KR100687745B1 - IPv6 Sender-based Network Processor for Multicast Packet Forwarding and Its Method - Google Patents

Abstract

A network processor and method for IPv6 sender based multicast packet forwarding are disclosed. The ingress forwarder of the network processor's forwarding engine unit determines whether the packet is an IPv6 sender-based multicast packet based on the packet's destination address. If the packet is an IPv6 sender-based multicast address, the ingress forwarder uses the packet's source address and destination address as a hash key. Retrieve the multicast forwarding table. The egress forwarder searches the IPv6 multicast output port table to determine the number of output ports corresponding to the packet, and copies and forwards the packet by the number of the identified ports. As a result, the network processor can divide the IPv6 sender-based multicast packet and forward the packet at high speed.

Network Processor, IPv6 Sender Based Multicast

Description

IP processor 6 Network processor for sender-based multicast packet forwarding and method therefor {Network processor for IPv6 source-specific multicast packet forwarding and method therefor}

1 is a diagram showing the configuration of an embodiment of a packet forwarding engine unit of a network processor having an IPv6 sender based multicast packet forwarding function according to the present invention;

2 is a diagram illustrating an embodiment of a configuration of an egress forwarder of a network processor according to the present invention;

3 is a diagram illustrating an embodiment of a configuration of an ingress forwarder of a network processor according to the present invention; and

4A and 4B illustrate a method of forwarding an IPv6 sender based multicast packet in a network processor according to the present invention.

The present invention relates to a network processor, and more particularly, to a network processor and a method for forwarding an IPv6 sender based multicast packet.

Currently, most IP multicast applications are based on the Any-Source Multicast model, which supports many-to-many transmission. However, the current random sender-based multicast model is known to have problems with scalability and security. Currently, the Internet Engineering Task Force (IETF), a private international standards organization, has completed the standardization by proposing an IPv6 sender-based multicast method to solve the shortcomings of the random sender-based multicast model. An empirical method for transport based on IPv6 sender based multicast packets has not been developed.

Conventional application-specific integrated circuit (ASIC) -based systems provide high performance on a hardware basis, but do not quickly accommodate the demands for various Internet services. The network processor has a rapid development cycle and provides sufficient performance, providing a development environment that can flexibly cope with rapidly changing markets such as the IPv6 market.

An object of the present invention is to provide a network processor and a method for classifying IPv6 sender-based multicast packets and forwarding them at high speed.

In order to achieve the above technical problem, an embodiment of a network processor for forwarding an IPv6 sender based multicast packet may be configured to determine whether the packet is an IPv6 sender based multicast packet based on a destination address of the packet, and wherein the packet is an IPv6 An ingress forwarder for searching a multicast forwarding table using a source address and a destination address of the packet as a hash key if the sender is a multicast address; And an egress forwarder for searching an IPv6 multicast output port table to determine the number of output ports corresponding to the packet, and copying and transmitting the packet as many as the identified port number.

In order to achieve the above technical problem, an embodiment of a method of forwarding an IPv6 sender based multicast packet in a network processor includes: (a) determining whether the packet is an IPv6 sender based multicast packet based on a destination address of the packet; step; (b) if the packet is an IPv6 sender based multicast address, searching for a multicast forwarding table using the source address and destination address of the packet as a hash key; (c) if the search of the forwarding table is successful, searching the IPv6 multicast output port table to determine the number of output ports corresponding to the packet; And (d) copying and transmitting the packet by the number of the identified ports.

As a result, the network processor can divide the IPv6 sender-based multicast packet and forward the packet at high speed.

In the following, with reference to the accompanying drawings will be described in detail with respect to a network processor having a IPv6 sender-based multicast packet transmission function and a method according to the present invention.

1 is a diagram illustrating a configuration of an embodiment of a packet forwarding engine unit of a network processor having an IPv6 sender based multicast packet transmission function according to the present invention.

Referring to FIG. 1, a network processor according to the present invention includes an ingress forwarder 100 and an egress forwarder 150.

The ingress forwarder 100 determines whether the packet is an IPv6 sender-based multicast packet based on the destination address of the packet. If the packet is an IPv6 sender-based multicast address, the ingress forwarder 100 uses the source address and the destination address of the packet as a hash key. After searching the cast forwarding table and generating packet metadata reflecting the entry of the found forwarding table, the packet is forwarded to the egress forwarder 150.

The egress forwarder 150 searches the IPv6 multicast output port table to retrieve the number of output ports corresponding to the packet, and copies the packet by the number of the detected ports to forward the packet.

Specifically, the ingress forwarder 100 includes a packet classifier 102, an ingress multicast forwarder 104, a CSIX frame generator 108, and an exception packet processor 106.

Upon receiving a packet from the network, the packet classification unit 102 determines whether the packet is an IPv6 packet and discards the packet if it is not an IPv6 packet. Based on the destination address of the received packet, it is determined whether the packet is an IPv6 multicast packet or an IPv6 unicast packet. In addition, if the packet is an IPv6 unicast packet, the packet classification unit 102 transmits the packet to an IPv6 unicast forwarding module (not shown). If the packet is an IPv6 multicast packet, the packet classifier 102 determines whether the packet is an IPv6 transmitter-based multicast packet. The packet classifier 102 forwards the packet to the ingress multicast forwarder 104 if the packet is an IPv6 sender based multicast packet.

The ingress multicast forwarder 104 searches the multicast forwarding table using the source address and destination address of the received IPv6 sender-based multicast packet as a hash key, generates packet metadata that reflects the entries of the found forwarding table. The packet is transmitted to the CSIX frame generation unit 108.

The CSIX frame generation unit 108 generates an IPv6 sender-based multicast packet received from the ingress multicast forwarder 104 as a CSIX frame and delivers it to the egress forwarder 150.

The exception packet processing unit 106 receives a packet having a predetermined IPv6 link local multicast address from the ingress multicast forwarder 104 and performs exception processing on the received packet.

Next, the egress forwarder 150 will be described in detail. The egress forwarder 150 looks at the egress multicast forwarder 152, the packet replication unit 154, and the packet transmission unit 156. Include.

The egress multicast forwarder 152 searches the multicast output port table to determine the number of output ports of the packets received from the ingress multicast forwarder 104, and the packet replication unit 154 determines the number of output ports. The packet is duplicated and stored in a buffer. The packet transmitter 156 searches the multicast output port table for each duplicated packet, sets a corresponding next hop to each packet, and outputs the packet.

2 is a view showing an embodiment of the configuration of the egress forwarder of the network processor according to the present invention, Figure 3 is a view showing an embodiment of the configuration of the ingress forwarder of the network processor according to the present invention.

2 and 3, an egress forwarder and an ingress forwarder of a network processor are largely divided into a micro block module and a core module.

The micro block module checks the delivery information of the corresponding multicast packet based on the multicast tree information received from the control plane and, if necessary, sends a packet copy command to the packet copier 310 to copy the packet. An ingress multicast forwarder 215 for forwarding these duplicated packets to a corresponding member, an egress multicast forwarder 305 operating in an egress forwarder in response to the ingress multicast forwarder 215, And a packet copy unit 310 for copying as many packets as required for the packet copy command received from the ingress multicast forwarder 215.

The core module operates on the XScale core of the ingress forwarder and initializes resources for the ingress multicast forwarder 215 and handles exception packets from the microblock module 220, egress core component 330. ), And the like.

Hereinafter, each configuration will be described in detail.

2, the ingress forwarder includes an Ether Rx module 200, a decap / classify 205, an IPv6 forwarder 210, an ingress multicast forwarder 215, and an ingress core component 220. , Stack driver core component 225, queue manager 235, and CSIX TX module 240.

Modules indicated by squares in FIG. 2 are newly implemented in the present invention, and modules indicated by rhombuses extend modules provided by the conventional network processor forwarding engine unit, and modules indicated by ovals indicate the conventional network processor forwarding engine unit. This module is provided by. And the external packet processing processor 230 provided in a hexagon is a separate exception packet processing system.

The Ether Rx module 200 receives a packet of a data link layer through a network interface device. The classification unit (Decap / Classify) 205 determines whether the packet is an IPv6 packet by using a type value of a link layer packet received by the Ether Rx module 200. The classifier 205 discards the received packet if it is not an IPv6 packet. In general, for IPv6 packets, the type value is 0x86DD.

The IPv6 forwarder 210 performs a function of checking the validity of the header of the IPv6 packet received from the classifier 205, searching the IPv6 unicast forwarding table, and reducing the hop limit value of the IPv6 unicast packet.

In detail, the IPv6 forwarder 210 discards a packet having an error of a version or length of a packet, a multicast address, or a packet whose source address is 0 or :: 1 in validating an IPv6 packet header. In addition, the IPv6 forwarder 210 ingresses the packet if the destination address of the packet is an IPv6 link local address, if the packet includes a Hop-by-Hop option header, or if the IPv6 unicast forwarding table lookup fails. Pass to component 220. The IPv6 forwarder 210 also forwards the packet if the destination address of the packet is an IPv6 sender-based multicast address (i.e., FF3x :: / 96, where x is a value in the range 0 to F in hexadecimal). Forward to 215.

The ingress multicast forwarder 215 checks whether the destination address of the received packet is an IPv6 sender based multicast address. If the destination address is an IPv6 sender based multicast address, the ingress multicast forwarder 215 uses the source address and destination address of the packet as a hash key to search the IPv6 multicast forwarding table.

Upon successful retrieval of the multicast forwarding table, the ingress multicast forwarder 215 sets the header type of packet metadata generated for each packet to IPv6_MULTICAST for packet processing in the forwarding engine, and sets next_hop_id to the retrieved multicast forwarding table. It is set to the index (Multicast ID, MTID) of the entry of, and the hop limit of the packet is reduced, and then the packet is delivered to the queue manager 235.

The ingress multicast forwarder 215 discards the packet if the search of the multicast forwarding table fails. The ingress multicast forwarder 215 also Xscale packets when the destination address of the packet is one of the FF02 :: 1, FF02 :: 2, FF02 :: D, and FF02 :: 16 of the IPv6 link-local multicast addresses. It passes to the ingress core component 220 of the core module.

The ingress core component 220 creates and manages an IPv6 unicast forwarding table and an IPv6 multicast forwarding table for the transmission of IPv6 packets. In addition, the ingress core component 220 examines the packet received from the ingress multicast forwarder 215, and if the packet has an IPv6 Hop-by-Hop option header, and the type of the packet is an IPv6 multicast packet. If the destination address of the packet is a link-local multicast address (if any of FF02 :: 1, FF02 :: 2, FF02 :: D, or FF02 :: 16), then the packet is sent to the stack driver core component 225. To pass.

Further, the ingress core component 220 may check the Internet Control Message Protocol for IPv6 when the IPv6 Maximum Transmission Unit (MTU) check of the packet fails, a hop limit error occurs, or the search of the IPv6 unicast forwarding table fails. ) Send a message. The ingress core component 220 also drops the packet if the lookup of the IPv6 multicast forwarding table fails and the ICMPv6 redirect option is set.

The stack driver core component 225 transfers the packet to the external packet processing processor 230 in charge of processing the packet received from the ingress core component 220 and the packet processed by the external packet processing processor 230. It is responsible for the function of delivering to the queue manager 235.

The CSIX TX module examines the packet received from the queue manager 235 and generates a CSIX multicast frame if the packet is an IPv6 sender-based multicast packet. In the CSIX multicast frame, the value of the type field of the base header is set to the multicast ID (0x03), and the CSIX multicast extension header is added. The generated CSIX multicast frame is delivered to the egress forwarder of the packet forwarding engine unit.

3, the egress forwarder according to the present invention includes a CSIX RX module 300, an egress multicast forwarder 305, a packet replication unit 310, an Ether Encap module 315, a queue management unit 320, It consists of an Ether TX module 325 and an egress core component 330.

If the packet received from the ingress forwarder of the packet forwarding engine unit is a multicast packet, the CSIX RX module 300 sets the header type to multicast and delivers the packet to the egress multicast forwarder 305. The CSIX RX module 300 forwards the packet directly to the Ether Encap module 315 if the packet is an IPv6 unicast packet.

The egress multicast forwarder 305 searches the IPv6 multicast output port table (MOIT_Port_Table) to forward packets to the Ether Encap module 315 when the number of output ports is one or less, and when the number of output ports is two or more. Next, the multicast output interface table (MOIT_OIF_Table) is searched for, the MTU value and the hop limit value are checked, and the packet is delivered to the packet replicating unit 310.

The packet copier 310 generates a child buffer for each port designated as an output port of the current packet, and then copies and stores the multicast packets to be output in the buffer. The packet replication unit 310 transmits the duplicated packets to the Ether Encap module 315.

The Ether Encap module 315 retrieves the multicast output port table (MOIT_Port_Table) using the index of the forwarding table entry and the output interface number of the multicast output interface table to obtain the output port number of the network processor that will transmit the actual packet. The packet whose output port number is retrieved from the Ether Encap module 315 is transferred to the Ether Tx module 325 via the queue manager 320.

The Ether Tx module 325 searches the link layer table, adds the link layer header to the packet, and then transmits the packet. After that, the packet replicating unit 310 empties the child buffer.

4A and 4B illustrate a method of forwarding an IPv6 sender based multicast packet in a network processor according to the present invention.

2 and 4A, when the Ether Rx module 200 receives a packet through the network interface device of the ingress forwarder (S400), the classifier 205 determines whether it is an IPv6 packet using a type value of a link layer packet. Check (S405).

If it is an IPv6 packet, the classification unit 205 forwards the packet to the IPv6 forwarder 210, and the IPv6 forwarder 210 performs a validity check on the header of the received IPv6 packet (S410), if there is an error in a version or a header checksum. Discard the packet (S440).

The IPv6 forwarder 210 checks the destination address of the packet and if the destination address is an IPv6 multicast address, delivers the packet to the ingress multicast forwarder 215 (S415). The packet whose destination address is the IPv6 unicast region is delivered to the IPv6 unicast forwarding module (S445).

The ingress multicast forwarder 215 checks whether the destination address of the received packet is an IPv6 sender-based multicast address (S420). In the case of an IPv6 sender-based multicast address, the ingress multicast forwarder 215 uses the source address and the destination address of the packet as a hash key. The IPv6 multicast forwarding table is searched (S425).

If the multicast forwarding table search succeeds (S430), the CSIX TX module 240 generates a CSIX multicast frame for forwarding the packet to the egress forwarder, and then delivers it to the CSIX RX module 300 of the egress forwarder ( S435).

3 and 4B, the CSIX RX module 300 in the egress forwarder forwards the packet received from the ingress forwarder to the ingress multicast forwarder 305 (S450). The egress multicast forwarder 305 searches the IPv6 multicast output port table (S455).

If the number of the corresponding output ports is one or less (S460), the egress multicast forwarder 305 stores the packet in the parent buffer and delivers the packet to the Ether Encap module 315 (S470). If the number of output ports is two or more, the egress multicast forwarder 305 delivers the packet to the packet replicating unit 310, and the packet replication unit 310 creates a child buffer as many as the number of output ports, and then sets the parent. The packet stored in the buffer is replicated and stored in each child buffer, and then transferred to the Ether Encap module 315 (S465, S470).

The Ether Encap module 315 retrieves the next hop from the multicast output port table, sets the packet to the packet (S475), and then adds the link layer header to repeatedly transmit the packet as many as the number of child buffers (S480 and S485).

The invention can also be embodied as computer readable code on a computer readable recording medium. Computer-readable recording media include all types of recording devices that store data that can be read by a computer system. Examples of computer-readable recording media include ROM, RAM, CD-ROM, magnetic tape, floppy disk, optical data storage, and the like, and may also be implemented in the form of a carrier wave (for example, transmission over the Internet). Include. The computer readable recording medium can also be distributed over network coupled computer systems so that the computer readable code is stored and executed in a distributed fashion.

So far I looked at the center of the preferred embodiment for the present invention. Those skilled in the art will appreciate that the present invention can be implemented in a modified form without departing from the essential features of the present invention. Therefore, the disclosed embodiments should be considered in descriptive sense only and not for purposes of limitation. The scope of the present invention is shown in the claims rather than the foregoing description, and all differences within the scope will be construed as being included in the present invention.

According to the present invention, an IPv6 sender based multicast service can be provided using a network processor in an IPv6 network. Accordingly, IPv6 sender-based multicast service can be provided to a large number of users of IPv6 network at high speed, and the present invention can be applied to applications such as Internet broadcasting to efficiently use network resources.

Claims (11)

Determining whether the packet is an IPv6 sender based multicast packet based on a destination address of the packet, and searching the multicast forwarding table using the source address and the destination address of the packet as a hash key if the packet is an IPv6 sender based multicast address. Ingress forwarder; And And an egress forwarder for retrieving an IPv6 multicast output port table to determine the number of output ports corresponding to the packet, and to copy and transmit the packet by the identified number of ports. The method of claim 1, wherein the ingress forwarder, A packet classifier configured to determine whether the packet is an IPv6 multicast packet or an IPv6 unicast packet based on a destination address of the packet, and whether the packet is an IPv6 sender based multicast packet if the packet is an IPv6 multicast packet; If the packet is an IPv6 sender-based multicast packet, an ingress multicast that searches the multicast forwarding table using the source address and the destination address of the packet as a hash key and generates packet metadata reflecting an entry of the found forwarding table. A forwarder; comprising a network processor. The method of claim 1, wherein the ingress forwarder, And determining whether the packet is an IPv6 packet based on a type value of a data link layer of a packet received from a network. The method of claim 2, wherein the ingress forwarder, And a CSIX frame generator configured to generate an IPv6 sender-based multicast packet received from the ingress multicast forwarder as a CSIX frame and deliver the CSIX frame to the egress forwarder. The method of claim 2, wherein the ingress forwarder, And an exception packet processor for receiving a packet having a predetermined IPv6 link local multicast address from the ingress multicast forwarder and performing exception processing on the received packet. The method of claim 1, wherein the egress forwarder, An egress multicast forwarder for retrieving a multicast output port table to determine the number of output ports for packets received from the ingress multicast forwarder; And And a packet copying unit for replicating the packets as many as the determined number of output ports and storing the packets in a buffer. The method of claim 6, wherein the egress forwarder, And a packet transmitter configured to search the multicast output port table for each of the duplicated packets, and to set and transmit a corresponding next hop to each packet. (a) determining whether the packet is an IPv6 sender based multicast packet based on a packet's destination address; (b) if the packet is an IPv6 sender based multicast address, searching for a multicast forwarding table using the source address and destination address of the packet as a hash key; (c) if the search of the forwarding table is successful, searching the IPv6 multicast output port table to determine the number of output ports corresponding to the packet; And (d) copying and transmitting the packet by the number of the identified ports; IPv6 sender based multicast packet transmission method in a network processor, comprising: a. The method of claim 8, wherein step (a) comprises: (a1) determining whether the packet is an IPv6 packet based on a link layer type value of the packet, and discarding the packet if the packet is not an IPv6 packet; (a2) validating the header of the packet and discarding the packet if there is an error in the test result; (a3) determining whether the packet is an IPv6 multicast packet or an IPv6 unicast packet based on a destination address of the packet; And (a4) if the packet is an IPv6 multicast packet, determining whether the packet is an IPv6 sender-based multicast packet; and transmitting the IPv6 sender-based multicast packet in a network processor. According to claim 8, wherein step (b) is, (b1) if the packet is an IPv6 sender based multicast packet, searching for a multicast forwarding table using the source address and the destination address of the packet as hash keys; (b2) generating packet metadata reflecting an entry of the found forwarding table; And (b3) if the destination address of the packet is a predetermined IPv6 local multicast address, forwarding the packet to a core module of a forwarding engine unit and performing packet exception processing; and based on the IPv6 sender in the network processor How to send multicast packets. The method of claim 10, wherein step (d) (d1) if the number of the identified output ports is two or more, replicating the packet by the number of the identified output ports and storing the packet in a buffer; (d2) searching for the next hop corresponding to the packet in the multicast output port table, setting the packet in the packet, and then adding and transmitting a link layer header to transmit the IPv6 sender in the network processor; Based multicast packet transmission method.

Images