CN116545935A

CN116545935A - Method, device, equipment and storage medium for scheduling anti-affinity service flow

Info

Publication number: CN116545935A
Application number: CN202310608655.6A
Authority: CN
Inventors: 宋小恒
Original assignee: New H3C Technologies Co Ltd
Current assignee: New H3C Technologies Co Ltd
Priority date: 2023-05-27
Filing date: 2023-05-27
Publication date: 2023-08-04

Abstract

The invention provides a method, a device, equipment and a storage medium for scheduling anti-affinity service flows, which are used for solving the problem of scheduling the anti-affinity service flows in an SDN network. The invention introduces anti-affinity attribute to traffic engineering tunnel, SDN controller monitors traffic flow in traffic engineering tunnel in real time, when network congestion exists in a certain tunnel path, SDN controller timely identifies traffic flow on the congested tunnel path, carries out rerouting calculation to identified anti-affinity traffic flow, and dispatches partial anti-affinity traffic flow to other tunnel paths based on preset mechanism. The invention can reduce the occurrence of network retransmission caused by congestion and packet loss and improve the utilization rate of the whole bandwidth of the network.

Description

Method, device, equipment and storage medium for scheduling anti-affinity service flow

Technical Field

The present invention relates to the field of communications and technologies, and in particular, to a method, an apparatus, a device, and a storage medium for scheduling an anti-affinity service flow.

Background

SDN (Software Defined Network ) is a novel network innovation architecture, and the core idea is to separate the control layer and the forwarding layer of network equipment so as to realize flexible control of network traffic and provide a good platform for innovation of core network and application.

Network congestion is a major problem affecting backbone network performance. The congestion may be caused by insufficient network resources or by local congestion caused by unbalanced network resource loads. TE (Traffic Engineering ) is used to solve congestion due to load imbalance.

The traffic engineering TE dynamically adjusts traffic management parameters, routing parameters, resource constraint parameters and the like by monitoring the traffic of the network and the load of the network element in real time, so that the network operation state is migrated to an ideal state, the use of network resources is optimized, and the congestion caused by unbalanced load is avoided.

In general, the performance metrics of traffic engineering include two aspects:

(1) Performance index for service: enhancing QoS (Quality of Service ) performance of traffic, e.g. impact on packet loss, latency, throughput, and SLA (Service Level Agreement ).

(2) Resource-oriented performance index: optimizing resource utilization. Bandwidth is an important resource, and efficient management of bandwidth resources is a central task of traffic engineering.

Multiprotocol label switching traffic engineering (Multi-Protocol Label Switching-Traffic Engineering, MPLS-TE) combines MPLS technology with traffic engineering technology, and resource reservation is performed by establishing a label switched path (Lable Switch Path, LSP) tunnel to a specified path, so that network traffic bypasses congestion nodes, and the purpose of balancing network traffic is achieved. Multiple LSP tunnels may be required when rerouting (Reroute) is deployed or traffic needs to be transported through multiple paths. In TE, such a set of LSP tunnels is referred to as TE tunnels (Traffic Engineered Tunnel).

MPLS-TE tunnels are a virtual point-to-point connection from a head node to a destination node. Typically, MPLS-TE tunnels are formed by a Constraint-based label switched path (CRLSP). In the case where traffic needs to be transmitted through multiple paths, etc., multiple CRLSPs need to be established for the same traffic, in which case the MPLS-TE tunnel is made up of a set of CRLSPs.

The affinity properties of MPLS-TE tunnels determine the links used by the tunnel, which, in cooperation with the link management group, determine which links the tunnel can use. The affinity attribute (Affinity attribute) is a 32-bit vector that characterizes the links required for MPLS-TE tunneling and a 32-bit mask (similar to IPv4 address and subnet mask). The affinity attribute is configured under the Tunnel (Tunnel) interface of the Tunnel Ingress node, and can be carried along with the resource reservation protocol (Resource reSerVation Protocol-Traffic Engineering, RSVP-TE) signaling based on traffic engineering extension, so that the label switched path (Lable Switch Path, LSP) established for the Tunnel at the current device can be affected, so that the system can calculate and select a new link path for the Tunnel again.

The link management group attribute is a 32-bit vector representing the link attribute, and is configured under the physical interface of each link. Each bit in the link attribute represents an attribute, which may or may not be set, and may be associated in any desired sense. Such as to represent link bandwidth, performance, or management policies, etc. The link attribute only takes effect on the LSP newly created through the interface, and does not affect the established LSP.

After the affinity attribute is configured for the MPLS-TE tunnel, the tunnel will compare the affinity attribute with the management group attribute of the link when calculating the path, and decide to select or avoid the link with some attribute. The MPLS-TE tunnel uses a 32-bit mask to represent the bits to be compared, and the path is selected when the tunnel is routed by respectively phase-separating the affinity attribute and the mask of the link management group and the tunnel, and the path is abandoned if the obtained result is the same.

For example, during tunnel path selection, RSVP-TE signaling issued by a tunnel Ingress node is diffused along with flooding of OSPF-TE or ISIS-TE routing information, and an appropriate path for continuing transmission to a destination direction is selected according to an affinity attribute of the tunnel configuration by each node. The method of selecting is to compare the tunnel affinity attribute carried in the signaling with the link management group attribute of each link enabling MPLS-TE capability on the current forwarding device, and select the path with matched attribute as the path selected by the tunnel at the node.

When the traffic is routed by using the affinity attribute of the MPLS-TE tunnel, some drawbacks may occur in some cases, for example, traffic with many applications of the same type and the same service quality requirement may be introduced into the same TE tunnel, resulting in traffic aggregation and excessive tunnel load.

One common avoidance measure for solving the above problem is to configure a TE tunnel as a multi-LSP path to perform load sharing, but the hash algorithm for performing load sharing is a statically executed algorithm, and when the forwarding device performs hash operation on a service flow packet, it is easy to hash a plurality of service flows with the same affinity attribute onto the same next hop, so that these service flows cannot be balanced fundamentally. If the hash algorithm of the forwarding device is modified, the result of the message Wen Haxi is not necessarily uniform as expected, and the TCP/UDP field of the header of the traffic flow message will change frequently, so that the static hash algorithm is difficult to achieve real-time uniform distribution of the traffic flow.

In addition, the SDN controller directly issues policy routes for each service flow, so that the service flows are uniformly distributed, excessive hardware policy route resources are occupied, the requirement on forwarding equipment is too high, too many policy routes need to be configured, the maintenance cost is too high, and the realizability is too low.

Disclosure of Invention

In view of the above, the present invention provides a method, apparatus, device and storage medium for scheduling anti-affinity traffic, which are used for solving the technical problem of scheduling anti-affinity traffic in an SDN network.

Based on an aspect of the embodiment of the invention, the invention provides an anti-affinity service flow scheduling method, which is applied to a software defined network SDN controller, and comprises the following steps:

the SDN controller monitors traffic on a plurality of tunnel paths in a traffic engineering tunnel in real time, wherein the tunnel paths are configured with affinity attributes;

when the SDN controller monitors that a certain tunnel path on a certain forwarding device generates congestion, carrying out anti-affinity service flow identification on a service flow transmitted on the congested tunnel path, and determining the anti-affinity service flow;

the SDN controller performs rerouting pre-calculation on part of service flows in the anti-affinity service flows based on a preset priority order, and selects a service flow scheduling scheme which does not cause congestion of other tunnel paths according to a result of the rerouting pre-calculation;

and based on the selected service flow scheduling scheme, issuing a new route of the selected scheduled service flow to the forwarding equipment so that the forwarding equipment updates a routing table, and scheduling the selected scheduled service flow to other tunnel paths for transmission.

Further, the method for the SDN controller to monitor traffic on multiple tunnel paths in the traffic engineering tunnel in real time includes:

the SDN controller monitors the service flow state of each tunnel path on each forwarding device in real time through a flow monitoring mechanism based on the managed flow engineering tunnel information and routing information, and when the tunnel path generates congestion, the SDN controller triggers the processing step of anti-affinity service flow identification according to the congestion alarm reported by the forwarding device.

Further, the method for identifying the anti-affinity service flow comprises the following steps:

the SDN controller obtains the traffic transmission characteristics of the traffic flows on the congestion tunnel path, carries out correlation analysis on the traffic flows based on the traffic transmission characteristics, calculates the correlation between the traffic flows, and determines that the two traffic flows are anti-affinity traffic flows when the correlation between the two traffic flows is larger than a preset threshold.

Further, the method for acquiring the traffic transmission characteristics of the traffic flow on the congestion tunnel path is as follows:

the SDN controller obtains a historical fluctuation curve reflecting the service flow based on the flow sampling data of the service flow, and obtains one or more of the average flow speed, the fluctuation amplitude of the flow bandwidth, the fluctuation frequency and the fluctuation duration characteristic of the service flow as the flow transmission characteristic through the historical fluctuation curve.

Further, the method for performing rerouting pre-calculation to obtain the selected service flow scheduling scheme includes: selecting a target service flow from the anti-affinity service flow according to a preset selection strategy, determining the rest service flow as an alternative service flow, removing the next hop of the tunnel path corresponding to the alternative service flow, and selecting an optimal path in the rest tunnel path as a target path based on the preset path selection strategy;

and taking the next hop of the target path as the next hop of the target service flow, judging whether a service flow scheduling scheme for scheduling the target service flow to the target path causes congestion of the target path, and taking the target service flow and the next hop route corresponding to the target path as the selected service flow scheduling scheme when judging that congestion cannot be caused.

Further, when congestion is determined to be caused, a different service flow is reselected in the anti-affinity service flow as a target service flow according to a preset selection strategy, and the selection step of the target path is executed again until a target path which does not cause congestion is selected, and the finally selected target service flow and the next hop route corresponding to the target path are used as a selected service flow scheduling scheme;

If any combination of the target service flow and the target path finally cannot meet the condition of not causing congestion of the target path, the selection of the selected service flow scheduling scheme is terminated, and the scheduling flow is ended.

Further, the method for selecting an optimal path as the target path in the remaining tunnel paths based on a preset path selection policy includes:

selecting a tunnel path with the best matching affinity attribute from the rest tunnel paths as a target path; or (b)

And comprehensively selecting a tunnel path which meets the affinity attribute of the tunnel path and can balance the link load from the rest tunnel paths based on the affinity attribute and the link load as a target path.

Further, the plurality of tunnel paths within the traffic engineering tunnel include equivalent paths;

when the tunnel path corresponding to the alternative service flow is an equivalent path, the excluding the next hop of the tunnel path corresponding to the alternative service flow specifically includes: and excluding the equivalent next hop of the tunnel path corresponding to the alternative service flow.

Based on another aspect of the embodiment of the invention, the invention also provides an anti-affinity service flow scheduling device, which can be realized in a mode of software, hardware or combination of software and hardware. When implemented as a software module, the program code of the software module is loaded into a storage medium of the device, and the program code in the storage medium is read and executed by a processor, thereby implementing the functions of the respective constituent modules in the apparatus. The device is applied to a software defined network SDN controller, and comprises:

The monitoring module is used for monitoring the traffic on a plurality of tunnel paths in the traffic engineering tunnel in real time, and the plurality of tunnel paths are configured with affinity attributes;

the anti-affinity service flow determining module is used for carrying out anti-affinity service flow identification on the service flow transmitted on a certain tunnel path of certain forwarding equipment when the monitoring module monitors that the certain tunnel path generates congestion, and determining the anti-affinity service flow;

the scheduling scheme determining module is used for carrying out rerouting pre-calculation on part of the traffic flows in the anti-affinity traffic flows based on a preset priority order, and selecting a traffic flow scheduling scheme which does not cause congestion of other tunnel paths according to the result of the rerouting pre-calculation;

and the route issuing module is used for issuing the new route of the selected scheduled service flow to the forwarding equipment based on the selected service flow scheduling scheme so as to enable the forwarding equipment to update a route table and schedule the selected scheduled service flow to other tunnel paths for transmission.

Further, the monitoring module monitors the service flow state of each tunnel path on each forwarding device in real time through a traffic monitoring mechanism based on the managed traffic engineering tunnel information and routing information, and when the tunnel path generates congestion, the anti-affinity service flow determining module is triggered to perform anti-affinity service flow identification according to the congestion alarm reported by the forwarding device.

Further, the anti-affinity service flow determining module includes:

a feature acquisition unit, configured to obtain, based on traffic sample data of a traffic flow, a historical fluctuation curve reflecting the traffic flow, and obtain, through the historical fluctuation curve, a combination of one or more of characteristics of a traffic flow average flow velocity, a traffic bandwidth fluctuation amplitude, a fluctuation frequency, and a fluctuation duration of the traffic flow as the traffic transmission characteristic;

and the anti-affinity service flow determining unit is used for carrying out correlation analysis on the service flows based on the flow transmission characteristics, calculating the correlation between the service flows, and determining that the two service flows are anti-affinity service flows when the correlation between the two service flows is larger than a preset threshold.

Further, the scheduling scheme determining module includes:

the target service flow and target path determining unit is used for selecting a target service flow from the anti-affinity service flow according to a preset selection strategy, determining the rest service flow as an alternative service flow, and selecting an optimal path in the rest tunnel path as a target path based on the preset path selection strategy after the next hop of the tunnel path corresponding to the alternative service flow is eliminated;

and the service flow scheduling scheme determining unit is used for taking the next hop of the target path as the next hop of the target service flow, judging whether the service flow scheduling scheme for scheduling the target service flow to the target path causes congestion of the target path, and taking the target service flow and the next hop route corresponding to the target path as the selected service flow scheduling scheme when judging that congestion cannot be caused.

Further, the target service flow and target path determining unit selects a tunnel path with the best matching affinity attribute from the remaining tunnel paths as a target path; or comprehensively selecting a tunnel path which not only accords with the affinity attribute of the tunnel path but also can balance the link load from the rest tunnel paths based on the affinity attribute and the link load as a target path.

and the target service flow and target path determining unit excludes equivalent next hops of the tunnel paths corresponding to the alternative service flow when the tunnel paths corresponding to the alternative service flow are equivalent paths.

Drawings

In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following description will briefly describe the drawings required to be used in the embodiments of the present invention or the description in the prior art, and it is obvious that the drawings in the following description are only some embodiments described in the present invention, and other drawings may be obtained according to these drawings of the embodiments of the present invention for a person having ordinary skill in the art.

FIG. 1 illustrates a network application scenario diagram that causes traffic engineering tunnel congestion;

FIG. 2 is a schematic diagram of a network application scenario in which an anti-affinity traffic scheduling method is used to implement traffic scheduling of a host according to an embodiment of the present invention;

fig. 3 is a schematic flow chart of steps of a method for scheduling an anti-affinity service flow according to an embodiment of the present invention;

fig. 4 is a schematic structural diagram of an electronic device for implementing the method for scheduling an anti-affinity service flow according to an embodiment of the present invention.

Detailed Description

The terminology used in the embodiments of the invention is for the purpose of describing particular embodiments only and is not intended to be limiting of embodiments of the invention. As used in this embodiment of the invention, the singular forms "a," "an," and "the" are intended to include the plural forms as well, unless the context clearly indicates otherwise. It should be understood that although the terms first, second, third, etc. may be used in embodiments of the present invention to describe various information, these information should not be limited to these terms. These terms are only used to distinguish one from another or similar information, entity or step, but not to describe a particular sequence or order. For example, the first information may also be referred to as second information, and similarly, the second information may also be referred to as first information, without departing from the scope of embodiments of the present invention. Furthermore, the word "if" as used may be interpreted as "at … …" or "at … …" or "in response to a determination". The "and/or" in the present invention is merely an association relationship describing the association object, and indicates that three relationships may exist, for example, a and/or B may indicate: there are three cases, a alone, a and B together, and B alone, wherein a, B may be singular or plural. Also, in the description of the present invention, unless otherwise indicated, "a plurality" means two or more than two. "at least one of" or the like means any combination of these items, including any combination of single item(s) or plural items(s). For example, at least one (one) of a, b, or c may represent: a, b, c, a-b, a-c, b-c, or a-b-c, wherein a, b, c may be single or plural.

The invention provides a counteraffinity service flow scheduling scheme for solving the technical problem that traffic flow aggregation in a traffic engineering tunnel and overlarge tunnel path load are caused when traffic flows matched with TE tunnel affinity attributes are drained in an SDN network, wherein the core idea of the technical scheme is as follows: introducing an anti-affinity attribute for the traffic engineering tunnel, monitoring the traffic flow in the traffic engineering tunnel in real time by the SDN controller, identifying the traffic flow on the congested tunnel path by the SDN controller in time when network congestion exists in a certain tunnel path, carrying out rerouting calculation on the identified anti-affinity traffic flow, and scheduling part of the anti-affinity traffic flow on other tunnel paths based on a preset mechanism. The invention can reduce the occurrence of network retransmission caused by congestion and packet loss and improve the utilization rate of the whole bandwidth of the network.

The anti-affinity attribute is used for defining the mutual exclusion characteristic of the traffic flow in the traffic engineering tunnel, when the traffic transmission characteristics of a plurality of traffic flows are very similar, the traffic flows are overlapped to possibly generate large burst traffic.

The traffic engineering tunnel to which the technical scheme of the invention is applicable is not limited to MPLS-TE tunnel, but also is applicable to all tunnels with traffic engineering properties capable of realizing traffic scheduling, such as Segment Routing-Traffic Engineering (SR-TE) tunnel, and the invention is not particularly limited.

In order to analyze and explain the technical problem to be solved by the present invention more clearly, fig. 1 illustrates a network application scenario diagram that causes traffic engineering tunnel congestion. In an SDN network as shown in fig. 1, a traffic engineering tunnel is created between Leaf1 and Leaf2, and multiple tunnel paths exist between Leaf1 and Leaf2, and a controller defines multiple affinity attributes for each path of the traffic engineering tunnel. When the Leaf1 selects a route for the service flow, a tunnel path with the service flow characteristic close to the affinity attribute of the tunnel path is selected for transmission, for example, the tunnel paths between the Leaf1-Spine1-Leaf2 and the Leaf1-Spine2-Leaf2 have low-delay and high-burst affinity attributes, and the tunnel paths between the Leaf1-Spine3-Leaf2 have low-delay and high-reliability affinity attributes, but the delay affinity attribute weight is not larger than that of the two tunnel paths. There are multiple traffic interactions between Leaf1 and Leaf2, where policy drainage is performed on traffic flows of low latency characteristics, each using paths Leaf1-Spine1-Leaf2 and Leaf1-Spine2-Leaf2 that more closely match the tunnel affinity properties. The controller monitors the flow condition of each equipment node in real time, counts the forwarding condition of each node to the service flow, and counts the forwarding condition of each service flow on each node.

After a period of time, the flow monitoring module of the SDN controller monitors that traffic loss exists on the link path of the Spine2-Leaf2, congestion exists on the output ports of the Spine2 to the Leaf2, and if the affected traffic flows are host10- > host20 and host11- > host21. The controller monitors that the two traffic flows belong to the same network segment of the data center, and an exit port selected by the affinity attribute algorithm for the two traffic flows on the Leaf1 is an exit port of an equivalent path, but the two traffic flows are distributed on the same path, namely the Leaf1-Spine2-Leaf2, through hash operation when addressing, and congestion can occur on an exit link of the Spine2-Leaf2 when traffic burst peaks of the two traffic flows overlap. At this time, modifying the path affinity attribute cannot solve the problem, and the equivalent path hash algorithm is fixed, and once modified, the global is affected.

The SDN controller in the invention can regulate and control the service flow in fine granularity, the flow monitoring module can monitor the service flow condition of each tunnel path between Leaf devices, analyze the flow transmission characteristics of each service flow, extract the service flow with similar flow transmission characteristics to carry out anti-affinity attribute identification when finding out link congestion, set the anti-affinity attribute for the service flow with the highly similar flow transmission characteristics, and ensure that the service flows select different tunnel paths as much as possible when carrying out route scheduling.

The flow transmission characteristics described in the present invention refer to characteristics capable of reflecting flow trends and states, and may include, for example, a combination of one or more of characteristics such as average flow velocity of traffic flow, fluctuation amplitude (peak and trough values) of flow bandwidth, fluctuation frequency, and fluctuation duration.

Taking Host10 and Host11 on the Leaf1 device as examples, the Host10 selects a tunnel path as Leaf1-Spine1-Leaf2 and Leaf1-Spine2-Leaf2 according to a path priority and affinity attribute rule, the Host11 selects the same tunnel path as the Host10 according to the path priority and affinity attribute rule, if the traffic flows of the two hosts are overlapped to cause congestion of the tunnel path, and the SDN controller selects one of the two traffic flows to carry out rerouting according to the traffic transmission characteristic to exclude the next hop address of the other traffic flow, and then carries out path selection according to the affinity attribute set by the tunnel. As shown in the example of fig. 1, assuming that the traffic flow of the host11 is re-routed, the equivalent paths of the traffic flow of the host10 are removed first by using the next hops 1 and 2, and then the tunnel path is selected for the host11 traffic in the remaining paths of the traffic engineering tunnel by adopting the original same path method, so that the tunnel paths between the Leaf1-Spine3-Leaf2 are selected for forwarding.

Fig. 2 is a schematic diagram of a network application scenario in which an anti-affinity traffic scheduling method is used to implement traffic scheduling of a host according to an embodiment of the present invention. In the data center network shown in fig. 2, a leg 1 hangs down a leg 10, a leg 11 and a leg 12, a leg 2 hangs down a leg 20, a leg 21 and a leg 22, and a traffic engineering tunnel is established between the leg 1 and the leg 2, and includes three tunnel paths, respectively:

a first path: leaf1-Spine1-Leaf2

Second path: leaf1-Spine2-Leaf2

Third path: leaf1-Spine3-Leaf2

The first path and the second path are equivalent paths, and the path cost of the third path is greater than that of the first path and the second path. For the application scenario that all tunnel paths are non-equivalent paths, the invention is also applicable, and the purpose of taking the application scenario including equivalent paths as an embodiment is to highlight the technical problems and technical effects to be solved by the invention. The numbers marked on the forwarding devices (including Leaf and Spine) in the figures represent the physical port numbers of the devices.

In the embodiment illustrated in fig. 2, the controller can monitor the traffic flow on any tunnel path in real time, and when the congestion of the tunnel path is monitored, the SDN controller performs route calculation based on the anti-affinity attribute based on the monitoring result and issues a new route to the Leaf device, so that the Leaf dispatches some or some traffic flows with traffic transmission characteristics matching the anti-affinity attribute to other tunnel paths, thereby relieving the link congestion and guaranteeing the service quality requirement of the traffic flows. The method can avoid issuing a large number of strategy route or flow list items to the Leaf equipment, thereby reducing the influence on the Leaf equipment, and simultaneously, the flow monitoring capability of the controller can be fully utilized, and the controller can predict whether the scheduled target tunnel path can bear the scheduled service flow in advance, so that more reasonable and intelligent flow scheduling is realized.

Fig. 3 is a schematic flow chart of steps of a method for scheduling an anti-affinity service flow according to an embodiment of the present invention, where the method includes:

step 301, an sdn controller monitors traffic on a plurality of tunnel paths in a traffic engineering tunnel in real time, wherein the plurality of tunnel paths are configured with affinity attributes;

the SDN controller is used as intelligent management equipment of the network, maintains traffic engineering tunnel information and routing information on each forwarding equipment in the network, can monitor the traffic flow state on each physical link on each forwarding equipment in real time through a traffic monitoring mechanism, and can monitor the traffic flow state of each traffic flow when passing through the physical equipment. The SDN controller may also obtain, through statistical analysis, a traffic transmission characteristic of each traffic flow according to a history and a current traffic state of each traffic flow, where the obtained traffic transmission characteristic of the traffic flow is a necessary condition for performing anti-affinity attribute traffic scheduling, and may be generated by a dedicated monitoring analysis component or system by recording traffic flow logs, and the SDN controller may obtain the traffic transmission characteristic from the traffic flow logs.

In an initial situation, the first path and the second path have a low overhead, and the priority of the routing based on the affinity properties is higher than that of the third path. For example, the tunnel paths corresponding to the first path and the second path are configured with low-latency and high-burst affinity properties, the tunnel path corresponding to the third path is configured with low-latency and high-reliability affinity properties, but the latency affinity properties of the third path are not weighted as much as those of the two tunnel paths of the first path and the second path. After the Leaf1 performs routing calculation based on the affinity attribute, the traffic data stream is preferentially selected to be transmitted through the equivalent path, i.e. the first path and the second path, for the traffic stream of the down-hanging host.

After the Leaf1 receives the messages of Host10- > Host20 and Host11- > Host21 of the Host traffic flows, performing routing calculation based on the affinity attribute of the tunnel path, two next hops of the equivalent path respectively correspond to the Spine1 and the Spine2, and the Leaf1 performs hash calculation by using the header information (for example, source ip+ port or destination ip+ port) of the traffic flow messages, where the hash result may be that both traffic flows are hashed to the same output port (for example, port 1 of the illustrated Leaf 1), that is, the first path is selected for forwarding.

The controller can monitor the service flow of host10- > host20 and host11- > host21, wherein the service flow of host10- > host20 enters from the downstream port of Leaf1 and exits from port 1; entering from the port 1 of the Spine1 and exiting from the port 2; from port 3 of Leaf2, in, downstream out, the tunnel path is expressed in simplified terms:

(downstream port) Leaf1 (1) - (1) Spine1 (2) - (3) Leaf2 (downstream port);

by adopting the same expression, the forwarding path of the traffic flow of host11- > host21 is also:

(downstream port) Leaf1 (1) - (1) Spine1 (2) - (3) Leaf2 (downstream port);

when Leaf1 routes two traffic flows with highly similar transmission characteristics to the same tunnel path, if two burst flows are overlapped to exceed the bandwidth of the tunnel path, message congestion and even packet loss of the tunnel path are caused.

Step 302, when the SDN controller monitors that a certain tunnel path on a certain forwarding device generates congestion, carrying out anti-affinity service flow identification on a service flow transmitted on the congested tunnel path, and determining the anti-affinity service flow;

for example, at a certain moment, the SDN controller monitors that congestion or packet loss occurs in the egress direction of port 2 of the spline 1 on the spline 1 (2) - (3) Leaf2 path. The SDN controller can acquire the service flow sent to the host20 and the host11- > host21 by the host10 of the service flow of the link from the Spine1 port 2 to the Leaf2 port 3 according to the traffic engineering tunnel information, the routing information, the service flow state and the like on each forwarding device in the maintained network, after acquiring the service flow information on the congestion path, the SDN controller carries out the identification of the anti-affinity service flow on the service flow,

the authentication mode of the anti-affinity service flow can be as follows: the SDN controller obtains the traffic transmission characteristics of the traffic flows on the congestion tunnel path, carries out correlation analysis on the traffic flows based on the traffic transmission characteristics, calculates the correlation between the traffic flows, and determines that the two traffic flows are anti-affinity traffic flows when the correlation between the two traffic flows is larger than a preset threshold.

The traffic transmission characteristic of a traffic flow in the present invention may be understood as a curve obtained based on traffic sampling data of the traffic flow and capable of reflecting historical fluctuation conditions of the traffic flow, and one or more combinations of characteristics of average traffic flow velocity, fluctuation amplitude (peak and trough values) of traffic bandwidth, fluctuation frequency, fluctuation duration, etc. of the traffic flow may be obtained through the historical fluctuation curve as traffic transmission characteristics of the traffic flow. The SDN controller may analyze the correlation between two traffic history surge curves for a period of time before congestion occurs through a trained artificial intelligence model or correlation analysis function.

Referring to the example of fig. 2, assuming that three traffic flows forwarded through the first path of the traffic engineering tunnel are three, that is, three traffic flows that are sent from host10 to host20, three traffic flows that are sent from host11 to host21, and three traffic flows that are sent from host12 to host22, when congestion occurs in the tunnel path between the Spine1 and the Leaf2, the Spine1 reports congestion alarm information to the SDN controller, and the SDN controller performs anti-affinity traffic flow identification on traffic flows carried in the tunnel path based on the monitored and maintained tunnel path and traffic flow information carried by the Spine1, performs correlation analysis between every two of the three traffic flows, and determines traffic flows with highly relevant traffic transmission characteristics as anti-affinity traffic flows. If only two traffic flows are carried on the tunnel path, the two traffic flows can be directly judged as anti-affinity traffic flows when the congestion occurs on the tunnel path.

If the identification of the anti-affinity service flows does not determine that any two or more service flows meet the judgment conditions of the anti-affinity service flows, the scheduling processing step of the anti-affinity service flows is abandoned.

Step 303, the SDN controller performs rerouting pre-calculation on part of the traffic flows in the anti-affinity traffic flows based on a preset priority order, and selects a traffic flow scheduling scheme which does not cause congestion of other tunnel paths according to the result of the rerouting pre-calculation;

One method of selecting a traffic flow scheduling scheme is:

selecting a target service flow from the anti-affinity service flow according to a preset selection strategy, determining the rest service flow as an alternative service flow, removing the next hop of a tunnel path corresponding to the alternative service flow in the traffic engineering tunnel, selecting an optimal path in the rest tunnel path as a target path based on the preset path selection strategy, taking the next hop of the target path as the next hop of the target service flow, judging whether the scheduling of the target service flow can cause congestion of the target path according to the rerouting precalculation result, if congestion can be caused, reselecting a different service flow in the anti-affinity service flow as the target service flow according to the preset selection strategy, and executing the selection step of the target path again until a target path which can not cause congestion is selected, and taking the finally selected target service flow and the next hop route corresponding to the target path as the selected service flow scheduling scheme.

selecting a tunnel path with the best matching affinity attribute from the rest tunnel paths as a target path; or alternatively, the first and second heat exchangers may be,

Referring to the example of fig. 2, in an embodiment of the present invention, it is assumed that there are three traffic flows forwarded through the first path of the traffic engineering tunnel, which are respectively the first traffic flows: host10- > host20, second traffic: host11- > host21, third traffic: the host12- > host22, link costs of the first path, the second path and the third path are different (no equivalent path is provided), the matching degree of the affinity attribute of the link 1 and the three traffic flows is optimal, the matching degree of the affinity attribute of the second path and the third path is suboptimal and the same, in the initial case, the three traffic flows all travel the first path, congestion occurs at an outlet port from a Spin1 node to a Leaf2 of the first path at a certain moment, the SDN controller determines that the three traffic flows are all the anti-affinity traffic flows after performing anti-affinity traffic flow authentication for the tunnel inlet node Leaf1, the SDN controller selects one traffic flow from the three anti-affinity traffic flows based on a preset priority order (for example, randomly selects one traffic flow of the three anti-affinity traffic flows or selects a traffic flow with the minimum flow rate according to the traffic flow equalizing speed), other traffic flows as target traffic flows, the selected first traffic flow is assumed to be the target traffic flow, the second traffic flow and the third traffic flow is assumed to be the first path, the SDN controller selects the first path as the target flow path and the second path, the second path is selected as the target flow path, the first path is selected from the target path is the target flow path, the first path is selected from the first path and the second path is the target path, the first path is selected as the target path, the target path is selected from the first path is the target path, and the first path is the target-preferred, and the first path is the target path is selected from the first path, and the first path is the target path is selected as the target path and the second path is the target path and the optimal, judging whether the scheduling of the target first service flow is caused to congestion of the target second path according to the rerouting precalculation result, and if the scheduling of the target first service flow is not caused to congestion, issuing the finally selected route of the next hop (Spine 2) corresponding to the target first service flow and the target second path as a selected service flow scheduling scheme to the Leaf1. If congestion is caused, the SDN controller reselects a different service flow in the anti-affinity service flow as a target service flow according to a preset selection strategy to execute the selection step of the target path again until a target path which does not cause congestion is selected, and the finally selected target service flow and the next hop route corresponding to the target path are used as the selected service flow scheduling scheme. If any one of the three service flows is selected as the target service flow, the condition that the target path is not congested cannot be met, the selection process of the selected service flow scheduling scheme is terminated, and the scheduling process is terminated. After the current scheduling process is terminated, the scheduling period can be preset or the scheduling can be restarted after the congestion alarm is received again.

As illustrated in fig. 2, in an application scenario in which the first path and the second path are equivalent paths, the third path is a non-equivalent path, and the affinity attribute matching degree of the first path and the second path is higher than that of the third path, it is assumed that the first traffic flow: host10- > host20, second traffic: in the initial condition, host11- > host2 walks a first path, after congestion occurs from a Spin1 node of the first path to an exit port of Leaf2, after performing anti-affinity service flow identification for a tunnel entrance node Leaf1, an SDN controller determines that a first service flow and a second service flow are anti-affinity service flows, the SDN controller selects the first service flow as a target service flow from two anti-affinity service flows based on a preset priority order, and selects the second service flow as an alternative service flow, and only selects a third path as the target path after removing an equivalent next hop Spin1 and Spin 2 of the tunnel path corresponding to the second service flow. The SDN controller judges whether the scheduling of the first service flow can cause the congestion of the third path according to the pre-calculation result of the rerouting, if not, the finally selected route of the first service flow and the next hop (Spine 3) corresponding to the third path is used as the selected service flow scheduling scheme to be issued to the Leaf1. If congestion is caused, the SDN controller reselects the second traffic flow as a target traffic flow and performs the step of selecting the target path again, and similarly selects a third path as the target path, then determines whether the second traffic flow is scheduled to the third path to cause congestion, and if congestion is not caused, issues the finally selected second traffic flow and the route of the next hop (spin 3) corresponding to the third path as the selected traffic flow scheduling scheme to Leaf1. If any one of the first service flow and the second service flow is selected as the target service flow, the condition that the third path congestion is not caused cannot be met, the selection process of the selected service flow scheduling scheme is terminated, and the scheduling process is terminated.

And 304, based on the selected service flow scheduling scheme, issuing a new route of the selected scheduled service flow to the forwarding equipment so that the forwarding equipment updates a routing table, and scheduling the selected scheduled service flow to other tunnel paths for transmission.

In the application scenario with equivalent paths as shown in fig. 2, the finally selected first traffic flow and the route of the next hop (Spine 3) corresponding to the third path are used as the selected traffic flow scheduling scheme, the SDN controller issues a new route of the third path (the next hop is Spine 3) of the first traffic flow to the forwarding device Leaf1 based on the selected traffic flow scheduling scheme, so that the Leaf1 updates the routing table based on the new route, and schedules the first traffic flow to be scheduled on the third path for transmission.

By the technical scheme provided by the invention, the SDN controller can perform real-time balanced scheduling on the service flow in the data center network, so that high burst packet loss caused by superposition of the service flow is avoided, the result of routing selection is changed by using the anti-affinity characteristic, the strategy routing mode is avoided, the issuing of strategy rules is reduced, the hardware resource of the switch equipment is greatly saved, and the maintenance cost is reduced.

Fig. 4 is a schematic structural diagram of an electronic device for implementing the method for scheduling an anti-affinity service flow according to an embodiment of the present invention, where the device 400 includes: a processor 410 such as a Central Processing Unit (CPU), a communication bus 420, a communication interface 440, and a memory 430. Wherein the processor 410 and the memory 430 may communicate with each other via a communication bus 420. The memory 430 stores a computer program that, when executed by the processor 410, performs the functions of one or more steps of the anti-affinity traffic scheduling method provided by the present invention.

Memory refers to a device for storing computer programs and/or data based on some storage medium, which may be a Volatile Memory (VM) or a Non-Volatile Memory (NVM). The memory is an internal memory for directly exchanging data with the processor, and can read and write data at any time, and has high speed, and is used as a storage medium for temporary data of an operating system and other running programs. The memory may be synchronous dynamic random access memory (Synchronous Dynamic Random Access Memory, SDRAM), dynamic random access memory (Dynamic Random Access Memory, DRAM), or the like. The nonvolatile memory is a memory using a persistent storage medium, and has a large capacity and can store data permanently, and may be a storage class memory (Storage Class Memory, SCM), a Solid State Disk (SSD), a NAND flash memory, a magnetic Disk, or the like. SCM is a common name for new storage medium between memory and flash memory, and is a composite storage technology combining persistent storage characteristic and memory characteristic, and has access speed slower than that of DRAM and SSD hard disk.

The processor may be a general-purpose processor including a central processing unit (Central Processing Unit, CPU), a network processor (Network Processor, NP), etc.; but also digital signal processors (Digital Signal Processing, DSP), application specific integrated circuits (Application Specific Integrated Circuit, ASIC), field programmable gate arrays (Field-Programmable Gate Array, FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components.

It should be appreciated that embodiments of the invention may be implemented or realized by computer hardware, a combination of hardware and software, or by computer instructions stored in non-transitory (or referred to as non-persistent) memory. The method may be implemented in a computer program using standard programming techniques, including a non-transitory storage medium configured with a computer program, where the storage medium so configured causes a computer to operate in a specific and predefined manner. Each program may be implemented in a high level procedural or object oriented programming language to communicate with a computer system. However, the program(s) can be implemented in assembly or machine language, if desired. In any case, the language may be a compiled or interpreted language. Furthermore, the program can be run on a programmed application specific integrated circuit for this purpose. Furthermore, the operations of the processes described in the present invention may be performed in any suitable order unless otherwise indicated herein or otherwise clearly contradicted by context. The processes (or variations and/or combinations thereof) described herein may be performed under control of one or more computer systems configured with executable instructions, and may be implemented as code (e.g., executable instructions, one or more computer programs, or one or more applications), by hardware, or combinations thereof, collectively executing on one or more processors. The computer program includes a plurality of instructions executable by one or more processors.

Further, the method may be implemented in any type of computing platform operatively connected to a suitable computing platform, including, but not limited to, a personal computer, mini-computer, mainframe, workstation, network or distributed computing environment, separate or integrated computer platform, or in communication with a charged particle tool or other imaging device, and so forth. Aspects of the invention may be implemented in machine-readable code stored on a non-transitory storage medium or device, whether removable or integrated into a computing platform, such as a hard disk, optical read and/or write storage medium, RAM, ROM, etc., such that it is readable by a programmable computer, which when read by a computer, is operable to configure and operate the computer to perform the processes described herein. Further, the machine readable code, or portions thereof, may be transmitted over a wired or wireless network. When such media includes instructions or programs that, in conjunction with a microprocessor or other data processor, implement the steps described above, the invention described herein includes these and other different types of non-transitory computer-readable storage media. The invention also includes the computer itself when programmed according to the methods and techniques of the present invention.

The foregoing is merely exemplary of the present invention and is not intended to limit the present invention. Various modifications and variations of the present invention will be apparent to those skilled in the art. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. The method for scheduling the anti-affinity service flow is characterized by being applied to a software defined network SDN controller, and comprises the following steps:

and based on the selected service flow scheduling scheme, transmitting the new route of the selected scheduled service flow to the forwarding equipment so that the forwarding equipment updates a routing table, and scheduling the selected scheduled service flow to other tunnel paths for transmission.

2. The method of claim 1, wherein the SDN controller monitors traffic on a plurality of tunnel paths within a traffic engineering tunnel in real time by:

3. The method according to claim 1, wherein the method for identifying the anti-affinity traffic is:

4. A method according to claim 3, wherein the means for obtaining traffic transmission characteristics of traffic on the congested tunnel path is:

5. The method of claim 1, wherein the performing the rerouting precalculation selects a traffic scheduling scheme that does not cause congestion of other tunnel paths based on a result of the rerouting precalculation:

selecting a target service flow from the anti-affinity service flow according to a preset selection strategy, determining the rest service flow as an alternative service flow, removing the next hop of the tunnel path corresponding to the alternative service flow, and selecting an optimal path in the rest tunnel path as a target path based on the preset path selection strategy;

6. The method of claim 5, wherein the step of determining the position of the probe is performed,

when congestion is judged to be caused, a different service flow is reselected in the anti-affinity service flow as a target service flow according to a preset selection strategy, and the selection step of the target path is executed again until a target path which does not cause congestion is selected, and the finally selected target service flow and the next hop route corresponding to the target path are used as a selected service flow scheduling scheme;

If any combination of the target service flow and the target path finally cannot meet the condition of not causing congestion of the target path, the selection process of the selected service flow scheduling scheme is terminated, and the scheduling process is ended.

7. The method of claim 5, wherein the selecting an optimal path as the target path based on a preset path selection policy in the remaining tunnel paths comprises:

8. The method of claim 5, wherein the step of determining the position of the probe is performed,

the plurality of tunnel paths within the traffic engineering tunnel include equivalent paths;

9. An anti-affinity traffic scheduling device, wherein the device is applied to a software defined network SDN controller, the device comprising:

and the route issuing module is used for issuing the new route of the selected scheduled service flow to the forwarding equipment based on the selected service flow scheduling scheme so that the forwarding equipment updates the routing table and schedules the selected scheduled service flow to other tunnel paths for transmission.

10. The apparatus of claim 9, wherein the device comprises a plurality of sensors,

the monitoring module monitors the service flow state of each tunnel path on each forwarding device in real time through a flow monitoring mechanism based on the managed flow engineering tunnel information and routing information, and when the tunnel path generates congestion, the anti-affinity service flow determining module is triggered to carry out anti-affinity service flow identification according to the congestion alarm reported by the forwarding device.

11. The apparatus of claim 9, wherein the anti-affinity traffic flow determination module comprises:

12. The apparatus of claim 9, wherein the scheduling scheme determination module comprises:

13. The apparatus of claim 12, wherein the device comprises a plurality of sensors,

the target service flow and target path determining unit selects a tunnel path with the best matching affinity attribute from the remaining tunnel paths as a target path; or comprehensively selecting a tunnel path which not only accords with the affinity attribute of the tunnel path but also can balance the link load from the rest tunnel paths based on the affinity attribute and the link load as a target path.

14. The apparatus of claim 12, wherein the device comprises a plurality of sensors,

15. An electronic device is characterized by comprising a processor, a communication interface, a storage medium and a communication bus, wherein the processor, the communication interface and the storage medium are communicated with each other through the communication bus;

a storage medium storing a computer program;

a processor for carrying out the method steps of any one of claims 1-8 when executing a computer program stored on a storage medium.

16. A storage medium having stored thereon a computer program which, when executed by a processor, implements the method of any of claims 1 to 8.