KR101958943B1 - Method and apparatus for multicast routing in software defined networking environment - Google Patents

Abstract

The present invention relates to a multicast routing method, apparatus and computer program. According to one aspect of the present invention, a method for performing multicast routing in a software defined networking environment comprises the steps of selecting a rendezvous point (RP) of a first multicast group and performing multicast routing using the RP. The step of selecting the RP includes (a) setting a second network equipment connected to a server or a host of a first network equipment belonging to the first multicast group as an edge node and a remaining third network equipment as a node; (b) calculating a cost per node by using at least one of the number of hops between the node and the edge node or a link traffic; and (c) selecting a first node having a minimum path cost according to the calculation result as the RP. According to the present invention, a network traffic environment can be dynamically reflected by automatically selecting the RP based on a network topology and a link state.

Description

[0001] METHOD AND APPARATUS FOR MULTICAST ROUTING IN SOFTWARE DEFINED NETWORKING ENVIRONMENT [0002]

The present invention relates to a multicast routing method, apparatus and computer program, and more particularly, to a multicast routing method and apparatus in a software defined networking environment.

Software Defined Networking (SDN) is a technology that manages all the network devices in the network by an intelligent central management system. In the SDN technology, a controller provided in a form of software can perform processing instead of a control operation related to the packet processing performed by itself in a conventional hardware type network device, so that various functions can be developed and given over the existing network structure. The SDN system generally comprises a controller server for controlling the entire network, a plurality of open flow switches controlled by the controller server for processing packets, and a host corresponding to a lower layer of the open flow switch. Here, the open flow switch is only responsible for transmitting and receiving packets, and routing, management, and control of the packets are all performed in the controller server. In other words, separating the data planes and control planes that form the network equipment is the basic structure of the SDN system.

On the other hand, there is Protocol Independent Multicast (PIM) as a routing protocol that enables multicast routing that does not depend on protocols in a large network in relation to multicast routing. PIM can be divided into Dense Mode (DM) and Sparse Mode (SM).

Protocol Independent Multicast-Dense Mode (PIM-DM) is mainly used in a LAN environment where group members are concentrated in a specific place, such as a company or a school, and is used when each router is likely to participate in multicasting.

Protocol Independent Multicast-Sparse Mode (PIM-SM) is used in an environment where group members are distributed in various places. It is a method of constructing a multicast network by using a forwarding table that a router has basically while reducing network load to be. It is used when the possibility of each router participating in multicasting is relatively low, and a transmission path is created based on a shared tree. PIM-SM is designed to route multicast packets into multicast groups and efficiently construct distribution trees on the WAN. PIM-SM is protocol independent because it can use route information on which routing protocols enter the multicast routing information base (RIB).

In PIM-SM, multicast traffic of a source such as a stream server is first delivered to a Randezvous Point (RP), and the RP serves as a root of a shared tree for multicast traffic transmission. do. Thereafter, a shortest-path tree (hereinafter referred to as SPT) is generated for multicast traffic transmission through information exchange between the RP and the neighboring routers. SPT is also referred to as a source-based tree, where the delivery path is based on the shortest unicast path to the source.

What matters here is what router to choose as the RP. There are static selection and dynamic selection techniques for RP selection. The static selection scheme consists of each leaf router as the address of a group or group of groups and does not require any special algorithm.

The dynamic selection technique is as shown in FIG. Referring to FIG. 1, the dynamic selection technique uses a Bootstrap Router (BSR) to generate bootstrap messages. First, a router is set up to perform a BSR and a Candidate Rendezvous Point (C-RP) manually. When the C-RP periodically transmits a advertisement message containing its own information to the BSR (A) Based on the collected C-RP information such as priority, hash value, and IP address information of the multicast group, and generates RP-Set information in which the RP is mapped for each multicast group. Then, the BSR can select the RP by transmitting the RP-Set information to the BSR message (B) periodically to all the routers.

According to this conventional method, the RP must be manually set for multicast transmission, and even when the RP is set dynamically, the RP management is very inefficient because the configuration for the BSR and the C-RP is required for each router. In addition, since C-RP must continuously transmit its information through network, there is a problem that network management overhead increases. In addition, RP selection does not dynamically reflect the network traffic environment because RP is selected through comparison of C-RP priority, hash value, and IP address. If failure occurs in C-RP or BSR failure occurs, multicast It is difficult to transmit, and in case of failure, it is necessary to manually reset C-RP and BSR, which is vulnerable to network traffic and failure.

It is an object of the present invention to provide a multicast routing method capable of dynamically reflecting a network traffic environment by automatically selecting an RP based on a network topology and a link state.

In addition, the present invention recognizes the topology change occurring in the case of a link failure or a switch failure, and automatically reselects the RP for each multicast group based on the change in topology. Thus, multicast routing It is another purpose to provide a method.

Further, according to the present invention, when a change occurs in a host in a multicast group, the RP is reselected adaptively, thereby preventing excessive multicast traffic from being concentrated on a specific link, and consequently, multicast routing It is another purpose to provide a method.

Another object of the present invention is to provide a multicast group traffic load balancing function in a multicast environment.

In accordance with another aspect of the present invention, there is provided a method for multicast routing in a software defined networking environment, the method comprising: selecting a rendezvous point (RP) of a first multicast group; Wherein the RP selecting step sets a second network equipment connected to a server or a host among the first network equipment belonging to the first multicast group as an edge node and the remaining third network equipment as a node a step b) of calculating the path cost for each node by using at least one of the number of hops between the node and the edge node or the link traffic, selecting the first node having the minimum path cost according to the calculation result, c < / RTI >

The present invention also provides an apparatus for performing multicast routing in a software defined networking environment, the apparatus comprising: a controller for selecting a rendezvous point (RP) of a first multicast group and performing multicast routing using the RP; A communication unit for transmitting a flow rule according to the multicast routing to at least one first network equipment belonging to the first multicast group and for receiving a message relating to the network environment from the first network equipment, Wherein the application unit sets a second network equipment connected to a server or a host among the first network equipment belonging to the first multicast group as an edge node and the remaining third network equipment as a node, The number of hops or rings between the node and the edge node An RP selecting unit for selecting a first node having a minimum path cost according to the calculation result, and a transmission path tree having the RP as a root are generated And a multicast routing unit for converting the transmission path tree into a routing tree having an edge node connected to the server as a root.

According to the present invention as described above, the network traffic environment can be dynamically reflected by automatically selecting the RP based on the network topology and the link status.

Further, the present invention recognizes topology changes occurring in the case of a link failure or a switch failure, and automatically selects the RPs for each multicast group based on the topology change, thereby ensuring continuous transmission of multicast traffic even in a network failure situation.

Also, according to the present invention, when a change occurs in a host in a multicast group, the RP is adaptively selected, thereby preventing excessive multicast traffic from being concentrated on a specific link, and consequently, network utilization efficiency can be increased.

Also, the present invention can provide a multicast group traffic load balancing function in a multicast environment.

FIG. 1 is a view for explaining a conventional method of setting a rendezvous point (RP)
2 is a diagram illustrating a software defined networking environment according to an embodiment of the present invention;
3 is a block diagram for explaining a controller according to an embodiment of the present invention.
4 is a diagram for explaining RP selection and multicast routing of a controller according to an embodiment of the present invention;
5 is a diagram for explaining a transmission path tree and a routing tree according to an embodiment of the present invention;
6 is a flowchart illustrating a multicast routing method of a controller according to an embodiment of the present invention.
FIG. 7 illustrates one embodiment when a failure occurs in any link;
FIG. 8 is a diagram for explaining a transmission path tree and a routing tree according to the embodiment of FIG. 7;
Fig. 9 is a diagram showing an embodiment when a failure occurs in a switch, Fig.
FIG. 10 is a diagram for explaining a transmission path tree and a routing tree according to the embodiment of FIG. 9;
11 is a diagram illustrating an embodiment in which a multicast group is changed;
FIG. 12 is a diagram for explaining a transmission path tree and a routing tree according to the embodiment of FIG. 11;
13 is a diagram illustrating an embodiment in which multicast routing is performed by reflecting network traffic,
14 is a diagram for explaining a transmission path tree and a routing tree according to the embodiment of FIG.

The above and other objects, features, and advantages of the present invention will become more apparent by describing in detail exemplary embodiments thereof with reference to the attached drawings, which are not intended to limit the scope of the present invention. In the following description, well-known functions or constructions are not described in detail since they would obscure the invention in unnecessary detail. Hereinafter, preferred embodiments of the present invention will be described in detail with reference to the accompanying drawings. In the drawings, the same reference numerals are used to designate the same or similar components, and all combinations described in the specification and claims can be combined in any manner. It is to be understood that, unless the context requires otherwise, references to singular forms may include more than one, and references to singular forms may also include plural forms.

The term 'flow rule' in the description of the present invention should be understood to mean a network policy applied by a controller server in a software defined network in the ordinary artisan's field.

Also, the controller of the present invention manages a switch for each multicast group, and manages a switch and a host included in the entire multicast group, Can manage the network topology information of the network.

2 is a diagram illustrating a software defined networking environment according to an embodiment of the present invention. Referring to FIG. 2, a software defined networking environment may include a server 50, a controller 100, a network device 200, and a host 300. The network device 200 and the host 300 may be referred to as a node, and a link may denote a connection between two nodes.

The controller 100 manages the network equipment 200 and centrally manages and controls the plurality of network equipment 200. Specifically, the controller 100 is configured to perform network topology management including a server and a host included in the entire multicast group, path management related to packet processing, link discovery, (Software) that performs functions such as flow management and the like.

The switch 200 functions to process packets under the control of the controller 100. The switch 200 shown in the figure can be exchanged with a network device such as a mobile communication base station, a base station controller, a gateway device, a router, etc. for packet processing. However, for convenience of explanation, the case where the network equipment 200 is the open flow switch 200 will be mainly described, and the same reference numerals will be given.

Meanwhile, the server 50 and the host 300 that transmit the multicast traffic are connected to the open flow switch 200 managed by the controller 100, and the switch 200 is controlled according to a flow rule received from the controller 100 Multicast traffic can be transmitted and received by forwarding packets.

3 is a block diagram for explaining a controller 100 according to an embodiment of the present invention. 3, the controller 100 according to an exemplary embodiment of the present invention may include an application unit 130, a communication unit 150, and a storage unit 170. The application unit 130 may include an RP pre- A multicast routing unit 135, and a multicast routing unit 137.

The application unit 130 is an area in which various functions performed by the controller 100 are implemented, and may include an application for performing the network topology management, the path management, and the like described above. According to an embodiment of the present invention, an application unit 130 for selecting a rendezvous point (RP) of a first multicast group and performing multicast routing using the RP includes an RP selection unit 135, a multicast routing (137).

The RP selection unit 135 sets the second network equipment connected to the server or the host among the first network equipment belonging to the first multicast group as an edge node and the remaining third network equipment as a node, The path cost for each node may be calculated using at least one of the number of links and the link traffic, and the first node having the lowest path cost may be selected as the RP according to the calculation result.

More specifically, a rendezvous point (hereinafter referred to as RP) selection method used in multicast routing according to an embodiment of the present invention will be described with reference to FIG.

Rendezvous point, which is a concept used in the conventional PIM protocol, is a kind of multicast router that receives all the traffic from the server (source) and delivers all the traffic to the receiver. The RP is used in a structure where all the multicast streams are collected into one distribution node and then distributed to the receivers. The tree with the RP as the root is called a shared tree or RP tree. According to the PIM protocol, all multicast sources send a multicast stream to the RP, and all listeners join the RP to receive the multicast stream using the shared tree.

The RP used in one embodiment of the present invention is similar in functionality to the RP of the prior art but includes similar features in that it creates a root tree for the RP in multicast routing, The term widely used by. However, the RP selected by the RP selecting unit 135 of the present invention is not for receiving all the multicast traffic received from the source and delivering it to the receiver, but for generating the transmission path tree before the generation of the routing tree Its role is completely different.

Also, in the method of selecting the RP, the PIM protocol requires a step of manually setting the RP as described above. However, the present invention is not limited to the case where the RP selection unit 135 calculates the path cost between the edge node and the node, There is a big difference in that the RP is set dynamically according to the network conditions.

The RP selection unit 135 distinguishes the switch 200 that is connected only to the switch 200 connected to the server or the host and the switch 200 connected to the other switch. In this specification, the former is referred to as an edge node and the latter as a node. The RP selection unit 135 can calculate the minimum path cost based on the number of path hops and link traffic between each edge node and the node.

, 이전 트래픽을

라고 할 때, 다음과 같이 연산할 수 있다.For example, in the case of one hop, the path cost may be set to 1, and the link traffic may be additionally reflected as the weight W. Thus, the path cost between each switch 200 in the embodiment of FIG. 4 may be 1 + W. The weight (W) can be calculated as follows:

, Previous traffic

, The following operation can be performed.

In the example of FIG. 4, it is assumed that there is no previous traffic (W = 0). Only the number of hops will be reflected in the path cost, and the path cost of the link between each switch 200 can be calculated as shown in the following table.

SW # 2 SW # 3 SW # 1 One 2 SW # 4 2 3 SW # 5 2 2 SW # 6 One One Cost 6 8

The RP selection unit 135 can select a node having the minimum path cost as the RP of the multicast group. Therefore, in the above example, switch # 2 with the minimum path cost will be selected as the RP.

The multicast routing unit 137 may generate a transmission path tree having the root of the RP and convert the transmission path tree into a routing tree having an edge node connected to the server as a root. In the above example, the multicast routing unit 137 can generate a transmission path tree having the switch # 2 as the root, as shown in the <A> in FIG. The controller 100 can manage a transmission path for each multicast group. The multicast routing unit 137 assigns a routing tree having an edge node connected to a source (server) as a root in each multicast group to < B > You can create it together. That is, a transmission path tree can be created and a routing tree can be created by converting the tree so that the edge node connected to the server in the transmission path tree is at the root. The controller 100 may then cause the multicast traffic to be delivered along the routing tree. According to this method, there is no need to join multicast traffic receivers whenever the RP wants to, like PIM-SM, and there is no need to pruning where multicasting is unnecessary like PIM-DM. Thus, it is possible to reflect changes in the host (receiver) in the multicast group in a much simpler way.

The communication unit 150 can communicate with the switch 200. The communication unit 140 may transmit a flow rule according to multicast routing to one or more first network devices belonging to the first multicast group and receive a message regarding the network environment from the first network equipment. The message thus received can be used by the controller 100 to determine a network failure. The communication unit 150 can exchange information with the switch 200 through a secure channel and the information exchanged with the switch 200 can be encrypted.

 The storage unit 170 may store various information such as the status of the controller application and information necessary for operating the application, the flow rule transmitted from the controller to the switch 200, the status of the switch 200, and the network topology information. In particular, the storage unit 170 may store topology information for each multicast group.

Hereinafter, a multicast routing method of a controller and a routing method in a fault situation will be described in detail with reference to FIG. 6 to FIG.

6 is a flowchart illustrating a multicast routing method of a controller according to an embodiment of the present invention.

Referring to FIG. 6, the controller first selects a rendezvous point RP of the first multicast group (S100), and when the RP is selected, performs multicast routing using the selected RP (S200). More specifically, in step 100, the controller sets the second network equipment connected to the server or host among the first network equipment belonging to the first multicast group as an edge node and the remaining third network equipment as a node to distinguish the edge node and the node (S130). Next, the controller can calculate the path cost per node using at least one of the number of hops between the node and the edge node or the link traffic (S150). The first node having the lowest path cost may be selected as the RP according to the calculation result (S170).

In step 150, the controller may set an initial value of the path cost in proportion to the number of hops, and add link traffic to the initial value of the path cost as a weight. A concrete example of this has been described above with reference to [Table 1] and [Equation 1].

In step 170, the controller selects an arbitrary one of the nodes RP as the operation result of step 160 if there are two or more nodes having the lowest path cost. If either of the two nodes is the RP selected node, the controller can keep the existing RP.

FIG. 7 is a diagram showing a case where a failure occurs in the link between the switch 1 and the switch 2. FIG. If any link fails, the controller can receive a failure report message containing a port down message from the network equipment associated with the link. The controller can recognize the fault condition for the link using the fault report message.

 For example, as shown in FIG. 7, when a failure occurs in a link connecting switch 1 and switch 2, switch 1 and switch 2 transmit a port down message to the controller. The controller can use the port down messages received at switch 1 and switch 2 to acknowledge that the link connecting switch 1 and switch 2 has failed. The controller can change the topology information by deleting the link from the topology information of the multicast group network including the link. Then, the controller can re-determine the RP using the changed topology information and perform the multicast routing again using the re-selected RP.

In this example, if the link between switch 1 and switch 2 does not operate, the path cost to the edge node for each node may be computed as follows.

SW # 2 SW # 3 SW # 1 3 3 SW # 4 3 3 SW # 5 2 2 SW # 6 One One Cost 8 8

In the above embodiment, it is assumed that the path traffic is not reflected. The path cost of switches 2 and 3 is proportional to the number of hops and the result of summing the number of hops with each edge node is the same. If the route cost is the same as described above, either one of them can be arbitrarily selected as RP. If the previous RP was switch 2, it is not necessary to change the RP unnecessarily if the path cost is the same, so the RP can remain unchanged. However, since the link between the switches 1 and 2 does not operate, the network topology is changed, and the transmission path tree of the multicast group with the switch 2 as the root can be generated as shown in FIG. 8A.

The controller uses the transmission path tree to create the routing tree. After creating the transmission path tree, the controller can change the tree shape so that the server (source) included in the transmission path tree becomes the root. According to this method, the connection between each node constituting the routing tree is the same as the transmission path tree, but a new tree in which only the root is changed can be obtained.

9 is an example of a case where a failure occurs in the switch. The controller according to an embodiment of the present invention can receive a status message (hello message) from each of the switches constituting the multicast group at predetermined intervals to check whether the switch has failed or not.

When a failure occurs in any network equipment (switch 2) belonging to the multicast group shown in FIG. 9, it is impossible to transmit a packet to the adjacent switch, so that the switch 1, the switch 3, It is possible to transmit a link failure message to the controller that corresponds to a failure to be received for a predetermined time or longer.

If a problem occurs in the switch 2, the controller can recognize the problem situation of the switch 2 due to the disconnection of the status message received from the switch 2. Further, since the switch 2 receives a link failure message such as a port down message from the switch 2 and the adjacent switch, the controller can recognize the failure situation through the received message.

When a fault condition is recognized, the controller can change the network topology to reflect the fact that certain links and switches no longer work. The controller then re-selects the RP based on the changed topology information. That is, it is possible to calculate the total path cost per node by calculating the path cost between the edge nodes for each node and summing them. The path cost for each node in the above example is shown in the following table.

SW # 2 SW # 3 SW # 1 INF 3 SW # 4 INF 3 SW # 5 INF 2 SW # 6 INF One Cost INF 8

The above example also assumes that there is no previous traffic (BW = 0) and the weight w due to the route traffic is zero. In the case of switch 2, which is a node, operation is impossible, so that the number of hops between switch 2 and edge node can be set to infinity. The number of hops between switch 3 and each edge node is as shown in [Table 3], and since the node having the minimum path cost in the changed topology is switch 3, switch 3 can be selected as a new RP.

The transmission path tree with the switch 3 as the root is the same as the <A> in FIG. 10, and the routing tree with the edge node (switch 1) connected to the server (source) in the transmission path tree as the root is generated as shown in <B>. The controller sends flow rules for multicast routing to each switch based on the routing tree, and the switches carry multicast traffic based on this routing tree. Therefore, in the example of FIG. 10, the data stream transmitted from the server can be transmitted to each host via the switch 1 and the switch 5.

Fig. 11 shows an embodiment in which the multicast group is changed. If a host in the multicast group is added or deleted, the controller can reselect the RP.

In the example of FIG. 11, the host B has been deleted from the multicast group. When a host in a multicast group is deleted, the controller can recalculate the path cost of a node belonging to the multicast group. When host B is deleted, switch 5, which was previously set as an edge node, is reset to the node.

Therefore, the path cost to the edge node for each node is calculated as shown in the following table.

SW # 2 SW # 3 SW # 5 SW # 1 One 2 One SW # 4 2 3 One SW # 6 One One One Cost 4 6 3

The nodes are switches 2, 3, and 5, and the edge nodes are divided into 1, 4, and 6, and as a result of summing the number of hops between each node and the edge node, . Therefore, the switch 5 is selected as the RP, and a transmission path tree is generated based on the re-selected RP to obtain the tree shown in FIG. 12A. The controller converts <A> to a routing tree, and the conversion method is the same as described above.

13 is a diagram illustrating a case where multicast routing is performed by reflecting network traffic.

The traffic of the switch 1-4 link, the switch 1-2 link, the switch 1-5 link, and the switch 2-6 link is 1, 0.5, 0.5, 1 each as shown in the figure .

The controller monitors the link traffic using the traffic information transmitted from the switch. If it is determined that the monitoring result traffic is concentrated in an arbitrary path, the controller calculates the path cost for each node again using the changed path cost, The node having the path cost is re-selected as the RP.

In the example of FIG. 13, the path cost of switches 2 and 3, which are nodes, can be calculated as shown in the following table.

The path cost of switches 4-5, 5-6, and 6-3 is 1, and the path cost of switches 1-2 is 1 And the path cost of the switch 2-4 can be calculated as 3.5, which is the number of hops of 2 and the weight of 1.5 is added. Therefore, the total sum of the path cost of each node is calculated as 9 for switch 2 and 8.5 for switch 3, and the RP changed by the traffic change will be switch 3.

SW # 2 SW # 3 SW # 1 1.5 2.5 SW # 4 3.5 3 SW # 5 3 2 SW # 6 2 One Cost 9 8.5

When the RP is selected as the switch 3, the transmission path of the multicast traffic according to the newly generated routing tree is shown in FIG. That is, according to the transmission path shown in FIG. 14, as shown in FIG. 13, traffic is transmitted while avoiding a path in which traffic increases sharply, thereby increasing network utilization efficiency.

Some embodiments omitted in this specification are equally applicable if their implementation subject is the same. It is to be understood that both the foregoing general description and the following detailed description of the present invention are exemplary and explanatory and are intended to be exemplary and explanatory only and are not restrictive of the invention, The present invention is not limited to the drawings.

50: Server
100: controller
200: Network equipment
300: Host

Claims (8)

In a method for a controller to perform multicast routing in a software defined networking environment,
Selecting a rendezvous point (RP) of a first multicast group;
Generating a transmission path tree having the RP as a root, converting the transmission path tree into a routing tree having an edge node connected to the server as a root, and performing routing according to the routing tree,
The RP selection step
Setting a second network equipment connected to a server or a host among the first network equipment belonging to the first multicast group as an edge node and the remaining third network equipment as a node;
B) calculating the per-node path cost using at least one of the number of hops between the node and the edge node or the link traffic;
And selecting a first node having a minimum path cost as an RP according to the calculation result.
The method according to claim 1,
The step b
Setting an initial value of a path cost in proportion to the number of hops;
And adding the link traffic as a weight to an initial value of the path cost.
The method according to claim 1,
Step c
And selecting an arbitrary one of the first nodes as an RP if the first node has more than two first nodes.
The method according to claim 1,
If it is determined that a failure has occurred in an arbitrary link,
Receiving a port down message from a network device connected to the link;
Changing topology information of the first multicast group using the port down message;
And executing the RP selection step and the multicast routing step again using the changed topology information.
The method according to claim 1,
Further comprising receiving a status message from the first network device at predetermined intervals,
A fourth multicast group, a fourth network device, and a fourth network device, the fourth network device being connected to the fourth network device, the fourth network device being connected to the fourth network device, If it is determined that the equipment has failed,
Changing topology information of the first multicast group using the port down message and the status message reception;
And executing the RP selection step and the multicast routing step again using the changed topology information.
The method according to claim 1,
Monitoring the link traffic using a message received from the first network equipment;
If it is determined that the monitoring result traffic is concentrated in an arbitrary path,
And performing the step b to step c and the step of performing the multicast routing again using the modified path cost.
An apparatus for performing multicast routing in a software defined networking environment,
An application unit for selecting a rendezvous point (RP) of a first multicast group and performing multicast routing using the RP;
A communication unit for transmitting a flow rule according to the multicast routing to at least one first network equipment belonging to the first multicast group and receiving a message regarding the network environment from the first network equipment;
And a storage unit for storing topology information of the first multicast group,
The application unit
A second network device connected to a server or a host among the first network devices belonging to the first multicast group is set as an edge node and the remaining third network device is set as a node and the number of hops or link traffic between the node and the edge node An RP selection unit for calculating a path cost for each node by using at least one of the RPs and an RP and selecting a first node having a minimum path cost according to the calculation result;
And a multicast routing unit for generating a transmission path tree having the RP as a root and converting the transmission path tree into a routing tree having an edge node connected to the server as a root.

delete

Images