KR101155012B1 - Open flow network system and method of controlling the same - Google Patents

Abstract

The present invention relates to an open flow network system and a control method thereof, comprising a plurality of nodes for transmitting and receiving data, an open flow switch for relaying flow transmission and reception between the nodes, and a flow routing controller for controlling the open flow switch. In the openflow network system, the controller generates forwarding for a flow by reflecting flow information received from the openflow switch and a user's request, and forwarding information for the flow in the form of a flow entry in the openflow switch. Transmit to; The open flow switch may receive forwarding of the flow from the controller in the form of a flow entry to manage subsequent traffic based on a forwarding rule included in the entry. As a result, by dynamically assigning traffic transmission paths according to user-defined quality of service requirements and redefining flows according to network conditions, the traffic can be moved to maintain a certain level of network quality. The present invention relates to an open flow network system and a method of controlling the same.

Description

Open flow network system and its control method {OPEN FLOW NETWORK SYSTEM AND METHOD OF CONTROLLING THE SAME}

The present invention relates to an openflow network system and a control method thereof, and more particularly, by dynamically assigning a traffic transmission path according to a service quality requirement defined by a user and redefining the flow according to network conditions to move traffic. The present invention relates to an open flow network system capable of maintaining a certain level of network quality and efficiently using an idle band through dynamic flow movement, and a control method thereof.

OpenFlow technology is a technical standard for testing new protocols over real networks. OpenFlow technology allows you to test network protocols while maintaining existing applications.

The network system to which the open flow technology is applied includes an open flow protocol that provides communication security between a flow table, a controller, and a switch.

Flow tables are built into the switch and define the rules for how to handle packets and traffic. The controller is a kind of command computer that controls the OpenFlow system and communicates with the switch using a specially defined OpenFlow protocol. This protocol is encrypted by Secure Socket Layer (SSL).

OpenFlow systems provide an API for instructing switches, routers, or network controllers, and engineers can use this API to define their own protocols. For example, simple traffic can be defined with a given IP address.

According to this command, the controller can establish a path through the optimized network using network knowledge information. In other words, in the conventional openflow technology, a software-controlled transmission policy determines an optimal traffic path using existing switches and routers.

On the other hand, multimedia traffic such as video conferencing is sensitive to packet loss, so that a certain amount of idle bandwidth must always exist to enable good quality video and voice services.

Therefore, when transmitting multimedia traffic, there is a demand for a method of monitoring the state of a network path being used periodically and transmitting traffic using a bypass path as necessary.

SUMMARY OF THE INVENTION The present invention has been made to solve the above-mentioned problems. The present invention provides a network quality over a predetermined level by dynamically assigning traffic transmission paths according to user-defined service quality requirements, and redefining flows according to network conditions. There is a technical problem to provide an open flow network system and a method of controlling the same, which can maintain the network bandwidth and efficiently utilize the idle band through dynamic flow movement.

The openflow network system of the present invention for achieving the above object includes a plurality of nodes for transmitting and receiving data, an openflow switch for relaying flow transmission and reception between the nodes, and a flow routing controller for controlling the openflow switch. In the open flow network system, the controller generates the forwarding for the flow by reflecting the flow information received from the open flow switch and the user's request, the forwarding information for the flow in the form of a flow entry of the open flow Send to the switch; The openflow switch receives forwarding for the flow from the controller in the form of a flow entry to manage subsequent traffic based on forwarding rules included in the entry; The open flow switch and the controller is characterized in that the transmission and reception of data with each other by TCP or SSL.

The flow entry may include a header field including a MAC, an IP, a service port number of a destination and an origin of the flow; It may include an action field for performing forwarding to a specified port.

delete

In another aspect, the openflow network system of the present invention for achieving the above object, an openflow network system including a plurality of nodes for transmitting and receiving data, and an openflow switch for relaying flow transmission and reception between the nodes. A QoS and configuration information manager comprising: a database storing a QoS policy received from an administrator and network configuration information maintaining a network connection state; Periodically monitors and registers and manages information of the OpenFlow switch and the host, and sets a routing path based on the QoS profile declared for the received packet according to the received event and message type and includes the routing path in the set path. A forwarding and switch communicator for generating a flow entry of an open flow switch; And a routing processor for generating any one of the QoS routing topology and the general routing topology according to the received flow, calculating a shortest path based on each generated topology, and requesting reflection in a network configuration information database. do.

The QoS and configuration information manager may include: a QoS application target input unit configured to generate a table of the database and store and manage and update the QoS related information and network information in the table; A host / switch information processor for registering and managing the host and the open flow switch information; And a route information processor configured to store and manage route information of the flow.

The forwarding and switch communicator may include an open flow message and event processor; A host and switch register configured to register an openflow switch and a host detected by the controller based on the data path event; A physical topology acquiring unit recognizing a physical topology based on the LLDP packet; A bandwidth measuring unit for obtaining a band between nodes using port state information; An L2 switching processor configured to perform switching for transmission and reception of a general flow; A routing selector for making a routing decision based on the QoS profile declared for the received packet; And a forwarding processor configured to generate and transmit a flow entry to an open flow switch included according to a set path.

The routing processor may further include: a QoS routing topology generator configured to process a flow reflecting a user's network requirements; A general routing topology generation unit in charge of general flow processing; It may include a shortest path calculation processing unit for calculating the shortest path based on each of the topologies.

The database may store at least one or more of dijkstra path information, host information, open flow switch connection information, open flow switch list information, QoS policy information, QoS path information, and failed packet information. .

In addition, the Dijkstra Path information, QoS policy information, QoS path information, and failed packet information may include the MAC address of the transmitting and receiving side, the IP address of the transmitting and receiving side, the service port number of the transmitting and receiving side, the type in the data link layer, and It includes a network protocol number, and may further include additional information according to the attribute of each information.

The host information may include address information and access information of a host subscribed to the open flow network, bandwidth information of a receiving side and a transmitting side, and information on whether a host is connected and whether to use it when setting a path.

In addition, the open flow switch connection information may include address information and access information of the open flow switch, reception side and transmission side bandwidth information, and last received packet information.

According to another aspect of the present invention, there is provided a method of controlling an open flow network system, the method comprising: determining whether a flow of a received packet is a general flow or a QoS flow; Creating a QoS routing topology if the QoS flow; Setting a shortest path based on each generated topology; Generating a flow entry based on the shortest path; Sending the flow entry to an openflow switch; And managing, by the open flow switch, subsequent traffic based on a forwarding rule included in the flow entry.

Here, when the received flow is a general flow, generating a shortest path (Dijkstra) routing topology; Setting a shortest path based on each generated topology; And transmitting traffic by controlling the L2 switch according to the path.

On the other hand, periodically receiving port-specific statistical data of the open flow switch; Calculating bandwidth between links using the port-specific statistical information; Setting a routing flag according to the bandwidth; The method may further include updating the bandwidth information included in pre-stored host information and connection information of the open flow switch.

In addition, when a data path event occurs, the method may further include adding or deleting connection information of the corresponding host and the open flow switch.

As described above, the open flow network system and its control method of the present invention dynamically allocate traffic transmission paths according to user-defined quality of service requirements, redefine the flow according to network conditions, and move the traffic above a certain level. Network quality can be maintained and dynamic flow movements can efficiently utilize idle bands.

1 is a block diagram of an openflow network system according to an embodiment of the present invention;
2 is a control block diagram of an openflow network system according to an embodiment of the present invention;
3 is a schema structure diagram of a database of an openflow network system according to an embodiment of the present invention;
4 is a control flowchart of an openflow network system according to a first embodiment of the present invention;
5 is a control flowchart of an openflow network system according to a second embodiment of the present invention.

Hereinafter, exemplary embodiments of the present invention will be described in detail with reference to the accompanying drawings.

1 is a block diagram of an openflow network system according to an embodiment of the present invention.

The open flow network system includes a plurality of nodes 10 for transmitting and receiving data, an open flow switch 20 for relaying flow transmission and reception between each node 10, and flow information from the open flow switch 20 to receive flow information. It includes a flow routing controller 100 for controlling the routing. Here, the open flow switch 20 and the flow routing controller 100 may exchange information through a communication method such as TCP or SSL.

In the communication between the nodes 10, the transmitting node 10 requests a communication to the receiving node 10 to be connected to the open flow switch 20 side.

The open flow switch 20 stores the data received from the transmitting node 10 in a buffer and transmits information on the requested flow to the open flow routing controller 100. At this time, the exchange of information on the flow is performed using an open flow protocol.

The controller 100 generates a forwarding path for the flow by reflecting the flow information received from the open flow switch 20 and the user's request. The open flow routing controller 100 transmits the generated forwarding path to the open flow switch 20 in the form of a flow entry.

Receiving the flow entry, the open flow switch 20 manages subsequent traffic based on forwarding rules contained in the entry. The flow entry includes a header field including a destination / source MAC, an IP, a service port number, and the like, and an action field for forwarding to a designated port.

The present invention proposes a system capable of flow routing by distinguishing between a general flow and a QoS flow defined by the network policy manager 110 to ensure differentiated requirements in such an open flow network system.

2 is a control block diagram of an open flow network system according to an embodiment of the present invention, and shows a control configuration for performing a dynamic flow routing function in consideration of traffic load.

The open flow network system includes a QoS and configuration information manager 200, a routing processor 300, a forwarding and switch communicator 400.

The QoS and configuration information manager 200 includes a QoS profile and configuration information database 220 in which the QoS policy received from the openflow network policy manager 110 and network configuration information for maintaining a network connection state are stored.

The QoS and configuration information manager 200 generates a table in the database 220 and manages and updates the information stored in the table. The QoS application target input unit 210, the host / switch information processing unit 240, and the path information processing unit ( 230). Here, a description of the schema structure and fields of the database 220 will be described in detail with reference to FIG. 3 to be described later.

The forwarding and switch communicator 400 includes an openflow message and event processor 410, a host and switch register 420, a physical topology acquirer 430, a bandwidth measurer 440, an L2 switching processor 450, and routing selection. The unit 470 includes a forwarding processor 460.

The open flow message and event processor 410 processes and responds to the open flow protocol.

The host and switch registration unit 420 registers the open flow switch 20 and the host detected by the controller 100 based on the datapath event.

The physical topology acquisition unit 430 recognizes the physical topology based on the LLDP packet.

The bandwidth measuring unit 440 obtains band information between the nodes 10 by using port states (port_stats).

The L2 switching processor 450 performs a conventional L2 switching role.

The routing selector 470 sets a routing path based on the QoS profile declared for the packets of three or more layers.

The forwarding processor 460 generates and transmits a flow entry to the open flow switch 20 included in the path according to the path set by the routing selector 470.

The routing processor 300 includes a QoS routing topology generator 310, a general routing topology generator 320, and a shortest path calculation processor 330.

The QoS routing topology generator 310 processes a flow that reflects a user's network requirements.

The general routing topology generation unit 320 processes general flows, not QoS routing. The general routing topology generator 320 may process a general flow by using a Dijkstra algorithm for searching the shortest distance.

The shortest path calculation processor 330 searches for the shortest path based on the generated topology, and requests the network configuration information database 220 to reflect the searched shortest path.

3 is a schema structure diagram of a database 220 of an openflow network system according to an embodiment of the present invention.

In FIG. 3, (a) shows Dijkstra Path information, (b) shows host information, (c) shows open flow switch connection information, (d) shows open flow switch list information, and (e) shows QoS policy information. (f) shows QoS path information and (g) shows failed packet information.

Dijkstra Path information (a) is used for storing path information of a general flow that does not require QoS guarantee. SM and DM are the MAC addresses of the transmitting and receiving sides, SP and DP are the IP addresses of the transmitting and receiving sides, respectively, Sp and Dp are the service port numbers of the transmitting and receiving sides, DLT is the type at the data link layer, and NWP is the network protocol number. . The PTH sequentially includes the IDs of the switches 20 that the flow must go through. USE indicates whether the route is used or not, and EXP records the last time the route was used.

The host information (b) is for storing information of hosts subscribed to the open flow network. MAC is the MAC address of the connected host, IP is the IP address, COF is the ID of the OpenFlow switch 20 to which the host is connected, CPT is the port number of the connected switch 20, RBW is the receiving bandwidth, TBW is the sending bandwidth , CRT indicates when the host is connected, and USE indicates whether to use it for routing.

The open flow switch connection information (c) indicates the state information in which the open flow switch is connected to another open flow switch, and has one table for each open flow switch 20.

In OpenFlow switch connection information (c), PT is the port number to which the switch is attached to another switch, COF is the ID of the other Openflow switch to which the switch is connected, RBW is the receiving bandwidth, TBW is the sending bandwidth, and CRT is the last. This is to check the connection status periodically by the time when LLDP packet is received. CST is the cost of the link. The default value is 1.

The open flow switch list information d stores the ID of the switch 20 subscribed to the open flow network.

The QoS policy information (e) is used for storing QoS application related information received from the open flow network policy manager 110 and supports wildcard (*). SM and DM are MAC addresses of the transmitting and receiving sides, SP and DP are IP addresses of the transmitting and receiving sides, respectively, Sp and Dp are service port numbers of the transmitting and receiving sides, DLT is the type at the data link layer, NWP is the network protocol number, and PRT. Is priority, and TYP is a routing protocol option used when establishing a QoS flow path. In the present invention, CSPF is used. The CT mainly uses bandwidth information as a threshold value to satisfy the constraint. The DRT indicates whether the QoS flow is unidirectional or bidirectional, and the STE indicates the state of use of the entry.

QoS path information (f) is used for storing information on the path of the QoS flow, and is substantially similar to the schema of Dijkstra Path information (a). SM and DM are MAC addresses of the transmitting and receiving sides, SP and DP are IP addresses of the transmitting and receiving sides, respectively, Sp and Dp are service port numbers of the transmitting and receiving sides, DLT is the type at the data link layer, NWP is the network protocol number, and PRT. Is priority, and TYP is a routing protocol option used when establishing a QoS flow path. In the present invention, CSPF is used. The CT mainly uses bandwidth information as a threshold value to satisfy the constraint.

The PTH sequentially includes the IDs of the switches 20 that the flow must go through. The DRT records whether the QoS flow is unidirectional or bidirectional, the USE is used to set the path, and the EXP records the last time the path was used. The ABW indicates the available bandwidth of this path, and the CPT indicates whether the links between the hops constituting the path are measured in list form. CBW is used to temporarily store the smallest available bandwidth among several links, and CAY indicates whether the path was invaded by another path.

The failed packet information g is used for storing information about the failed packet when the path establishment fails. The failed packet information (g) is similar to the structure of the QoS path information (f) schema, and records the time when the path establishment fails with a CRT value.

4 is a control flowchart of the openflow network system according to the first embodiment of the present invention, and shows a processing step when a message or event occurs from the openflow switch 20 managed in the present system.

When an open flow message or event occurs, it corresponds to the type of event or message (S410).

This correspondence is performed by calling the open flow message and event processing unit 410 (S412), and the open flow message and event processing unit 410 determines various types of messages and types defined in the OpenFlow protocol (S414). Accordingly, the event and the message, in the present system is a datapath event (signal-in) message to the controller 100 from the switch 20 informing the join / leave of the switch 20, the switch ( A Stats-reply message for Stats-request for management information request of 20) is defined, and the OpenFlow message and event processing unit 410 determines the type of generated message or event.

The occurrence of a datapath event means that the open flow switch 20 joins / leaves the area of the controller 100. Thus, the host and switch registration unit 420 is called (S420).

The host and switch registration unit 420 processes the generated event to join or leave the open flow switch 20 to the area of the controller 100 (S422). Here, the host and switch registration unit 420 adds / deletes an entry to the open flow switch list database 220, generates an open flow switch connection information table for the corresponding switch 20, and creates and adds an entry for each port. / Delete

In the case of a packet-in message, various types of messages are encapsulated and delivered to the controller 100. The packet-in message is divided into two layers or less and a packet including three or more layers of information (S430). S450).

If the packet is three or more layers, the routing selector 470 is called and processed (S450).

If the packet is less than two layers, it is determined whether the packet is an LLDP packet (S430).

If it is not an LLDP packet and contains host information (S432), the host and switch registration unit 420 is called (S420). The host and switch registration unit 420 adds an entry to the open flow switch list database 220, generates an open flow switch connection information table for the corresponding switch 20, and creates and adds an entry for each port.

If it is not an LLDP packet and does not include host information, packets of unknown destination address such as ARP, and packets such as ICMP operate in the same manner as switches used in general use. Thus, the general L2 switching processing unit 450 is called (S434), and processed in a packet processing scheme such as ARP / ICMP (S436).

When receiving the LLDP packet, the physical topology acquisition unit 430 is called (S440) to obtain connection information between the switches (S442).

The chassid and portid of the LLDP packet represent the information of the switch that received the first LLDP packet, and the dpid and inport represent the information of the switch that received the LLDP packet from the original switch. Reload the poem and use it. In addition, in order to obtain the host information, the host detects how many ports are connected to which switch, stores the host information table, and updates the database 220 when the change occurs (S444).

Stats-reply messages are received periodically to collect switch-specific statistics. The bandwidth between links is calculated using statistical information for each port included in the message reception. The bandwidth calculation method calculates the difference between the cumulative number of bytes and the time reported previously, and divides the byte difference by the time difference. This value is stored in RBW and TBW of OpenFlow switch connection information table and host information table, and this value is used bandwidth of each link.

To calculate the available bandwidth between hops on the QoS path, the maximum supported bandwidth is subtracted from the used bandwidth. However, since this value is the value of one of the links on the path, it is not the available bandwidth of the path. This is because the available bandwidth of the path should be the minimum available bandwidth of the related links. Therefore, after checking all available bandwidths of each link, the minimum available bandwidth must be derived.

The CBW of the QoS route information table temporarily stores the minimum value of the available bandwidths of the checked links. The CPT represents the available bandwidth check state of each link included in the path. If there are five links included in the path, '11111' should be checked. '1' is checked according to the link order. If the first and third of the five links are checked, it is '10100'. If the available bandwidth of all included links is checked (ie all 1s) then the CBW is stored as an ABW.

5 is a control flowchart of an openflow network system according to a second embodiment of the present invention, and illustrates a flow processing method in the routing processor 300.

 If a packet-in message including information of three layers or more is received, or if rerouting is required (S512), the routing selector 470 is called (S512).

It is determined whether the flow of the received packet is a general flow or a QoS flow (S514).

In the case of the QoS flow, the CSPF QoS routing topology generator 310 is called (S520), and the link that does not satisfy the constraint CT is removed from the openflow switch connection information table (S522).

Thereafter, a new topology is generated and a path is generated by setting the shortest path (S524).

Here, it is checked whether the path setting is successful (S526). If the path setting is successful, the corresponding path is stored in the PTH field of the QoS path information table in the form of a list (S528). Accordingly, forwarding is required for each switch 20 on the path by forwarding using a packet-out message of open flow (S540).

On the other hand, if the path setting fails, the relevant information is stored in the fail packet information database 220 (S538).

On the other hand, if it is determined that the flow of the received packet is a general flow, the general routing topology generation unit 320 using the Dijkstra Path technology is called (S530).

The general routing topology generator 320 simply generates the topology and then generates the shortest path (S532).

If the shortest path is successfully set, the path is stored in the PTH field of the Dijkstra Path information database 220 in the form of a list (S536).

On the other hand, if the path setting fails, the relevant information is stored in the fail packet information database 220 (S538).

As described above, the present invention manages and controls user traffic by dynamically assigning traffic transmission paths according to user-defined quality of service requirements and moving traffic through redefinition of flows according to network conditions. Can be. In addition, there is a feature that can further define and combine attributes constituting the flow, for example, MAC address, IP address, service port number, and the like.

As a result, network policy managers can easily support network applications that require more than a certain level of network quality, and can use the idle bandwidth efficiently through dynamic flow movement, which can increase the efficiency of network utilization. have.

As such, those skilled in the art will appreciate that the present invention can be implemented in other specific forms without changing the technical spirit or essential features thereof. Therefore, the above-described embodiments are to be understood as illustrative in all respects and not as restrictive. The scope of the present invention is defined by the appended claims rather than the detailed description and all changes or modifications derived from the meaning and scope of the claims and their equivalents are to be construed as being included within the scope of the present invention do.

The present invention can maintain the network quality more than a certain level by dynamically allocating traffic transmission paths according to user-defined service quality requirements and redefining flows according to network conditions to move traffic, and moving idle bandwidth through dynamic flow movement. It can be used for open flow network operation that can be used efficiently.

10: node 20: openflow switch
100: flow routing controller 110: manager
200: QoS and configuration information manager 210: QoS application target input unit
220: database 230: route information processing unit
240: host / switch information processing unit 300: routing processor
310: QoS routing topology generation unit
320: general routing topology generation unit
330: shortest path calculation processing unit
400: Forwarding and Switch Communicator
410: overflow message and event processing unit
420: host and switch register 430: physical topology acquisition
440: bandwidth measurement unit 450: L2 switching processing unit
460: forwarding processing unit 470: routing selection unit

Claims (15)

An open flow network system including a plurality of nodes for transmitting and receiving data, an open flow switch for relaying flow transmission and reception between the nodes, and a flow routing controller for controlling the open flow switch,
The controller generates forwarding for the flow by reflecting the flow information received from the openflow switch and the user's request, and sends the forwarding information for the flow to the openflow switch in the form of a flow entry;
The openflow switch receives forwarding for the flow from the controller in the form of a flow entry to manage subsequent traffic based on forwarding rules included in the entry;
The open flow switch and the controller is an open flow network system, characterized in that for transmitting and receiving data to each other in a TCP or SSL manner.
delete The method of claim 1,
The flow entry,
A header field containing a MAC, IP, service port number of a destination and an origin of the flow;
An open PRO network system comprising an action field for performing forwarding to a designated port.
An open flow network system comprising a plurality of nodes for transmitting and receiving data, and an open flow switch for relaying flow transmission and reception between the nodes,
A QoS and configuration information manager including a database storing a QoS policy received from an administrator and network configuration information maintaining a network connection state;
Periodically monitors and registers and manages information of the OpenFlow switch and the host, and sets a routing path based on the QoS profile declared for the received packet according to the received event and message type and includes the routing path in the set path. A forwarding and switch communicator for generating a flow entry of an open flow switch; And
A routing processor for generating any one of the QoS routing topology and the general routing topology according to the received flow, calculating a shortest path based on each generated topology, and requesting the network configuration information database to be reflected;
The QoS and configuration information manager may include: a QoS application object input unit which generates a table of the database, stores, manages, and updates the QoS related information and network information in the table; A host / switch information processor for registering and managing the host and the open flow switch information; And a route information processor for storing and managing route information of the flow.
delete The method of claim 4, wherein
The forwarding and switch communicator,
An open flow message and event processor;
A host and switch register configured to register an openflow switch and a host detected by the controller based on the data path event;
A physical topology acquiring unit recognizing a physical topology based on the LLDP packet;
A bandwidth measuring unit for obtaining a band between nodes using port state information;
An L2 switching processor configured to perform switching for transmission and reception of a general flow;
A routing selector for making a routing decision based on the QoS profile declared for the received packet;
An open flow network system comprising a forwarding processing unit for generating and transmitting a flow entry to an open flow switch included according to a set path.
The method of claim 4, wherein
The routing processor,
QoS routing topology generation unit for processing a flow reflecting a user's network requirements;
A general routing topology generation unit in charge of general flow processing;
And a shortest path calculation processor configured to calculate the shortest path based on each of the topologies.
The method of claim 4, wherein
In the database,
Open PRO network, characterized in that at least one or more of the Dijkstra Path information, host information, open flow switch connection information, open flow switch list information, QoS policy information, QoS path information and failed packet information is stored system.
The method of claim 8,
The shortest path (Dijkstra Path) information, QoS policy information, QoS path information and failed packet information,
Open including the MAC address of the transmitting and receiving side, the IP address of the transmitting and receiving side, the service port number of the transmitting and receiving side, the type and network protocol number in the data link layer, and further comprises additional information according to the attribute of each information. Prow network system.
The method of claim 8,
The host information is,
The open flow network system, comprising: address information and access information of a host subscribed to the open flow network, bandwidth information of a receiving side and a transmitting side, and information on whether a host is connected and whether to use it when setting a path.
The method of claim 8,
The open flow switch connection information,
And the address information and access information of the open flow switch, the reception side and the transmission side bandwidth information, and time information of the last received packet.
In the control method of the open PRO network system,
Determining whether the flow of the received packet is a general flow or a QoS flow;
Creating a QoS routing topology if the QoS flow;
Setting a shortest path based on each generated topology;
Generating a flow entry based on the shortest path;
Sending the flow entry to an openflow switch;
Managing subsequent traffic at the openflow switch based on a forwarding rule included in the flow entry; And
And adding or deleting connection information of a corresponding host and an open flow switch when a data path event occurs.
The method of claim 12,
Generating a Dijkstra routing topology when the received flow is a general flow;
Setting a shortest path based on each generated topology; And
And controlling the L2 switch according to the path to transmit the traffic.
The method of claim 12,
Periodically receiving port-specific statistical data of the open flow switch;
Calculating bandwidth between links using the port-specific statistical information;
Setting a routing flag according to the bandwidth;
The method of claim 1, further comprising updating previously stored host information and connection information of the open flow switch.
delete

Images