Fat tree-based network failure recovery

By detecting the fault link in the fat tree network and blocking its associated ports, sending a fault notification message to the remote nodes, and updating the routing table using the equivalent multi-path routing table, the rapid failure recovery of the data center network is achieved and the existing solution is solved The long recovery time of failure in technology reduces cost and complexity and avoids packet loss.

CN118923097APending Publication Date: 2024-11-08HUAWEI TECH CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202380025104.5
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2023-03-07
Publication Date
2024-11-08

AI Technical Summary

Technical Problem

Existing network technologies cannot quickly identify and recover link failures in data center networks, resulting in service interruptions and data packet loss, and cannot achieve seamless failure recovery in milliseconds.

Method used

Adopt the fault recovery mechanism based on fat tree network topology, by detecting the fault link, blocking the ports associated with the fault link, and sending fault notification messages to the remote nodes, and using equivalent multi-path routing tables to update the routing table to avoid fault links , realize traffic rerouting of local and remote nodes.

Benefits of technology

Reduces the cost and network complexity of using additional professional hardware, achieves continuity of connections and packet loss avoidance, enables repair of failures in milliseconds, and ensures continuity of services.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118923097A_ABST
    Figure CN118923097A_ABST
Patent Text Reader

Abstract

In some examples, there is provided a fault recovery method in an IP network, the network comprising a plurality of nodes, the network comprising a logical topology representing hierarchical relationships between respective ones of the plurality of nodes, the logical topology comprising a plurality of logical layers for the network, the method includes detecting a faulty link between a first node of the plurality of nodes and a second node of the plurality of nodes, where the first node is associated with a first logical layer of the network and the second node is associated with a second logical layer of the network, the second layer comprises a layer logically higher than the first layer of the topology; in response to detecting the faulty link, the first node of the plurality of nodes blocks a port associated with the faulty link between the first node of the plurality of nodes and the second node of the plurality of nodes.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention generally relates to fast recovery in a network environment. Various aspects of the present invention relate to fast recovery in a fat-tree based network. Background Art

[0002] Data center networks may contain tens or hundreds of thousands of nodes, so they may frequently experience hardware and software failures as well as link congestion. Solving network failures has been recognized as a problem of great significance to security, stability, and prosperity.

[0003] In data center networks, packets are routed along the shortest paths with sufficient resources to promote efficient network utilization and minimize latency. Typically, traditional recovery schemes identify link failures and recalculate routes within seconds or even minutes. Current network standards are unable to meet the challenge of avoiding failures quickly to ensure seamless integrity of running computations. In order to maintain service continuity, faster recovery schemes are needed that can repair failures in milliseconds. Summary of the invention

[0004] The present invention aims to provide a fast recovery mechanism for fault recovery in a network environment.

[0005] These and other objects are achieved by the features of the independent claims.

[0006] Further implementations are apparent from the dependent claims, the description and the drawings.

[0007] A first aspect of the present invention provides a fault recovery method in a fat-tree based IP network, the network comprising a plurality of nodes, the network comprising a logical topology representing a hierarchical relationship between each of the plurality of nodes, the logical topology comprising a plurality of logical layers for the network, the method comprising detecting a faulty link between a first node of the plurality of nodes and a second node of the plurality of nodes, wherein the first node is associated with a first logical layer of the network, the second node is associated with a second logical layer of the network, the second layer comprising a layer of the topology that is logically higher than the first layer; in response to detecting the faulty link, the first node of the plurality of nodes blocks a port associated with the faulty link between the first node of the plurality of nodes and the second node of the plurality of nodes.

[0008] Therefore, an effective recovery mechanism is provided for failure recovery of IP networks (such as data centers). Since the mechanism is based on the inherent characteristics of the fat tree network topology, the cost associated with the use of additional specialized hardware and the complexity of the network can be reduced. In addition, by providing a recovery mechanism in the event of a failure, the continuity of the connection can be achieved and packet loss can be avoided.

[0009] In an implementation of the first aspect, blocking the port associated with the failed link may include: the first node among the multiple nodes updates an equal-cost multi-path (ECMP) routing table associated with the first node among the multiple nodes by deleting the port associated with the failed link from the table, thereby blocking the port. Thus, the first node can locally reroute traffic to avoid the failed link.

[0010] The method may also include the following steps: in response to detecting the faulty link, the second node of the plurality of nodes sends a fault notification message to at least one remote node of a plurality of remote nodes, the plurality of remote nodes being associated with the first logical layer of the network. Thus, the remote node may be notified of the fault, enabling the remote node to respond to the notification and take action as needed.

[0011] Sending the fault notification message may include: the second node among the plurality of nodes identifies the at least one remote node among the plurality of remote nodes, the at least one remote node among the plurality of remote nodes is configured to send traffic through the faulty link, generates the fault notification message including an Internet Control Message Protocol (ICMP) data packet, and sends the fault notification message to the identified remote node. Therefore, an existing, widely used protocol may be used without inventing a new protocol to notify a remote node of a fault.

[0012] Identifying the at least one remote node that is set to send traffic through the faulty link may include: using an equal-cost multi-path (ECMP) routing table associated with the second node among the multiple nodes, identifying a set of Internet Protocol (IP) addresses that are set to send traffic through the faulty link.

[0013] The method may further include the following steps: in response to receiving the fault notification message, the remote node blocks the port that receives the fault notification message. Therefore, the port where the faulty link is located can be easily identified and blocked.

[0014] Blocking the port that receives the fault notification message may include: the remote node updates an ECMP routing table associated with the remote node by deleting the port that receives the fault notification message, thereby blocking the port. Therefore, the remote node can reroute traffic to avoid the faulty link.

[0015] The IP network may include a fat-tree based IP network.

[0016] A second aspect of the present invention provides an IP network comprising multiple nodes, the network comprising a logical topology representing a hierarchical relationship between corresponding nodes among the multiple nodes, the logical topology comprising multiple logical layers of the network, the network comprising a first node among the multiple nodes associated with a first logical layer of the network, a second node among the multiple nodes associated with a second logical layer of the network, the second layer comprising a layer of the topology that is logically higher than the first layer, wherein, in response to detecting a faulty link between the first node among the multiple nodes and the second node among the multiple nodes, the first node among the multiple nodes is used to block a port associated with the faulty link between the first node among the multiple nodes and the second node among the multiple nodes.

[0017] Therefore, an effective recovery mechanism is provided for failure recovery of IP networks (such as data centers). Since the mechanism is based on the inherent characteristics of the fat tree network topology, the cost associated with the use of additional specialized hardware and the complexity of the network can be reduced. In addition, by providing a recovery mechanism in the event of a failure, the continuity of the connection can be achieved and packet loss can be avoided.

[0018] In an implementation of the second aspect, the first node of the plurality of nodes may also be configured to update an equal-cost multi-path (ECMP) routing table by deleting the port associated with the failed link from the table, thereby blocking the port. Thus, the first node may locally reroute traffic to avoid the failed link.

[0019] The network may also include a plurality of remote nodes associated with the first layer of the network, wherein the second node of the plurality of nodes is further configured to send a fault notification message to at least one of the plurality of remote nodes. Thus, the remote node may be notified of the fault, enabling the remote node to respond to the notification and take action as required.

[0020] The second node among the plurality of nodes may also be used to identify the at least one remote node among the plurality of remote nodes, the at least one remote node among the plurality of remote nodes being configured to send traffic through the faulty link, generate the fault notification message including an Internet Control Message Protocol (ICMP) data packet, and send the fault notification message to the identified remote node. Therefore, an existing, widely used protocol may be used without inventing a new protocol to notify remote nodes of a fault.

[0021] The second node of the plurality of nodes may also be used to use its ECMP routing table to identify a set of Internet Protocol (IP) addresses configured to send traffic through the faulty link, thereby identifying the at least one remote node to which the fault notification message is to be sent. Thus, the remote node that sends traffic using the faulty link may be identified in a simple manner.

[0022] The remote node may also be used to block the port that receives the fault notification message in response to receiving the fault notification message. Therefore, the port where the faulty link is located can be easily identified and blocked.

[0023] The remote node may also be configured to update its ECMP routing table by deleting the port that received the fault notification message, thereby blocking the port. Thus, the remote node may reroute traffic to avoid the faulty link.

[0024] The IP network may include a fat-tree based IP network.

[0025] A third aspect of the invention provides a data center comprising a fat-tree based network as described herein.

[0026] A fourth aspect of the present invention provides a computer-readable storage medium, the computer-readable storage medium comprising a computer program code, the computer program code being accessible by a device comprising a processor to provide instructions and / or data to the device, the computer program code being used, together with the processor, to enable the device to detect, in an IP network comprising a plurality of nodes, a faulty link between a first node among the plurality of nodes and a second node among the plurality of nodes, the network comprising a logical topology representing a hierarchical relationship between corresponding nodes among the plurality of nodes, the logical topology comprising a plurality of logical layers of the network, wherein the first node is associated with a first logical layer of the network, the second node is associated with a second logical layer of the network, the second layer comprising a layer of the topology that is logically higher than the first layer; in response to detecting the faulty link, the first node among the plurality of nodes blocks a port associated with the faulty link between the first node among the plurality of nodes and the second node among the plurality of nodes.

[0027] The above method provides an effective recovery mechanism for failure recovery of IP networks (such as data centers). Since the mechanism is based on the inherent characteristics of the fat tree network topology, the cost associated with the use of additional specialized hardware and the complexity of the network can be reduced. In addition, by providing a recovery mechanism in the event of a failure, the continuity of the connection can be achieved and packet loss can be avoided.

[0028] In an implementation of the fourth aspect, the computer program may also be used to cause the device, through the processor, to delete the port associated with the failed link from an equal-cost multi-path (ECMP) routing table associated with the first node among the multiple nodes, thereby causing the first node among the multiple nodes to update the table, thereby blocking the port. Therefore, the first node can locally reroute traffic to avoid the failed link.

[0029] The computer program may also be configured to cause the apparatus, through the processor, to send a fault notification message from the second node of the plurality of nodes to at least one remote node of a plurality of remote nodes associated with the first layer of the network in response to detecting the faulty link, so that the remote node can be notified of the fault, enabling the remote node to respond to the notification and take action as needed.

[0030] The computer program may also be used to, through the processor, cause the apparatus to identify at least one of the plurality of remote nodes through the second node of the plurality of nodes, at least one of the plurality of remote nodes being configured to send traffic through the faulty link, generate the fault notification message including an Internet Control Message Protocol (ICMP) data packet, and send the fault notification message to the identified remote node. Therefore, an existing, widely used protocol may be used to notify remote nodes of a fault without inventing a new protocol.

[0031] The computer program may also be used to cause the apparatus, through the processor, to use an equal-cost multi-path (ECMP) routing table associated with the second node of the plurality of nodes to identify a set of Internet Protocol (IP) addresses configured to send traffic through the faulty link, thereby identifying the at least one remote node. Thus, the remote node that sends traffic using the faulty link can be identified in a simple manner.

[0032] The computer program may also be used to cause the device to respond to receiving the fault notification message through the processor, and the remote node blocks the port that receives the fault notification message. Therefore, the port where the faulty link is located can be easily identified and blocked.

[0033] The computer program may also be configured to cause the remote node to update an ECMP routing table associated with the remote node by deleting the port that received the fault notification message, thereby blocking the port. Thus, the remote node may reroute traffic to avoid the faulty link.

[0034] These and other aspects of the invention will be apparent from one or more of the embodiments described below. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] In order to make the present invention more easily understood, embodiments of the present invention will now be described by way of example with reference to the accompanying drawings, in which:

[0036] Figure 1 A two-layer fat tree is schematically depicted;

[0037] Figure 2 A fat-tree based network is schematically depicted;

[0038] Figure 3 is a flow chart of a method according to an example;

[0039] Figure 4 is a schematic diagram of a failure between the second layer and the first layer of a fat-tree network according to an example;

[0040] Figure 5A is a flow chart of a method according to an example shown from the perspective of a first node;

[0041] Figure 5B is a flow chart of a method according to an example shown from the perspective of a second node;

[0042] Figure 5C is a flow chart of a method according to an example, shown from the perspective of at least one remote node of a plurality of remote nodes. DETAILED DESCRIPTION

[0043] The exemplary embodiments are described below in sufficient detail to enable one of ordinary skill in the art to implement and practice the systems and processes described herein.It is important to understand that the embodiments may be provided in many alternative forms and should not be construed as limited to only the examples described herein.

[0044] Therefore, although the embodiments may be modified in various ways and take various alternative forms, specific embodiments thereof are shown in the drawings and described in detail below as examples. It is not intended to be limited to the specific forms disclosed. On the contrary, all modifications, equivalents and alternatives within the scope of the appended claims should be included. In all drawings and appropriate specific embodiments, the elements of the exemplary embodiments are consistently represented by the same reference numerals.

[0045] The terms used herein to describe the embodiments are not intended to limit the scope. The articles "a" and "the" are singular forms because they have only one referent, but the use of the singular form herein should not exclude the presence of multiple referents. In other words, unless the context otherwise clearly indicates, the elements mentioned in the singular form can be one or more in number. It should be further understood that the term "comprising" used herein is used to illustrate the presence of the features, items, steps, operations, elements and / or parts, but does not exclude the presence or addition of one or more other features, items, steps, operations, elements, parts and / or combinations thereof.

[0046] Unless otherwise defined, all terms (including technical and scientific terms) used in this article should be interpreted as customary usage in the art. It should also be understood that unless explicitly defined in this article, terms in common usage should also be interpreted as customary usage in the relevant art, and should not be interpreted as idealized or overly formal meanings.

[0047] Modern data centers usually use fat tree networks. To help understand the present invention, Figure 1A two-layer fat tree is schematically depicted. The two-layer fat tree 100 includes a complete bipartite graph between two layers of switches, namely spines 101 and leaves 102. Each spine 101 is connected to all leaves 102, and each leaf 102 is connected to all spines 101. Servers 103 are connected to the leaf layer. The level of a node is defined as the height in the tree, for example, leaves are at the first level and spines are at the second level. This structure can be extended by creating multiple sets of complete bipartite graphs between the layers of the fat tree.

[0048] Routing in a fat-tree network is a simple process that is always performed via the shortest path. Routing is sometimes referred to as "up-down" routing because the path always goes up to any spine (e.g. Figure 1 101) and then down to the target. Figure 2 A fat-tree based network is schematically depicted to help describe the up-down routing from source 201 to destination 202. To reach destination 202, a packet sent from source 201 may follow an up-path from source 201 to node 203-1, from node 203-1 to node 203-2, and from node 203-2 to node 203-3. The packet may then follow a down-path from node 203-3 to node 203-4, from node 203-4 to node 203-5, and finally from node 203-5 to the destination. Importantly, due to the inherent characteristics of the fat-tree topology, there are multiple up-paths to reach destination 202. That is, a packet sent from source 201 may reach destination 202 through any of the core nodes (i.e., the nodes at the highest level of the fat-tree hierarchy above the spine), not just through node 203-3. In contrast, there is only one downstream path from source 201 to destination 202 because there is only one path connecting node 203-3 to destination 202. In other words, in the upstream-downstream routing scheme, on the upstream way, there are multiple paths to the destination and all upstream ports belong to the path to the destination, while on the downstream way, there is only one path to the destination and only one downstream port belongs to the path.

[0049] Equal-cost multi-path (ECMP) routing is a routing strategy in which packets are forwarded to a single destination over multiple best paths with equal routing preference. Multipath routing can be used in conjunction with most routing protocols because multipath routing is a per-hop local decision made independently on each router. Multipath routing can significantly increase bandwidth by load balancing traffic over multiple paths. The mechanism is typically implemented by having the routing table contain multiple egress (output) ports for a given destination. The ports are typically selected by a hash function on the packet header data (e.g., a 5-tuple).

[0050] The Internet Control Message Protocol (ICMP) is a support protocol in the Internet Protocol Suite. The Internet Control Message Protocol is used by network devices (including routers) to send error messages and operational information indicating success or failure when communicating with another IP address. ICMP is a management protocol in IP networks, usually generated by routers. The ICMP header begins after the IPv4 header and is identified by the IP protocol number "1". All ICMP packets have an 8-byte header and variable-sized data sections. The first 4 bytes of the header have a fixed format, while the second 4 bytes depend on the type / code of the ICMP packet.

[0051] According to one example, a fast recovery mechanism is provided for failure recovery in a network environment. More specifically, aspects relate to bypassing link failures in a fat tree-based network environment, such as a fat tree-based IP network in a data center. The provided mechanism is advantageously based on the inheritance property of a fat tree topology. The mechanism assumes that multiple paths in the fat tree are used through equal-cost multipath routing. Advantageously, the scheme uses existing management protocols, such as ICMP, so no new protocols need to be invented.

[0052] Examples of the present invention may be provided as methods, systems or machine-readable instructions, such as any combination of software, hardware, firmware, etc. Such machine-readable instructions may be included in a computer-readable storage medium (including but not limited to disk storage, CD-ROM, optical storage, etc.) having a computer-readable program code thereon or on the medium.

[0053] The present invention is described with reference to the flow chart and / or block diagram of the method, device and system according to the example of the present invention. Although the flow chart described above shows a specific execution order, the execution order may be different from that described. The frame associated with a flow chart can be combined with the frame of another flow chart. In some examples, some frames of the flow chart may be unnecessary and / or additional frames can be added. It should be understood that each process and / or frame in the flow chart and / or block diagram, and the combination of the process and / or block diagram in the flow chart and / or block diagram can be realized by machine-readable instructions.

[0054] For example, machine-readable instructions can be executed by machines such as general-purpose computers, user devices such as smart devices (such as processors of smart phones, special-purpose computers, embedded processors or other programmable data processing devices) to implement the functions described in the embodiments and figures. Specifically, a processor or a processing device can execute machine-readable instructions. Therefore, the modules of the device (for example, a module that implements a comparator unit, or a firewall structure, etc.) can be implemented by a processor that executes machine-readable instructions stored in a memory, or a processor that operates according to instructions embedded in a logic circuit. The term "processor" is broadly interpreted to include a CPU, a processing unit, an ASIC, a logic unit or a programmable gate group, etc. These methods and modules can be executed by a single processor or by multiple processors.

[0055] Such machine-readable instructions may also be stored in a computer-readable storage device, which may direct a computer or other programmable data processing device to operate in a specific mode. For example, the instructions may be provided on a non-transitory computer-readable storage medium encoded with instructions executable by a processor.

[0056] Figure 3 is a flow chart of a method according to an example; in block 301, a fault recovery method in an Internet Protocol (IP) network includes the step of detecting a faulty link between a first node and a second node. For example, the fault may be detected by a controller for routing data packets through the nodes, or by one of the nodes itself. The IP network may be a fat tree network.

[0057] To help understand the present invention, Figure 4 A diagram of an IP network 400 according to an example is shown. The network 400 may include a fat tree network. For illustrative purposes, the link between the first node 403-1 and the second node 403-2 has been arbitrarily selected as the failed link. However, it should be understood by those skilled in the art that the fault recovery method described herein can be applied regardless of the location of the fault in the network. The first node 403-1 is located on (in other words, associated with) a first logical layer 410 of the network 400. The second node 403-2 is located on a second logical layer 420 of the network 400. The second layer 420 includes a layer that is logically higher than the first layer 410 in the topology. That is, the second layer 420 is located at a position higher than the first layer 410 in the tree hierarchy. In other words, see Figure 1 , the first node 403 - 1 can be described as a leaf node, and the second node 403 - 2 can be described as a spine node.

[0058] In block 302, the method includes: in response to detecting a failed link, the first node 403-1 blocks a port associated with the failed link between the first node 403-1 and the second node 403-2. In other words, since the first node 403-1 uses the failed link routing path on the upstream path, and due to the inherent characteristics of the fat tree topology, there is more than one upstream path to the target, therefore, the first node 403-1 (i.e., the downstream node) can be configured to repair the route locally.

[0059] Blocking the port associated with the failed link may include: the first node 403-1 updates an equal-cost multi-path (ECMP) routing table associated with the first node 403-1 by deleting the port associated with the failed link from the table, thereby blocking the port. Figure 4 In the example of FIG. 4 , the first node 403 - 1 is directly connected to two nodes 403 - 2 and 403 - 6. Since the port associated with the failed link has been blocked, any data packet sent upstream from the first node 403 - 1 will be sent through the node 403 - 6 instead of the node 403 - 2.

[0060] However, due to the downlink path characteristics of the fat tree topology, the second node 403-2 cannot repair the fault locally because there is only one path connecting the second node 403-2 and the first node 403-1 in the downlink direction, and the link fails. Therefore, an alternative repair must be applied. Therefore, the method may also include: in response to detecting the fault, the second node 403-2 sends a fault notification message to at least one of the multiple remote nodes 403-5 and 403-7 to 403-11. The multiple remote nodes 403-5 and 403-7 to 403-12 can be located on (i.e., associated with) the first layer 410 of the network. That is, the multiple remote nodes 403-5 and 403-7 to 403-12 can be located on the same layer as the first node 403-1, i.e., on the first layer 410. In other words, the plurality of remote nodes 403 - 5 and 403 - 7 to 403 - 12 may be located on (ie, associated with) a layer below the layer where the second node 403 - 2 is located, for example, the first layer 410 located below the second layer 420 .

[0061] Sending a fault notification message may include: the second node 403-2 identifies at least one remote node among the multiple remote nodes based on whether the remote node is used to send traffic using the faulty link. Since paths from multiple sources may use the faulty link, several remote nodes among the multiple remote nodes may be required to repair their routes. For a given allocation in a fat tree network, the remote node address is known in advance and can therefore be pre-configured at the second node 403-2. For example, remote nodes 403-5 and 403-7 to 403-11 can be pre-identified by a hierarchical allocation strategy. Sending a fault notification message may also include: generating a fault notification message including an Internet Control Message Protocol (ICMP) data packet, and sending the fault notification message to the identified remote node.

[0062] In order to identify at least one remote node configured to send traffic through the failed link, the method may include using an ECMP routing table associated with the second node 403-2 to identify a set of IP addresses configured to send traffic through the failed link. In other words, the second node 403-2 may search its ECMP routing table for an entry including a target IP, the target IP including a port blocked by the first node 403-1, the blocked port being the only output port in the entry.

[0063] The list of IPs using the failed link is the set of target IPs in these entries. An ICMP message can be constructed for each such IP, including the list of IPs using the failed link, for sending to a group of remote node IPs. The number of messages sent is equal to the number of remote nodes.

[0064] In response to receiving the fault notification message, at least one of the multiple remote nodes 403-5 and 403-7 to 403-11 (i.e., at least one remote node for sending traffic through the faulty link) can block the port that has received the fault notification message. For example, when the remote node receives the ICMP message, the remote node can search for the faulty IP table entry in its ECMP table and block the egress ICMP port. To this end, the remote node can update its ECMP routing table by deleting the port that received the fault notification message, thereby blocking the port. To further help understand the present invention, it is provided FIG. 5A to FIG. 5C . Figure 5A is a flow chart of a method according to an example, shown from the perspective of a first node. In block 501, a failed link is detected. In block 502, a port associated with the failed link between the first node and the second node is blocked by the first node.

[0065] Figure 5Bis a flow chart of a method according to an example shown from the perspective of a second node. In block 503, a failed link is detected. In block 504, the second node sends a notification message to at least one of the plurality of remote nodes, the notification message including a set of IP addresses configured to send traffic over the failed link.

[0066] Figure 5C is a flow chart of a method according to an example shown from the perspective of at least one remote node in a plurality of remote nodes. In block 505, at least one remote node in a plurality of remote nodes receives a notification message sent by a second node in block 504. In block 506, the remote node blocks an ingress port of notification messages from a group of associated target IPs.

[0067] According to one example, machine-readable instructions may be loaded onto a computer or other programmable data processing device so that the computer or other programmable data processing device performs a series of operations to produce a computer-implemented process, whereby the instructions executed on the computer or other programmable device provide operations for implementing the functions specified by the processes in the flowchart and / or the boxes in the block diagram.

[0068] Furthermore, the teachings herein may be implemented in the form of a computer or software product, such as a non-transitory machine-readable storage medium, which is stored in a storage medium and includes a plurality of instructions, such as machine-readable instructions, for causing a computer device to implement the methods described in examples of the invention.

[0069] In some examples, some methods may be performed in a cloud computing or network-based environment. A cloud computing environment may provide various services and applications over the Internet. These cloud-based services (e.g., software as a service, platform as a service, infrastructure as a service, etc.) may be accessed through a web browser or other remote interface of a user device. The various functions described herein may be provided through a remote desktop environment or any other cloud-based computing environment.

[0070] Although various embodiments have been described and / or shown in the context of a fully functional computing system herein, one or more of these exemplary embodiments can be distributed as a program product in various forms, regardless of the specific type of the computer-readable storage medium used to actually perform the distribution. The embodiments disclosed herein can also be implemented using software modules that perform certain tasks. These software modules can include scripts, batch processing or other executable files, which can be stored on a computer-readable storage medium or in a computing system. In some embodiments, these software modules can configure a computing system to perform one or more exemplary embodiments disclosed herein. In addition, one or more modules described herein can convert the representation of data, physical devices and / or physical devices from one form to another.

[0071] The above description is provided to enable those skilled in the art to best utilize the various aspects of the exemplary embodiments disclosed herein. This exemplary description is not intended to be exhaustive or limited to any precise form disclosed. Many modifications and variations may be made without departing from the spirit and scope of the present invention. The embodiments disclosed herein should be considered illustrative and not restrictive in all aspects. When determining the scope of the present invention, reference should be made to the attached claims and their equivalents.

Claims

1. A fault recovery method in an IP network, characterized in that: The network includes a plurality of nodes, the network includes a logical topology representing a hierarchical relationship between respective nodes of the plurality of nodes, the logical topology includes a plurality of logical layers of the network, the method including: detecting a failed link between a first node of the plurality of nodes and a second node of the plurality of nodes, wherein the first node is associated with a first logical layer of the network, the second node is associated with a second logical layer of the network, the second layer comprising a layer of the topology that is logically higher than the first layer; In response to detecting the failed link, the first of the plurality of nodes blocks a port associated with the failed link between the first of the plurality of nodes and the second of the plurality of nodes.

2. The method according to claim 1, characterized in that Blocking the port associated with the faulty link includes: The first of the plurality of nodes blocks the port by removing the port associated with the failed link from an equal-cost multi-path (ECMP) routing table associated with the first of the plurality of nodes, thereby updating the table.

3. The method according to claim 1 or 2, characterized in that: The following steps are also included: In response to detecting the failed link, the second node of the plurality of nodes sends a failure notification message to at least one remote node of a plurality of remote nodes associated with the first logical layer of the network.

4. The method according to claim 3, characterized in that Sending the fault notification message includes: The second node of the plurality of nodes identifies the at least one remote node of the plurality of remote nodes, the at least one remote node of the plurality of remote nodes being configured to send traffic through the failed link; generating the fault notification message including an Internet Control Message Protocol (ICMP) data packet; The fault notification message is sent to the identified remote node.

5. The method according to claim 4, characterized in that Identifying the at least one remote node configured to send traffic through the faulty link includes: Using an equal-cost multi-path (ECMP) routing table associated with the second node of the plurality of nodes, a set of Internet Protocol (IP) addresses configured to send traffic over the failed link is identified.

6. The method according to any one of claims 3 to 5, characterized in that The following steps are also included: In response to receiving the fault notification message, the remote node blocks the port that receives the fault notification message.

7. The method according to claim 6, characterized in that Blocking the port that receives the fault notification message includes: The remote node updates the ECMP routing table associated with the remote node by deleting the port that receives the fault notification message, thereby blocking the port.

8. The method according to any of the preceding claims, characterized in that The IP network includes a fat-tree based IP network.

9. An IP network, characterized in that: The IP network includes a plurality of nodes, the network includes a logical topology representing a hierarchical relationship between corresponding nodes in the plurality of nodes, the logical topology includes a plurality of logical layers of the network, and the network includes: a first node of the plurality of nodes associated with a first logical layer of the network; a second node of the plurality of nodes associated with a second logical layer of the network, the second layer comprising a layer of the topology that is logically higher than the first layer; Wherein, in response to detecting a faulty link between the first node among the multiple nodes and the second node among the multiple nodes, the first node among the multiple nodes is used to block the port associated with the faulty link between the first node among the multiple nodes and the second node among the multiple nodes.

10. The network according to claim 9, characterized in that The first node of the plurality of nodes is further configured to update an equal-cost multi-path (ECMP) routing table by deleting the port associated with the failed link from the table, thereby blocking the port.

11. The network according to claim 9 or 10, characterized in that The network also includes: a plurality of remote nodes, the plurality of remote nodes being associated with the first layer of the network; The second node among the multiple nodes is further configured to send a fault notification message to at least one of the multiple remote nodes.

12. The network according to claim 11, characterized in that The second node among the plurality of nodes is further configured to: identifying the at least one remote node among the plurality of remote nodes, the at least one remote node among the plurality of remote nodes being configured to send traffic through the faulty link; generating the fault notification message including an Internet Control Message Protocol (ICMP) data packet; The fault notification message is sent to the identified remote node.

13. The network according to claim 12, characterized in that The second node among the multiple nodes can also be used to use its ECMP routing table to identify a set of Internet Protocol (IP) addresses configured to send traffic through the faulty link, thereby identifying the at least one remote node to which the fault notification message is to be sent.

14. A network according to any one of claims 11 to 13, characterized in that The remote node may also be configured to, in response to receiving the fault notification message, block a port that receives the fault notification message.

15. The network according to claim 14, characterized in that The remote node may also be configured to update its ECMP routing table by deleting the port that receives the fault notification message, thereby blocking the port.

16. A network according to any preceding claim, characterised in that The IP network includes a fat-tree based IP network.

17. A data center, characterized in that: Comprising a fat-tree based network according to any one of claims 9 to 16.

18. A computer-readable storage medium, characterized in that: The computer readable storage medium includes computer program code, which can be accessed by an apparatus including a processor to provide instructions and / or data to the apparatus, and the computer program code is used to cause the apparatus, through the processor, to: In an IP network including a plurality of nodes, detecting a failed link between a first node of the plurality of nodes and a second node of the plurality of nodes, the network including a logical topology representing a hierarchical relationship between respective nodes of the plurality of nodes, the logical topology including a plurality of logical layers of the network, wherein the first node is associated with a first logical layer of the network, the second node is associated with a second logical layer of the network, the second layer comprising a layer of the topology that is logically higher than the first layer; In response to detecting the failed link, the first of the plurality of nodes blocks a port associated with the failed link between the first of the plurality of nodes and the second of the plurality of nodes.

19. The computer-readable storage medium according to claim 18, wherein: The computer program code is further configured to cause the apparatus, through the processor, to: The first of the plurality of nodes blocks the port by removing the port associated with the failed link from an equal-cost multi-path (ECMP) routing table associated with the first of the plurality of nodes, thereby updating the table.

20. The computer-readable storage medium according to claim 18 or 19, characterized in that: The computer program code is further configured to cause the apparatus, through the processor, to: In response to detecting the failed link, the second node of the plurality of nodes sends a failure notification message to at least one remote node of a plurality of remote nodes associated with the first layer of the network.

21. The computer-readable storage medium according to claim 20, wherein: The computer program code is further configured to cause the apparatus, through the processor, to: The second node of the plurality of nodes identifies the at least one remote node of the plurality of remote nodes, the at least one remote node of the plurality of remote nodes being configured to send traffic through the failed link; generating the fault notification message including an Internet Control Message Protocol (ICMP) data packet; The fault notification message is sent to the identified remote node.

22. The computer-readable storage medium according to claim 21, wherein: The computer program code is further configured to cause the apparatus, through the processor, to: Using an equal-cost multi-path (ECMP) routing table associated with the second node of the plurality of nodes, a set of Internet Protocol (IP) addresses configured to send traffic over the failed link is identified, thereby identifying the at least one remote node.

23. The computer-readable storage medium according to any one of claims 20 to 22, characterized in that: The computer program code is further configured to cause the apparatus, through the processor, to: In response to receiving the fault notification message, the remote node blocks the port that receives the fault notification message.

24. The computer-readable storage medium according to claim 23, wherein: The computer program code is further configured to cause the apparatus, through the processor, to: The remote node updates the ECMP routing table associated with the remote node by deleting the port that receives the fault notification message, thereby blocking the port.