly and accurately, and fault information which can indicate the residual bandwidth of the first path is generated.

In a third aspect, an embodiment of the present application provides a first network node, which may include: a receiving unit, configured to receive first failure information from a second network node; the second network node is a neighbor node of the first network node, the first fault information is used for indicating a fault condition of a first path, and the first path is a path from the second network node to a destination network node; the analyzing unit is used for analyzing the first fault information to obtain the target network node and the first address; the first address is the address of an interface where the second network node communicates with the destination network node; a sending unit, configured to send second fault information to a third network node when a next hop of the first network node includes the first address in a routing path from the first network node to the destination network node; the third network node is a neighboring node of the first network node, the second fault information is used to indicate a fault condition of a second path, and the second path is a path from the first network node to the destination network node.

In an alternative implementation, the network node further comprises: and the discarding unit is used for discarding the first fault information when the next hop of the first network node does not comprise the first address in a route path from the first network node to the destination network node.

In an alternative implementation, the first failure information includes a bandwidth field, where the bandwidth field is used to indicate a remaining bandwidth of the first path.

In an alternative implementation, the network node further comprises: a determining unit configured to determine a remaining bandwidth of the second path; a first generating unit, configured to generate the second fault information according to the second path and a remaining bandwidth of the second path; the second fault information is also used for analyzing and obtaining the second bandwidth.

In an optional implementation manner, the parsing unit is further configured to parse the first failure information to obtain a residual bandwidth of the first path; the determining unit is specifically configured to take the remaining bandwidth of the first path as the bandwidth of a third path in the routing table; taking the sum of bandwidths of all paths included in the second path in the routing table as the residual bandwidth of the second path; the third path is included in the second path, and a next hop of the third path is the first address.

In an alternative implementation, the network node further comprises: a query unit, configured to query a route path from the first network node to the destination network node in a route table, where a next hop of the first network node is a path of the first address to obtain a fourth path; the fourth path is included in the second path; an updating unit, configured to update a state of the fourth path in the routing table and/or update a bandwidth of the fourth path to a remaining bandwidth of the first path according to the remaining bandwidth of the first path; the state of the fourth path is used to determine the bandwidth of the fourth path.

In an optional implementation manner, the updating unit is specifically configured to update the state of the fourth path in the routing table to be a fault when the remaining bandwidth of the first path is zero; determining a bandwidth reduction condition of the fourth path if the remaining bandwidth of the first path is greater than zero; updating the state of the fourth path in the routing table to be the bandwidth reduction condition.

In an alternative implementation, the network node further comprises: a flow distribution unit, configured to determine a bandwidth of a fifth path in the routing table, where the fifth path is included in the second path; and distributing flow for the fourth path and the fifth path according to the bandwidth of the fourth path and the bandwidth of the fifth path by adopting a non-equivalent load sharing UCMP technology.

In an alternative implementation, the network node further comprises: a second generation unit configured to generate third failure information in the case where the sixth path fails; the sixth path is a path from the first network node to a reference network node, and the third fault information is used for analyzing and obtaining the residual bandwidth of the sixth path; the sending unit is further configured to send the third failure information to the third network node.

In a fourth aspect, embodiments of the present application provide a second network node, which may comprise: a generating unit for generating fault information in case of a fault of the first path; a sending unit, configured to send the fault information to a first network node; the fault information is used for indicating the fault condition of a first path, and the first network node is a neighbor node of the second network node; the failure information includes a bandwidth field indicating a remaining bandwidth of the first path.

In an alternative implementation, the network node further comprises: the detection unit is used for detecting that a reference path is failed, and the reference path is included in the first path; a setting unit configured to set a state of the reference path in the routing table as a failure; and the determining unit is used for determining that the first path fails under the condition that the state of the reference path in the routing table is failure.

In an optional implementation manner, the generating unit is specifically configured to, in a case where the first path fails, take a sum of bandwidths of paths included in the first path in a routing table as a remaining bandwidth of the first path; and generating the fault information according to the first path and the residual bandwidth of the first path.

In a fifth aspect, embodiments of the present application provide a network device, which may include: a memory for storing a program; a processor for executing the program stored in the memory, the processor being configured to perform the methods of the first aspect to the second aspect and optional implementations described above when the program is executed.

In a sixth aspect, embodiments of the present application provide a computer readable storage medium having instructions stored therein, which when run on a computer, cause the computer to perform the method of the above aspects.

In a seventh aspect, embodiments of the present application provide a computer program product comprising instructions which, when run on a computer, cause the computer to perform the method of the above aspects.

Drawings

Fig. 1 is a schematic diagram of a network including a plurality of network nodes according to an embodiment of the present application;

fig. 2 is a schematic diagram of another network including a plurality of network nodes according to an embodiment of the present application;

Fig. 3 is a schematic diagram of yet another network including a plurality of network nodes according to an embodiment of the present application;

Fig. 4 is a schematic diagram of a network architecture according to an embodiment of the present application;

Fig. 5 is a flowchart of a network failure transmission method according to an embodiment of the present application;

fig. 6A to fig. 6C are schematic diagrams of first fault information provided in an embodiment of the present application;

FIG. 7 is a schematic diagram of a fault information diffusing process according to an embodiment of the present application;

FIG. 8 is a schematic diagram of another fault information diffusing process according to an embodiment of the present application;

FIG. 9 is a flowchart of another network failure delivery method according to an embodiment of the present application;

FIG. 10A is a schematic diagram illustrating a fault delivery process in a fabric network according to an embodiment of the present application;

FIG. 10B is a schematic diagram of a routing table of a network node in a fabric network according to an embodiment of the present application;

FIG. 11 is a schematic diagram illustrating a fault delivery process in another fabric network according to an embodiment of the present application;

Fig. 12 is a schematic structural diagram of a first network node according to an embodiment of the present application;

fig. 13 is a schematic structural diagram of a second network node according to an embodiment of the present application;

Fig. 14A is a schematic diagram of an application scenario of a network failure delivery method according to an embodiment of the present application;

fig. 14B is a schematic structural diagram of a network node according to an embodiment of the present application;

fig. 15 is a schematic diagram of an application scenario of another network failure delivery method according to an embodiment of the present application;

fig. 16 is a schematic structural diagram of another network node according to an embodiment of the present application.

Detailed Description

The technical solutions in the embodiments of the present application will be clearly described below with reference to the drawings in the embodiments of the present application.

The terms first, second, third and the like in the description and in the claims and in the above-described figures are used for distinguishing between similar elements and not necessarily for describing a particular sequential or chronological order. Furthermore, the terms "comprise" and "have," as well as any variations thereof, are intended to cover a non-exclusive inclusion, such as a series of steps or elements. The method, system, article, or apparatus is not necessarily limited to those explicitly listed but may include other steps or elements not explicitly listed or inherent to such process, method, article, or apparatus. "and/or" is used to mean that one or both of the objects to which it is connected are selected between. For example, "a and/or B" means A, B or a+b.

Communication networks (e.g., the internet, autonomous networks, etc.) composed of multiple network nodes are typically required to meet the following reliability requirements: (1) fast (millisecond or microsecond) perception of faults; (2) Bandwidth transfer, the remaining bandwidth of the path that is about to fail is transferred to other network nodes in the network. In the present application, the network node may be a router, a switch, or other network devices with routing functions. The necessity of satisfying both of these reliabilities is described below in conjunction with fig. 1 and 2, respectively.

Fig. 1 is a schematic diagram of a network including a plurality of network nodes according to an embodiment of the present application. In fig. 1, a01, C11, C21, and a71 each represent a network node, and a01 may transmit data to a71 through C11 and C21, respectively. It can be seen that a01 in fig. 1 sends data to a71 with two paths that form load sharing. In the network of fig. 1, a01 transmits data to a71 via C11, and a01 stops transmitting data to a71 via C11 after detecting that the path for transmitting data to a71 via C11 is interrupted. That is, a01 continues to send data to a71 via the path of C11 before it senses an interruption of the path, which results in a loss of a large amount of data. Therefore, the network node needs to sense the fault through express delivery, so that the network node can timely stop sending data through the interrupted path, and further, the loss of the data is avoided or reduced. It will be appreciated that the faster the network node perceives a failure, the faster it can cease sending data over the interrupted path. Therefore, in many scenarios, it is necessary for the network node to quickly perceive the failure, especially those with high requirements for packet loss rate.

Fig. 2 is a schematic diagram of another network including a plurality of network nodes according to an embodiment of the present application. In fig. 2, a01, C11, C21 and a71 respectively represent a network node, a01 may send data to a71 through C11 and C21 respectively, C11 forwards data to a71 through D path and E path respectively, C21 forwards data to a71 through F path and G path respectively, and bandwidths of D path, E path, F path and G path are all 10G. As shown in fig. 2, the ratio of the flow of data sent by a01 to a71 through C11 to the flow of data sent by C21 to a71 is 1:1, and after the E path is interrupted, the bandwidth between C11 and a71 is reduced to 10G, and the bandwidth between C21 and a71 is still 20G. If the ratio of the traffic of a01 sending data to a71 through C11 to the traffic of a71 through C21 remains unchanged after the bandwidth between C11 and a71 is reduced to 10G, it is likely to cause the path of C11 to a71 to continue to be congested. Since the traffic of a01 sending data to a71 through C11 remains unchanged after the bandwidth between C11 and a71 decreases, the bandwidth between C11 and a71 is insufficient to support the transmission of the traffic, which naturally results in continuous congestion of the path between C11 and a 71. Congestion refers to the phenomenon that a certain part of a communication sub-network is too many packets to reach the part of the network, so that the part of the network is not processed so much that the performance of the part and even the whole network is reduced, and when the congestion is serious, network communication service is stopped, namely a deadlock phenomenon occurs. When vehicles in the holiday road network are greatly increased, various trend traffic flows are mutually interfered, so that the time for each vehicle to reach a destination is relatively increased (namely delay is increased), and even vehicles on a certain section of road cannot start due to blockage (namely partial deadlock occurs).

Several schemes for perceiving faults and achieving route convergence are described below.

Route convergence refers to the process of re-establishing a route table, sending, learning and stabilizing after the topology structure of the network is changed, and notifying all related routers in the network of the change. I.e. the behavior of discovering alternative routes by recalculating routes caused by network topology changes. All routers (i.e., network nodes) in the routing domain can be brought into an agreed state for the current network structure and route forwarding by route convergence. Convergence time refers to the state transition process from the change in topology of the network to the re-maintenance of the routing tables in all routing devices in the network.

The first scheme is to implement route convergence by a route protocol (e.g., a border network protocol), and the route convergence speed is slow and the fault perception speed is slow (second level). In this scheme, each network node in the network will not stop sending data over the broken path until the route of the entire network converges. It will be appreciated that this approach does not meet the need to perceive faults quickly (in milliseconds or microseconds) well.

The second scheme is to establish an end-to-end label path, and when detecting that the end-to-end label path for transmitting certain data fails, switch the data to the end-to-end label path which does not fail for transmission. Here the end refers to the network node. This scheme is described below in connection with fig. 3. Fig. 3 is a schematic diagram of another network including a plurality of network nodes according to an embodiment of the present application. In fig. 3, a01, C11, C21, and a71 each represent a network node, a01 may send data to a71 through C11 and C21, a01 path sending data to a71 through C11 is an end-to-end label path P, and a01 path sending data to a71 through C21 is an end-to-end label path Q. Referring to fig. 3, a01 transmits data to a71 through an end-to-end tag path P, and when a communication link (also referred to as a path) between C11 and a71 is interrupted, a01 detects that the end-to-end tag path P is failed, and a01 switches from the end-to-end tag path P to an end-to-end tag path Q, that is, stops transmitting data through the end-to-end tag path P and transmits data through the end-to-end tag path Q. The disadvantage of this approach is that the network node needs to have label path capability and needs to configure the bidirectional forwarding detection (Bidirectional Forwarding Detection, BFD) protocol. The BDF protocol is a network protocol for detecting a failure between two forwarding points. BDF is a bidirectional forwarding detection mechanism, which can provide millisecond detection and can realize rapid detection of links.

The above schemes for sensing faults and achieving route convergence have one or more disadvantages of low convergence speed, wide fault spread range, detouring, need of having label path capability and the like. The implementation of these schemes is not described in detail here. In addition, none of the several failure-aware and route convergence-achieving schemes described above currently deliver the remaining bandwidth of the failed path. Thus, there is a need to investigate new perceived failures and schemes for achieving route convergence, i.e. network failure delivery schemes, that do not suffer from the above-mentioned drawbacks. In order to solve the problem of fast (millisecond or microsecond) perception of a network node in a network, the embodiment of the application provides a network fault transmission scheme. The network fault transmission scheme overcomes the defects of low convergence speed, detouring, needing to have label path capability and the like, can spread faults to all networks affected by the faults in a smaller range, and can effectively avoid unnecessary route convergence of other network nodes. The implementation of the network failure delivery scheme is described in further detail below. In order to solve the problem of bandwidth transfer, the embodiment of the application further provides a scheme of residual bandwidth passing through the fault information transfer path. The network failure transmission scheme provided by the embodiment of the application is described below.

Fig. 4 is a schematic diagram of a network architecture according to an embodiment of the present application. It should be understood that the network architecture in fig. 4 is only an example, and the method in the embodiment of the present application may be applied to various networks such as fabric networks. Fabric networks refer to regularized networking of networks. As shown in fig. 4, a01 represents a first network node, C11 represents a second network node, T00 represents a third network node, T70 is a destination network node (a node that receives data), T00 is a source network node (i.e., a node that sends data), and C21, C12, C22, a02, a71, and a72 are all network nodes. In fig. 4, T00 sends data to T70, network nodes other than T00 and T70 are used to forward data, a01 receives first failure information from C11, and a01 sends second failure information to both T00 and C21. In the network, if a communication link is provided between two network nodes, the two network nodes are neighboring nodes. For example, in fig. 4, the neighbor nodes of C11 include a01 and a71, the neighbor nodes of a01 include T00, C11, and C21, the neighbor nodes of T00 include a01 and a02, and the neighbor nodes of T70 include a71 and a72. The method of the first network node a01 in fig. 4 to communicate failure information is described below. Fig. 5 is a flowchart of a network failure transmission method according to an embodiment of the present application. As shown in fig. 5, the method may include:

501. the first network node receives first failure information from the second network node.

The first network node may be a01 in fig. 4 and the second network node may be C11 in fig. 4. The second network node is a neighboring node of the first network node, and the first fault information is used to indicate a fault condition of a first path, where the first path is a path from the second network node to a destination network node. The destination network node may be T70 in fig. 4. The first path may be a C11 to T70 path.

In an alternative implementation, the first fault information may include: a bandwidth field, a prefix field, and an interface field; the bandwidth field is used to indicate the remaining bandwidth of the first path, the interface field is used to indicate a first address, and the prefix field is used to indicate the destination network node. The first address is an address of an interface through which the second network node communicates with the destination network node. Fig. 6A is a schematic diagram of first fault information according to an embodiment of the present application. As shown in fig. 6A, the first failure information includes a bandwidth field 601, an interface field 602, and a prefix field 603. Fig. 6B is a schematic diagram of another first fault information provided in an embodiment of the present application. Fig. 6B is a specific example of the first failure information in fig. 6A, where the information in the bandwidth field 601 is the remaining bandwidth of the first path, the information in the interface field 602 is the first address, and the information in the prefix field 603 is an interface address of the destination network node, i.e. the destination address. In FIG. 6B, 20.1.1.2/3 is the first address, 10.1.1.0/24 is the destination address, and a bandwidth field of 0 indicates that the remaining bandwidth of the first path is 0. In some embodiments, the first failure information may also carry a serial number. And when the first network node receives a plurality of identical messages, the message with the largest sequence number is used as the reference. The same message refers to the fault information that the carried interface field and the prefix field are the same, namely the fault information indicating the fault condition of the same path. In practical application, the second network node may send two or more fault information to the first network node successively, where interface fields and prefix fields of the fault information are the same, and a sequence number carried by the fault information sent later is greater than a sequence number carried by the fault information sent earlier. It will be appreciated that if the second network node only transmits fault information once to the first network node, the first network node cannot obtain the fault information when the first network node does not successfully receive the fault information. The second network node sends the fault information of the same path to the first network node twice or more, so that the residual bandwidth of the path can be updated in time, and the first network node can be ensured to receive the fault information. Alternatively, the second network node may periodically (e.g. every 3 seconds) detect the remaining bandwidth of the first path after sending the failure information of the first path to the first network node, and send the failure information indicating the failure condition of the first path to the first network node. Fig. 6C is a schematic diagram of still another first fault information provided in an embodiment of the present application. As shown in fig. 6C, the first failure information may also carry a serial number 604.

502. And analyzing the first fault information to obtain the target network node and the first address.

The first address is an address of an interface through which the second network node communicates with the destination network node. Optionally, the first fault information includes an interface field and a prefix field, the first network node may parse the interface field to obtain the first address, and parse the prefix field to obtain the destination network node. The information of the prefix field may be an interface address of the destination network node, the first network node may store a correspondence between the destination network node and the interface address of the destination network node, and the first network node may obtain the destination network node according to the correspondence and the interface address. The first network node may store a correspondence between the second network node and the first address, and the first network node may determine the second network node according to the correspondence and the first address.

503. And when the next hop of the first network node comprises the first address in a routing path from the first network node to the destination network node, the first network node sends second fault information to the third network node.

The third network node is a neighbor node of the first network node, which may be T00 in fig. 4. The second fault information is used for indicating a fault condition of a second path, and the second path is a path from the first network node to the destination network node. The format of the second fault information may be the same as the format of the first fault information. Optionally, the second fault information includes a bandwidth field, an interface field, and a prefix field. The bandwidth field in the second failure information is used to indicate the remaining bandwidth of the second path, the prefix field in the second failure information is used to indicate the destination network node, and the interface field in the second failure information is used to indicate an interface address of the first network node. In some embodiments, the second fault information further includes a serial number. It should be appreciated that each network node in the network may send failure information in the same format, and that since the information carried by the first failure information has been described in detail above, the second failure information will not be described in detail herein. The next hop in the routing path from the first network node to the destination network node comprises the first address indicating that at least one path of the first network node sending or forwarding data to the destination network node passes through the second network node. Since the second network node fails to the first path of the destination network node and at least one path of the first network node sending data to the destination network node passes through the second network node, the first network node can determine that the second path thereof to the destination network node fails. Thus, the first network node needs to send the second failure information to the third network node. The routing table of the first network node may store routing paths for the first network node to send or forward data to one or more network nodes. Table 1 is an example of a routing table of a first network node according to an embodiment of the present application.

TABLE 1

Destination address	Next hop	Status of	……
				10.1.1.0/24	20.1.1.2/32	Failure of	……
10.1.1.0/24	30.1.1.2/32	Normal state	……
				……	……	……	……
10.1.1.2/32	20.1.1.2/32	Failure of

In table 1, the destination addresses represent the interface addresses of the destination network nodes, i.e. each destination address represents a destination network node, and each next hop is the interface address of a network node. In table 1, the path corresponding to any row is the path from the first network node to the destination network node represented by the destination address of any row, and the state of any row represents the fault condition of the path corresponding to any row. In practical application, after the first network node analyzes the first fault information to obtain the destination network node and the first address, the first network node can query whether the next hop of the destination address corresponding to the destination network node in the routing table includes the first address; if yes, sending second fault information to a third network node; if not, discarding the first fault information. Referring to fig. 4, 10.1.1.0/24 is an interface address of T70, 20.1.1.2/32 is an interface address of C11, 30.1.1.2/32 is an interface address of C21, and the next hops for forwarding data from a01 (i.e., the first network node) to T70 (i.e., the destination network node) are C11 and C21. For example, the destination network node obtained by the first network node analyzing the first fault information is T70 in fig. 4, and the first network node may query whether the next hop of the destination address 10.1.1.0/24 in the routing table includes the first address; in case the next hop of the destination address 10.1.1.0/24 comprises the first address, the second failure information is sent to the third network node.

In the embodiment of the application, a first network node determines whether a second path fails according to the condition that whether the next hop of the first network node comprises a first address in a route path from the first network node to a destination network node, and sends failure information to a neighbor node after determining that the second path fails; the fault information can be quickly transmitted out, so that each network node stops forwarding data through the disconnected path in time.

The purpose of the network nodes in the network to transmit the fault information is to enable the network nodes affected by the fault to update the states of all paths in the routing table in time, further to stop transmitting data through the disconnected paths in time and to distribute traffic for all paths more reasonably. The following describes how to update the state of each path in the routing table, taking the first network node as an example.

The first network node updates the status and/or the remaining bandwidth of each path in its routing table according to the first failure information before performing step 502 as follows: the first network node queries a route path from the first network node to the destination network node in a route table, and the next hop of the first network node is the path of the first address to obtain a fourth path; and updating the state of the fourth path in the routing table and/or updating the bandwidth of the fourth path to the residual bandwidth of the first path according to the residual bandwidth of the first path. The fourth path is included in the second path; the state of the fourth path is used to determine the bandwidth of the fourth path. The fourth path may be the third path described above.

In some embodiments, the first network node may update the bandwidth of the fourth path in the routing table to be the remaining bandwidth of the first path according to the remaining bandwidth of the first path. Wherein the routing table stores bandwidths of the paths. Table 2 is an example of another routing table provided in an embodiment of the present application.

TABLE 2

Destination address	Next hop	Status of	……
				10.1.1.0/24	20.1.1.2/32	0	……
10.1.1.0/24	30.1.1.2/32	100G	……
				……	……	……	……
10.1.1.2/32	20.1.1.2/32	50G

The column of states in table 2 may store the bandwidth of each path. The first network node may update the bandwidth of each path based on the received failure information or the detected failure. For example, before the first network node receives the first failure information, the bandwidth of the fourth path in the routing table is 100G, and if the remaining bandwidth of the first path obtained by the first network node analyzing the first failure information is 0, the first network node updates the bandwidth of the fourth path in the routing table to be 0. Also for example, the first network node detects a path disruption and updates the state of the path in its routing table.

In some embodiments, the first network node may update the state of the fourth path in the routing table to be a failure if the remaining bandwidth of the first path is zero; and under the condition that the residual bandwidth of the first path is larger than zero, determining the bandwidth reduction condition of the fourth path, and updating the state of the fourth path in the routing table to be the bandwidth reduction condition. The states of the paths in the routing table can be divided into three types, namely normal, fault and bandwidth reduction conditions. If the state of a certain path in the routing table is normal, indicating that the bandwidth of the path is the stored bandwidth of the path in the routing table; if the state of the path is a fault, indicating that the bandwidth of the path is 0; if the state of the path is a bandwidth reduction condition, indicating that the path is faulty and the remaining bandwidth is not 0, the first network node may determine the bandwidth of the path according to the bandwidth reduction condition and the bandwidth of the path stored in the routing table. The bandwidth reduction condition may reflect a bandwidth reduction condition of the path, such as a halving of the bandwidth, a third of the bandwidth reduction, etc. For example, if the state of a path in the routing table is halved, and the bandwidth of the path stored in the routing table is 100G, the current remaining bandwidth of the path is 50G. Table 3 is an example of yet another routing table provided by an embodiment of the present application.

TABLE 3 Table 3

Destination address	Next hop	Status of	……
				10.1.1.0/24	20.1.1.2/32	Failure of	……
10.1.1.0/24	30.1.1.2/32	Normal state	……
				……	……	……	……
10.1.1.2/32	20.1.1.2/32	Halving bandwidth

In Table 3, the column of states may store the state of each path. In practical applications, each network node may update the state of each path in each routing table according to the received fault information in a plurality of ways, which is not limited by the embodiment of the present application.

One advantage of the network failure delivery method of the embodiments of the present application compared to other network failure delivery schemes is that the remaining bandwidth of the paths can be delivered, so that each network node can allocate traffic for each path according to the remaining bandwidth of each path. The manner in which traffic is allocated to each path based on the remaining bandwidth of the path is described below.

The first network node may perform the following operations after updating the state of the fourth path in the routing table and/or updating the bandwidth of the fourth path to the remaining bandwidth of the first path according to the remaining bandwidth of the first path:

The first network node determines the bandwidth of a fifth path in the routing table; the first network node allocates traffic for the fourth path and the fifth path using a non-equal value load sharing (Unequal Cost Multiple Path, UCMP) technique based on the bandwidth of the fourth path and the bandwidth of the fifth path.

The fifth path and the fourth path are both included in the second path. The first network node may adjust a ratio of traffic allocated to the fourth path and the fifth path to a ratio of the bandwidth of the current fourth path and the bandwidth of the fifth path according to the bandwidth of the fourth path and the bandwidth of the fifth path using UCMP technique. UCMP refers to the fact that if there are multiple equivalent physical links (i.e., paths) with different bandwidths to the destination, the traffic will be shared onto each physical link proportionally according to the bandwidths. Therefore, all links can share the flow with different proportions according to different bandwidths, so that the load sharing is more reasonable. Assume that the ratio of the bandwidth of the current fourth path to the bandwidth of the fifth path is M: n, the first network node adjusts the ratio of traffic allocated for the fourth path and the fifth path to M: n; m and N are both real numbers greater than 0. For example, T00 in fig. 4 is a first network node, a second path is a path between T00 and T70 in fig. 4, a fourth path is a path between T00 and a01 to T70, and a fifth path is a path between T00 and a02 to T70; before the paths from C11 to T70 are not disconnected, the bandwidths of the fourth path and the fifth path in the routing table of T00 are 200G, and the ratio of the traffic allocated to the fourth path and the fifth path is 1:1; after the paths from C11 to T70 are disconnected, T00 updates the bandwidth of the fourth path in its routing table to 100G, and adjusts the ratio of the traffic allocated to the fourth path and the fifth path to 1:2.

In practical application, each network node in the network can analyze the fault information to obtain the residual bandwidth of the fault path, and the ratio of the traffic distributed to each path is adjusted according to the residual bandwidth of each path, so that the bandwidth utilization rate can be effectively improved, and the congestion of the link can be reduced.

In an alternative implementation, step 503 in fig. 5 may be replaced with the following operations: and discarding the first fault information when the next hop of the first network node does not comprise the first address in the route path from the first network node to the destination network node.

The next hop in the route path from the first network node to the destination network node does not include the first address, indicating that none of the paths through which the first network node sends data to the destination network node passes through the second network node. That is, the first path fails and does not affect the first network node. Since the failure of the first path does not affect the first network node, each path of the first network node is unchanged. So that the first network node does not have to send failure information to its neighboring nodes. The first network node discards the first failure information, which may reduce the spread of the first failure information. In the embodiment of the application, the fault information is kept to be propagated on the routing path, and the fault information which is not on the routing path needs to be discarded (also called pruning). When fault information is diffused, a flooding method (flooding) -pruning-flooding-pruning mode is used, and the first stage are advanced. In the present application, flooding refers to the fact that a failure packet sent from any network node will be sent to all neighboring nodes to that network node (except the node from which the failure message was sent). For example, a01 in fig. 4 receives first fault information from C11, and a01 generates second fault information according to the first fault information and sends the second fault information to T00 and C21. The failure information is just as a clue, and each network node only needs to announce the clue (i.e. the failure information) to the neighbor node, and the neighbor node determines whether to continue to deliver the clue. According to the clues, the fault information is transmitted to all network nodes affected by the fault.

Pruning of fault information during diffusion is described below with the aid of fig. 7. Fig. 7 is a schematic diagram of a fault information diffusion process according to an embodiment of the present application. In fig. 7, A, B, C, D, E, F represents one network node, and a connection line between two points represents a path between network nodes corresponding to the two points. For example, when a in fig. 7 detects a fault, the delivery and pruning of the fault message may include the steps of: (1) The A sends the fault information 1 to each neighbor node (B and C), namely the flooding fault information 1 to the neighbor nodes. (2) C, after receiving the fault information 1 from A, inquiring whether the next hop in the routing table comprises A or not; if yes, send fault information 2 to B, D and E. (3) After receiving the fault information 1 from A, B inquires whether the next hop in the routing table comprises A; if not, discarding the fault information 1; b, after receiving the fault information 2 from C, inquiring whether the next hop in the routing table comprises C or not; if so, fault information 3 is sent to a and D.

The diffusion path and the influence range of the fault information are described below with reference to fig. 8. Fig. 8 is a schematic diagram of another fault information diffusing process according to an embodiment of the present application. The network in fig. 8 is an example of a fabric network, each circle represents a network node, paths between C11 and a71 are disconnected, the diffusion paths of fault information are shown by arrows in fig. 8, a01, T00, T01 and T02 affected by the disconnection of paths between C11 and a71 receive corresponding fault information, fault information transferred in other paths is timely discarded, a01 determines that a path for forwarding data through C11 is faulty according to the received fault information from C11, and T00, T01 and T02 each determine bandwidth reduction conditions of respective paths.

It will be appreciated that by diffusing the fault information to the neighbor nodes and by each neighbor node itself determining whether to continue to communicate the fault information, the fault information can be diffused to all network nodes affected by the fault within a minimum range.

In this implementation manner, the first network node discards the first fault information when the next hop in the routing path from the first network node to the destination network node does not include the first address, which can effectively reduce the spread of the fault information.

The foregoing embodiments do not detail how the second failure information is generated, and the manner in which the second failure information is generated is obtained is described below.

The first network node may perform the following operations to generate the second failure information before sending the second failure information to the third network node: determining a remaining bandwidth of the second path; and generating the second fault information according to the second path and the residual bandwidth of the second path.

The second fault information is also used for resolving to obtain the second bandwidth. Optionally, the second fault information includes a bandwidth field, an interface field, and a prefix field. The bandwidth field in the second failure information is used to indicate the remaining bandwidth of the second path, the prefix field in the second failure information is used to indicate the destination network node, and the interface field in the second failure information is used to indicate the address of the interface where the first network node communicates with the destination network node. Optionally, the first network node determines the remaining bandwidth of the second path as follows: analyzing the first fault information to obtain the residual bandwidth of the first path; taking the residual bandwidth of the first path as the bandwidth of a third path in the routing table; the third path is included in the second path, and a next hop of the third path is the first address; and taking the sum of bandwidths of all paths included in the second path in the routing table as the residual bandwidth of the second path. The second path may refer to all paths between the first network node to the destination network node. The routing table may store bandwidths of the paths included in the second path. Referring to table 1, the state column may include the bandwidths of the paths.

For example, the interface address of the destination network node is 10.1.1.0/24, the second path is each path corresponding to the destination address 10.1.1.0/24 in table 1, and the remaining bandwidth of the second path is the sum of bandwidths of each path corresponding to the destination address 10.1.1.0/24. Also for example, a01 and T70 in fig. 4 represent a first network node and a destination network node, respectively, and for a01 in fig. 4, the second path of a01 to T70 includes two paths, a01-C11-a71-T70 and a01-C21-a71-T70, respectively, and the remaining bandwidth of the second path is the sum of the bandwidths of the two paths. Assuming that the bandwidth between a01 and C11 is 100G and the bandwidth between a01 and C21 is 100G, the remaining bandwidth of the second path is 200G. Assuming that the bandwidth between a01 and C11 is 100G and the bandwidth between a01 and C21 is 100G before the path between C11 and a71 is not broken, the remaining bandwidth of the second path is 100G after the path between C11 and a71 is broken. For another example, a01, C11, T70 in fig. 4 sequentially represent a first network node, a second network node, and a destination network node, where the first path is a path from C11 to T70, and the third path is a01-C11-a71-T70; when the first path does not fail, the bandwidth of the first path is the bandwidth of the third path; after the first path fails, the remaining bandwidth of the first path is the bandwidth of the third path. It will be appreciated that the first network node may update the status and/or the remaining bandwidth of each path in its routing table according to the first failure information, and further send the failure information of the failed path to its neighboring nodes.

The network nodes in the network can be classified into two types, the first is a network node that detects a failure and transmits failure information, and the second is a network node that receives failure information and transmits failure information. In the foregoing embodiment, the first network node is a second network node, and the second network node is the first network node. The foregoing embodiments describe the operations implemented by the first network node during network failure delivery, i.e., the operations implemented by the second network node. The following is an example of a second network node, and describes the operation that can be implemented by the first network node in the network during the network failure transmission process.

Fig. 9 is a schematic diagram of another network failure transmission method according to an embodiment of the present application. As shown in fig. 9, the method may include:

901. And the second network node generates fault information under the condition that the first path fails.

The fault information is used to indicate a fault condition of the first path. The second network node may be the second network node in fig. 5. The manner in which the second network node detects that a certain path (e.g., the first path) in the network fails is a common technical means in the art, and the implementation manner in which the second network node detects that the first path fails will not be described in detail herein.

In some embodiments, the second network node may perform the following operations before performing step 901: the second network node detects that a reference path is failed, the reference path being included in the first path; the second network node sets the state of the reference path in the routing table as a fault; and under the condition that the state of the reference path in the routing table is a fault, determining that the first path is faulty. For example, after the second network node is C11 in fig. 4, C11 detects that the reference path (i.e., the path between C11 and a 71) is disconnected, C11 sets the state of the reference path in its routing table to be a failure. In practical application, a network node needs to query the state of a certain path in its routing table before forwarding or transmitting data through the path; if the state of the path is a fault, the data is not forwarded or sent through the path; if the state of the path is a non-fault state, forwarding or transmitting data through the path. The bandwidth available to the first path may be reduced due to any path included in the first path failing. Thus, when at least one path included in the first path fails, the first path fails. After detecting that a certain path fails, the second network node can update the state of the path in the routing table in time so as to quickly determine the state of each path according to the updated routing table and transmit corresponding failure information.

In some embodiments, the second network node implements step 901 in the following manner: the second network node takes the sum of bandwidths of all paths included in the first path in the routing table as the residual bandwidth of the first path under the condition that the first path fails; the fault information is generated based on the first path and the remaining bandwidth of the first path. In this way, the second network node can quickly and accurately obtain the residual bandwidth of the first path, thereby generating fault information which can indicate the residual bandwidth of the first path.

902. And sending fault information to the first network node.

The first network node is a neighbor node of the second network node. The fault information may be the first fault information described above. The failure information includes a bandwidth field indicating a remaining bandwidth of the first path. The first path may be a path from the second network node to a destination network node. The failure information further includes an interface field for indicating an address of an interface through which the second network node communicates with a destination network node, and a prefix field for indicating the destination network node. Since the foregoing embodiments have detailed the content contained in the first failure information, a description thereof will not be repeated here.

The application of the network failure transmission method provided by the embodiment of the application in the actual scene is further described by two more complete embodiments.

First, a specific example of the application of the network failure delivery method provided by the application in the fabric network will be described.

Fig. 10A is a schematic diagram of a fault delivery process in a fabric network according to an embodiment of the present application. In fig. 10A, the fabric network includes 10 network nodes, each circle represents one network node, C11 corresponds to a second network node, a01 corresponds to a first network node, T00 corresponds to a third network node, T70 corresponds to a destination network node, 1001 is first failure information, 1002 is second failure information, 10.1.1.0/24 is an interface address of T70, 16.1.1.8/24 is an interface address of a71, 20.1.1.2/32 is an interface address of C11, 15.1.1.2/32 is an interface address of a01, 9.1.1.2/32 is an interface address of T00, 30.1.1.2/32 is an interface address of C21, and 16.1.1.2/32 is an interface address of a 02. FIG. 10B is a schematic diagram illustrating a fault delivery process in another fabric network according to an embodiment of the present application. The network in fig. 10B and the network in fig. 10A are the same network. In fig. 10B, 1003 is a partial path in the routing table of C11, 1004 is a partial path in the routing table of a01, 1005 is a partial path in the routing table of T00, and the traffic direction (i.e., the data transmission direction) is T00 to T70. Referring to fig. 10A and 10b, one example of fault delivery in a fabric network is as follows:

(1) C11 detects a break in the path between C11 and a 71.

(2) C11 updates the state of the path whose destination address is 10.1.1.0/24 and whose next hop is 16.1.1.8/24 in its routing table as a failure.

The path of C11 with destination address 10.1.1.0/24 and next hop 16.1.1.8/24 in the routing table is the path of C11 through a71 to T70 (corresponding to the reference path described above).

(3) C11 sends first failure information to a 01.

1001 In fig. 10A is one example of the first failure information. Where 01 denotes the sequence number of the first failure information, and the remaining bandwidth 0 denotes that the remaining bandwidth of the path from C11 to T70 is 0, 20.1.1.2/32 is an interface address of C11, and 10.1.1.0/24 is an interface address of T70. It can be seen that the interface field and the prefix field in this first failure information may indicate a path of C11 to T70 (corresponding to the first path described above).

(4) A01 updates the state of the path with the destination address 10.1.1.0/24 and the next hop 20.1.1.2/32 in the routing table to the fault according to the first fault information.

The path of A01 with destination address 10.1.1.0/24 and next hop 20.1.1.2/32 in the routing table is the path of A01 through C11 to A70. 20.1.1.2/32 is an interface address of C11. In fig. 10B, 1004 is one path a01 through C11 to a70, and the other path a01 through C21 to a 70.

(5) A01 sends a second fault message to T00.

1002 In fig. 10A is one example of this second failure information. Where 01 denotes the sequence number of the second failure information, and the residual bandwidth 100 denotes an interface address of 100g,15.1.1.2/32 a01, and 10.1.1.0/24T 70 for the paths a01 to T70. It can be seen that the interface field and the prefix field in this second failure information may indicate the path of a01 to T70 (corresponding to the second path described above). Before the path between C11 and a71 is not broken, the bandwidth between C11 and a71 is 100G, and the bandwidth between C21 and a71 is 100G; after the path between C11 and a71 is disconnected, the bandwidth between C11 and a71 is 0G, the bandwidth between C21 and a71 is 100G, and the remaining bandwidth of the path of a01 to T70 is 100G. Therefore, the remaining bandwidth in the second failure information is 100G.

(6) T00 updates the state of the path with the destination address 10.1.1.0/24 and the next hop 15.1.1.2/32 in the routing table to be halved according to the second fault information.

The path of T00 with destination address 10.1.1.0/24 and next hop 15.1.1.2/32 in the routing table is the path of T00 through A01 to A70. 15.1.1.2/32 is an interface address of A01. In fig. 10B, 1005 has one path T00 passing through a01 to T70 and the other path T00 passing through a02 to T70. Before the path between C11 and A71 is not broken, the bandwidth between A01 and T70 is 200G, and the bandwidth of the path of T00 through A01 to T70 is 200G; after the path between C11 and a71 is broken, the bandwidth between a01 and T70 is 100G, and the bandwidth of the path of T00 through a01 to T70 is 100G. Thus, T00 updates the state of the path whose destination address is 10.1.1.0/24 and whose next hop is 15.1.1.2/32 in its routing table to half the bandwidth.

(7) T00 adjusts the ratio of the traffic of the data transmitted by A01 and A02 from 1:1 to 1:2.

Before the path between C11 and A71 is not broken, the bandwidth of the path of T00 through A01 to T70 is 200G, the bandwidth of the path of T00 through A02 to T70 is 200G, and thus the ratio of traffic of T00 transmitting data through A01 and A02 is 1:1. After the path between C11 and A71 is broken, the bandwidth of the path of T00 through A01 to T70 is 100G, the bandwidth of the path of T00 through A02 to T70 is 200G, and the ratio of the traffic of the T00 through A01 and A02 can be adjusted from 1:1 to 1:2 by UCMP. It should be appreciated that the network node allocates traffic for each path according to the bandwidth of each path, which may effectively alleviate congestion.

According to the embodiment, each network node in the network can obtain the fault information of the residual bandwidth by analyzing the fault information transmitted to the neighbor nodes, and each network node affected by the fault in the network can update the routing table of the network node rapidly and adjust the flow distribution of each path in time according to the residual bandwidth of each path.

Next, a specific example of the application of the network failure delivery method provided by the present application in another fabric network will be described.

FIG. 11 is a schematic diagram illustrating a fault delivery process in another fabric network according to an embodiment of the present application. Network nodes inside the fabric network configure interior gateway protocols (Interior Gateway Protocol, IGP), and network nodes at the edge configure both IGP and border gateway protocols (Border Gateway Protocol, BGP). IGP is a protocol that exchanges routing information between gateways (hosts and routers) within an autonomous network. BGP is a core de-centralized autonomous routing protocol on the internet for exchanging routing information between different autonomous systems (Autonomous System, AS). When two ases need to exchange routing information, each AS must designate a BGP-running node to exchange routing information with other ases on behalf of the AS. In fig. 11, S01 and S03 are network nodes at the same AS edge, B01 and B02 are network nodes inside the AS, S04 is a network node at another AS edge, 40.1.1.2/32 is an interface address of S04, 60.1.1.2/32 is an interface address of S03, 11.1.1.8/32 is an interface address of B01, 12.1.1.8/32 is an interface address of B02, 1101 is a partial path in an IGP routing table of B01, 1102 is a partial path in an IGP routing table of S01, 1103 is a partial path in a BGP routing table of S01. Referring to fig. 11, one example of fault delivery and handling in a fabric network is as follows:

(1) B01 detects a path break between B01 and S03.

(2) B01 updates the state of the path whose destination address is 60.1.1.2/32 and whose next hop is 60.1.1.2/32 in its IGP routing table as a failure.

The path in the IGP routing table of B01 with destination address 60.1.1.2/32 and next hop 60.1.1.2/32 is the path between B01 and S03. 60.1.1.2/32 is an interface address of S03, i.e. 60.1.1.2/32 is used to indicate S03. 1101 in fig. 11 is the path between B01 and S03 in the IGP routing table of B01.

(3) And B01 searches an IGP route table, generates fault information (also called fault notification message) according to the IGP route with the destination address 60.1.1.2/32, and notifies network nodes in the IGP domain. The failure information is used to indicate failure conditions of the paths of B01 to S03. The method of propagation and pruning of fault messages is the same as in the previous embodiments. A network node within an IGP domain refers to a network node within the network, i.e. a network node that configures only IGPs.

(4) S01 updates the state of the path with the destination address of 60.1.1.2/32 and the next hop of 11.1.1.8/32 in the IGP routing table to be the fault after receiving the fault information.

1102 In fig. 11 are two paths in the IGP routing table of S01, the former path being the path of S01 through B01 to S03, and the next path being the path of S01 through B02 to S03. After the path between B01 and S03 is disconnected, S01 fails through the path of B01 to S03, and S01 fails through the path of B02 to S03. Therefore, upon receiving the failure information, S01 updates the state of the path of S01 through B01 to S03 to failure.

(5) S01 checks the BGP routing table, performs fault processing on the route with the next hop of 60.1.1.2/32 in the BGP routing table, and inherits the fault state information of 60.1.1.2/32.

In FIG. 11, 1103 is the path with the next hop 60.1.1.2/32 in the BGP routing table for S01. S01 can update the states of all paths and the residual bandwidths in the IGP routing table rapidly. How to update BGP routing tables based on IGP routing tables is a common technical approach in the art and will not be described in detail herein. In this embodiment, IGP routing is protected by the failure delivery method provided by the embodiment of the present application; BGP routes are protected by iterating the state of IGP routes.

In the embodiment of the application, the network node affected by the fault actively generates the fault message by checking the IGP routing table, so that the propagation speed of the fault message in the IGP network node can be accelerated, and the BGP routing depending on the IGP can sense the fault more quickly.

Fig. 12 is a schematic structural diagram of a first network node according to an embodiment of the present application. As shown in fig. 12, the first network node may include:

a receiving unit 1201, configured to receive first failure information from a second network node; the second network node is a neighbor node of the first network node, the first fault information is used for indicating a fault condition of a first path, and the first path is a path from the second network node to a destination network node;

A parsing unit 1202, configured to parse the first failure information to obtain the destination network node and a first address; the first address is an interface address of the second network node;

A sending unit 1203, configured to send second fault information to a third network node when a next hop of the first network node includes the first address in a routing path from the first network node to the destination network node; the third network node is a neighboring node of the first network node, and the second fault information is used to indicate a fault condition of a second path, where the second path is a path from the first network node to the destination network node.

Referring to fig. 12, the first network node may store a port list, and each port (also called an interface) in the port list is coupled to ports of other network nodes, and the sending unit 1203 diffuses the fault information by querying the port list. For example, a first port in the port list is coupled to one port of the third network node, and the sending unit 1203 outputs the second fault information through the first port, so that the third network node may receive the second fault information.

In an alternative implementation, as shown in fig. 12, the first network node further includes:

A discarding unit 1204, configured to discard the first failure information when the next hop of the first network node does not include the first address in a routing path from the first network node to the destination network node.

In an alternative implementation, the first failure information includes a bandwidth field that indicates a remaining bandwidth of the first path.

In an alternative implementation, the network node further comprises:

a determining unit 1205 for determining a remaining bandwidth of the second path;

A first generating unit 1206, configured to generate the second fault information according to the second path and a remaining bandwidth of the second path; the second fault information is also used for resolving to obtain the second bandwidth.

In an optional implementation manner, the parsing unit 1202 is further configured to parse the first failure information to obtain a residual bandwidth of the first path;

A determining unit 1205, configured to specifically use the remaining bandwidth of the first path as the bandwidth of the third path in the routing table; taking the sum of bandwidths of all paths included in the second path in the routing table as the residual bandwidth of the second path; the third path is included in the second path, and a next hop of the third path is the first address.

A query unit 1207, configured to query a route path from the first network node to the destination network node in the routing table, where a next hop of the first network node is a path of the first address to obtain a fourth path; the fourth path is included in the second path;

An updating unit 1208, configured to update a state of the fourth path in the routing table and/or update a bandwidth of the fourth path to a remaining bandwidth of the first path according to the remaining bandwidth of the first path; the state of the fourth path is used to determine the bandwidth of the fourth path.

A traffic allocation unit 1209, configured to determine a bandwidth of a fifth path in the routing table, where the fifth path is included in the second path; and distributing flow for the fourth path and the fifth path according to the bandwidth of the fourth path and the bandwidth of the fifth path by adopting a non-equivalent load sharing UCMP technology.

a second generation unit 1210 configured to generate third failure information in a case where the sixth path fails; the sixth path is a path from the first network node to a reference network node, and the third fault information is used for analyzing and obtaining the residual bandwidth of the sixth path;

the sending unit 1203 is further configured to send the third failure information to the third network node. The second generation unit 1210 and the first generation unit 1206 may be the same unit or different units.

Fig. 13 is a schematic structural diagram of a second network node according to an embodiment of the present application. As shown in fig. 13, the second network node may include:

A generating unit 1301 configured to generate failure information when the first path fails;

A transmitting unit 1302, configured to transmit the fault information to a first network node; the fault information is used for indicating the fault condition of a first path, and the first network node is a neighbor node of the second network node; the failure information includes a bandwidth field indicating a remaining bandwidth of the first path.

In an alternative implementation, the first path is a path from the second network node to a destination network node; the failure information further comprises an interface field for indicating an interface address of the second network node and a prefix field for indicating the destination network node.

In an alternative implementation, as shown in fig. 13, the second network node further includes:

A detecting unit 1303, configured to detect that a reference path is failed, where the reference path is included in the first path;

a setting unit 1304 for setting the state of the reference path in the routing table as a failure;

a determining unit 1305, configured to determine that the first path is faulty when the state of the reference path in the routing table is faulty.

In an optional implementation manner, the generating unit 1301 is specifically configured to, in a case where the first path fails, take, as a remaining bandwidth of the first path, a sum of bandwidths of paths included in the first path in the routing table; the fault information is generated based on the first path and the remaining bandwidth of the first path.

The first network node and the second network node are distinguished from the perspective of perceiving the failure and communicating the failure. In the network, the network node which perceives the fault is a second network node, and the network node which transmits the fault is a first network node. It should be appreciated that in some embodiments, each network node in the network may be provided with both the functionality of the second network node (fault-aware) and the functionality of the first network node (fault-delivery). That is, one network node may implement the operation of both the first network node and the second network node. In these embodiments, the generating unit 1301 may be the same unit as the first generating unit 1206 and/or the second generating unit 1210, and the transmitting unit 1302 and the transmitting unit 1203 may be the same unit.

Two practical application scenarios of the network node provided by the embodiment of the present application are described below.

Fig. 14A is an application scenario schematic diagram of a network failure delivery method according to an embodiment of the present application. As shown in fig. 14A, the network nodes 1 to 5 form a network, the structure of each network node is shown in fig. 14B, the network failure transmission method provided in the foregoing embodiment is effective on the forwarding plane, and the functions are implemented on the network processor (Network Processor, NP) without the participation of the control plane. That is, the NP in the network node implements the network failure delivery method, i.e., the NP is used to implement the functions of the units in fig. 12 and 13.

Fig. 15 is a schematic application scenario diagram of another network failure transmission method according to an embodiment of the present application. Fig. 15 can also be regarded as an internal connection structure of a switching network of network nodes. In the switching network of the network node, the fault transmission between the components of the switching network may also use the network fault transmission method provided in the foregoing embodiment, where the functions are implemented on the components of the switching network. Take the example of a switching network structure of a switching network Interface (FIC) +switching unit (SWITCHING ELEMENT, SE):

The FIC is an interface identifiable by the switching network, and the interface of the network node finally corresponds to a certain FIC for data exchange. One SE can connect all FICs and increasing SE can increase switching capacity. Both the SE and the FIC in the network node may implement the network failure delivery method, and both the SE and the FIC are used to implement the functions of the units in fig. 12 and fig. 13.

It should be understood that the above division of the units in the network node is merely a division of a logic function, and may be fully or partially integrated into a physical entity or may be physically separated. For example, the above units may be processing elements set up separately, may be implemented in a certain chip of the terminal, or may be stored in a memory element of the controller in the form of program codes, and the functions of the above units may be called and executed by a certain processing element of the processor. In addition, the units can be integrated together or can be independently realized. The processing element here may be an integrated circuit chip with signal processing capabilities. In implementation, each step of the above method or each unit above may be implemented by an integrated logic circuit of hardware in a processor element or an instruction in the form of software. The processing element may be a general-purpose processor, such as a network processor or a central processing unit (English: central processing unit; CPU), or may be one or more integrated circuits configured to implement the above methods, such as: one or more application-specific integrated circuits (ASIC), or one or more microprocessors (DSP), or one or more field-programmable gate arrays (FPGA), etc.

Referring to fig. 16, fig. 16 is a schematic structural diagram of another network node according to an embodiment of the present application, where the network node includes a processor 1601, a memory 1602, and an input/output device 1603, and the processor 1601, the memory 1002, and the input/output device 1603 are connected to each other by a bus 1604.

Memory 1602 includes, but is not limited to, random access memory (random access memory, RAM), read-only memory (ROM), erasable programmable read-only memory (erasable programmable read only memory, EPROM), or portable read-only memory (compact disc read-only memory, CD-ROM), with memory 1602 for associated instructions and data (including routing tables). The processor 1601 may be a network processor or other processor, such as a CPU. And an input/output device 1603 for transmitting/receiving data and fault information.

The processor 1601 in the network node reads the program code stored in the memory 1602, and may perform the following operations: receiving first fault information from a second network node; the second network node is a neighbor node of the first network node, the first fault information is used for indicating a fault condition of a first path, and the first path is a path from the second network node to a destination network node; analyzing the first fault information to obtain the target network node and a first address; the first address is the address of an interface where the second network node communicates with the destination network node; in a route path from the first network node to the destination network node, when the next hop of the first network node comprises the first address, sending second fault information to a third network node; the third network node is a neighboring node of the first network node, and the second fault information is used to indicate a fault condition of a second path, where the second path is a path from the first network node to the destination network node.

The processor 1601 in the network node reads the program code stored in the memory 1602, and may also perform the following operations: generating fault information under the condition that a first path fails; transmitting the fault information to a first network node; the failure information is used for indicating the failure condition of the first path, the first network node is a neighbor node of the second network node, and the failure information comprises a bandwidth field, and the bandwidth field is used for indicating the residual bandwidth of the first path.

In the network node depicted in fig. 16, an input-output device 1603 may be used to implement the functionality of the receiving unit 1201 and the transmitting unit 1203 shown in fig. 12; the processor 1601 may be further configured to perform the functions of the parsing unit 1202, the discarding unit 1204, the determining unit 1205, the first generating unit 1206, the querying unit 1207, the updating unit 1208, the traffic distribution unit 1209, and the second generating unit 1210 shown in fig. 12. The input-output device 1603 is also used to realize the function of the transmission unit 1302 in fig. 13; the processor 1601 may also be used to perform the functions of the generating unit 1301, the detecting unit 1303, the setting unit 1304, and the determining unit 1305 shown in fig. 13.

The embodiment of the application also provides a computer readable storage medium, wherein instructions are stored in the computer readable storage medium, when the computer readable storage medium runs on a computer, the computer is caused to execute the network fault transmission method provided by the embodiment.

Optionally, the instructions may be implemented when run on a computer: receiving first fault information from a second network node; the second network node is a neighbor node of the first network node, the first fault information is used for indicating a fault condition of a first path, and the first path is a path from the second network node to a destination network node; analyzing the first fault information to obtain the target network node and a first address; the first address is the address of an interface where the second network node communicates with the destination network node; in a route path from the first network node to the destination network node, when the next hop of the first network node comprises the first address, sending second fault information to a third network node; the third network node is a neighboring node of the first network node, and the second fault information is used to indicate a fault condition of a second path, where the second path is a path from the first network node to the destination network node.

Optionally, the above instructions when run on a computer may further implement: generating fault information under the condition that a first path fails; transmitting the fault information to a first network node; the failure information is used for indicating the failure condition of the first path, the first network node is a neighbor node of the second network node, and the failure information comprises a bandwidth field, and the bandwidth field is used for indicating the residual bandwidth of the first path.

An embodiment of the present application provides a computer program product containing instructions that, when run on a computer, cause the computer to perform the network failure delivery method provided by the foregoing embodiment.

In the above embodiments, it may be implemented in whole or in part by software, hardware, firmware, or any combination thereof. When implemented in software, may be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions. When loaded and executed on a computer, produces a flow or function in accordance with embodiments of the present invention, in whole or in part. The computer may be a general purpose computer, a special purpose computer, a computer network, or other programmable apparatus. The computer instructions may be stored in or transmitted from one computer-readable storage medium to another, for example, by wired (e.g., coaxial cable, optical fiber, digital Subscriber Line (DSL)), or wireless (e.g., infrared, wireless, microwave, etc.). The computer readable storage medium may be any available medium that can be accessed by a computer or a data storage device such as a server, data center, etc. that contains an integration of one or more available media. The usable medium may be a magnetic medium (e.g., floppy disk, hard disk, tape), an optical medium (e.g., DVD), or a semiconductor medium (e.g., solid state disk Solid STATE DISK (SSD)), etc.

While the invention has been described with reference to certain preferred embodiments, it will be understood by those skilled in the art that various changes and substitutions of equivalents may be made and equivalents will be apparent to those skilled in the art without departing from the scope of the invention. Therefore, the protection scope of the invention is subject to the protection scope of the claims.

Claims

1. A method of network failure delivery, comprising:

The first network node receives first fault information from the second network node; the second network node is a neighbor node of the first network node, the first fault information is used for indicating a fault condition of a first path, and the first path is a path from the second network node to a destination network node;

Analyzing the first fault information to obtain the target network node and a first address; the first address is the address of an interface where the second network node communicates with the destination network node;

In a route path from the first network node to the destination network node, when a next hop of the first network node comprises the first address, the first network node sends second fault information to a third network node; the third network node is a neighboring node of the first network node, the second fault information is used to indicate a fault condition of a second path, and the second path is a path from the first network node to the destination network node.

2. The method of claim 1, wherein after said resolving said first failure information to obtain said destination network node and a first address, the method further comprises:

And discarding the first fault information when the next hop of the first network node does not comprise the first address in a routing path from the first network node to the destination network node.

3. The method of claim 1, wherein the first failure information includes a bandwidth field indicating a remaining bandwidth of the first path.

4. A method according to claim 3, wherein before said sending the second failure information to the third network node, the method further comprises:

the first network node determining a remaining bandwidth of the second path;

Generating the second fault information according to the second path and the residual bandwidth of the second path; the second fault information is also used for analyzing to obtain a second bandwidth.

5. The method of claim 4, wherein the first network node determining the remaining bandwidth of the second path comprises:

the first network node analyzes the first fault information to obtain the residual bandwidth of the first path;

The first network node takes the residual bandwidth of the first path as the bandwidth of a third path in a routing table; the third path is included in the second path, and a next hop of the third path is the first address;

And taking the sum of bandwidths of all paths included in the second path in the routing table as the residual bandwidth of the second path.

6. The method according to any one of claims 1 to 5, wherein after said resolving said first failure information to obtain said destination network node and a first address, the method further comprises:

The first network node queries a route path from the first network node to the destination network node in a route table, and the next hop of the first network node is a path of the first address to obtain a fourth path; the fourth path is included in the second path;

Updating the state of the fourth path in the routing table and/or updating the bandwidth of the fourth path to the residual bandwidth of the first path according to the residual bandwidth of the first path; the state of the fourth path is used to determine the bandwidth of the fourth path.

7. The method of claim 6, wherein after updating the state of the fourth path in the routing table and/or updating the bandwidth of the fourth path to the remaining bandwidth of the first path according to the remaining bandwidth of the first path, the method further comprises:

the first network node determines the bandwidth of a fifth path in the routing table, wherein the fifth path is contained in the second path;

and the first network node adopts a non-equivalent load sharing UCMP technology to distribute flow for the fourth path and the fifth path according to the bandwidth of the fourth path and the bandwidth of the fifth path.

8. The method according to any one of claims 1 to 5, further comprising:

The first network node generates third fault information under the condition that a sixth path fails; the sixth path is a path from the first network node to a reference network node, and the third fault information is used for analyzing and obtaining the residual bandwidth of the sixth path;

the first network node sends the third fault information to the third network node.

9. A first network node, comprising:

A receiving unit, configured to receive first failure information from a second network node; the second network node is a neighbor node of the first network node, the first fault information is used for indicating a fault condition of a first path, and the first path is a path from the second network node to a destination network node;

the analyzing unit is used for analyzing the first fault information to obtain the target network node and the first address; the first address is the address of an interface where the second network node communicates with the destination network node;

A sending unit, configured to send second fault information to a third network node when a next hop of the first network node includes the first address in a routing path from the first network node to the destination network node; the third network node is a neighboring node of the first network node, the second fault information is used to indicate a fault condition of a second path, and the second path is a path from the first network node to the destination network node.

10. The first network node of claim 9, wherein the first network node further comprises:

And the discarding unit is used for discarding the first fault information when the next hop of the first network node does not comprise the first address in a route path from the first network node to the destination network node.

11. The first network node of claim 9, wherein the first failure information includes a bandwidth field indicating a remaining bandwidth of the first path.

12. The first network node of claim 11, wherein the network node further comprises:

A determining unit configured to determine a remaining bandwidth of the second path;

a first generating unit, configured to generate the second fault information according to the second path and a remaining bandwidth of the second path; the second fault information is also used for analyzing to obtain a second bandwidth.

13. The first network node of claim 12, wherein the first network node,

The analyzing unit is further configured to analyze the first failure information to obtain a residual bandwidth of the first path;

The determining unit is specifically configured to take the remaining bandwidth of the first path as the bandwidth of a third path in the routing table; taking the sum of bandwidths of all paths included in the second path in the routing table as the residual bandwidth of the second path; the third path is included in the second path, and a next hop of the third path is the first address.

14. The first network node according to any of claims 9 to 13, characterized in that the first network node further comprises:

a query unit, configured to query a route path from the first network node to the destination network node in a route table, where a next hop of the first network node is a path of the first address to obtain a fourth path; the fourth path is included in the second path;

An updating unit, configured to update a state of the fourth path in the routing table and/or update a bandwidth of the fourth path to a remaining bandwidth of the first path according to the remaining bandwidth of the first path; the state of the fourth path is used to determine the bandwidth of the fourth path.

15. The first network node of claim 14, wherein the first network node further comprises:

A flow distribution unit, configured to determine a bandwidth of a fifth path in the routing table, where the fifth path is included in the second path; and distributing flow for the fourth path and the fifth path according to the bandwidth of the fourth path and the bandwidth of the fifth path by adopting a non-equivalent load sharing UCMP technology.

16. The first network node according to any of claims 9 to 13, characterized in that the first network node further comprises:

a second generation unit configured to generate third failure information in the case where the sixth path fails; the sixth path is a path from the first network node to a reference network node, and the third fault information is used for analyzing and obtaining the residual bandwidth of the sixth path;

the sending unit is further configured to send the third failure information to the third network node.