How to operate the OpenFlow standard protocol in the router

OpenFlow is an exchange technology that uses the OpenFlow protocol to establish a Software Defined network. It can manage the network as a whole rather than countless independent devices. Currently, many vendors have begun to consider using the OpenFlow protocol.

1. Traditionally, the Spanning Tree Protocol or some other new standards (such as multi-link transparent interconnection, TRILL) are used to determine the packet forwarding path. OpenFlow transfers the forwarding decision from each vro to the Controller, which is generally a server or workstation. It manages the application execution controller, interacts with all network routers, and configures the data forwarding path, this improves the bandwidth utilization. This application interacts with the cloud management software to ensure sufficient bandwidth to support load creation and change.

2. The OpenFlow standard defines the interaction protocol between the Controller and the router, and a group of router operations. The Controller-Router Protocol runs on the secure transport layer protocol (TLS) or unprotected TCP connection. The controller sends commands to the router, controls the packet forwarding mode, and configures parameters, such as VLAN priority. The router will send a message notification controller when the link is interrupted or when packets with no forwarding instructions are specified.

3. the forwarding command is based on the stream. This stream is composed of the common features of all data packet sharing. Defining a stream requires specifying many parameters, which may include: the router port to which the data packet arrives, the source Ethernet port, the Source IP port, the VLAN label, the target Ethernet or IP port, and many other packet characteristics. The Controller sets a set of parameters on the vro to define each stream and the packet processing method that matches the stream.

4. Each vro maintains many route tables, and each table contains multiple route records. Each route record contains a matching domain, which defines the stream, counter, and instruction set. Items recorded in the matching domain may contain a comparison parameter value that corresponds to the arriving data packet, or a parameter set that indicates the Stream does not have the indication value for this record.

5. The number of the route table starts from 0, and the packets that arrive at the table are compared with the records of Table 0. If yes, the route count increases and the specified instruction set is executed. If the incoming data packet does not match any route table record, a new stream must be created.

Some routers may directly discard undefined streams, but in most cases, data packets are forwarded to the Controller. The Controller then defines a new stream for the packet and creates one or more route table records. It then sends records to the vro and adds the route table. Finally, the data packet is sent back to the router and processed using the newly created route record.

6. The route table command modifies the operation set for each data packet. At the beginning, data packets are processed using an empty operation set. These operations may require data packets to be forwarded through the specified port, or you need to modify the packet TTL, VLAN, MPLS label, or packet QoS.

7. The command in the first route table may perform operations on data packets or add some future operations. These commands compare data packets with other route table records to control the subsequent processing of data packets. Subsequent commands recorded in the route table may add operations, delete or modify previously added operations, or perform other operations.

8. Before sending a data packet to the next route table, the command may also add an element data value to the data packet. This value becomes an additional parameter, which matches the metadata value of the record in the next route table. Data packets are processed one by one until all commands are processed and the data packets are forwarded successfully.

The command may specify a group identifier. A group is an efficient way to specify the same operation set for data packets in multiple streams. Group Operation records are defined in the vro group table. Each record includes the Identifier value, group type, counter, and operation set.

The group type specifies whether all operation sets are executed. This is suitable for broadcast or multicasting, or it specifies that only some specific operation sets are executed.