High-performance multi-user traffic replication system and method based on combination of switching chip and network processor

Abstract

The invention discloses a high-performance multi-user traffic replication system and method based on combination of a switching chip and network processors. The system comprises the switching chip, a first network processor and a second network processor, original flow firstly enters the first network processor, the first network processor adds a private message header required by the switching chip to a flow message and then transmits the flow to the switching chip, and the switching chip copies and forwards the flow according to three-layer multicast information in the message header. The replicated flow enters a second network processor to be processed, the second network processor replicates the received flow message, then restores all the replicated flow into original message data, carries out unicast forwarding in sequence, and respectively forwards the original message data to different outlets of the switching chip or equipment; each exit corresponds to one different user, and each user obtains one flow. On the basis of original hardware of the shunt switch, the performance of multi-user traffic replication of the whole equipment is improved by utilizing the respective functional characteristics of the switching chip and the network processing chip and combining the digital communication forwarding principle.

Description

Technical Field

[0001] The present invention relates to the field of digital communication technology, and in particular to a high-performance multi-user traffic replication system and method based on a combination of a switching chip and a network processor. Background Art

[0002] With the rapid development and maturity of digital communication technology, communication security issues have become an increasingly important focus of attention, and the application of diversion switches has also increased. Diversion switches often need to provide traffic to different users or multiple users, so the requirements for traffic replication capabilities are getting higher and higher. In specific scenarios, traffic messages need to be copied and distributed to different users. Generally, they are copied directly through network processors, but the replication capabilities of network processors are often limited. Therefore, new technologies or methods are needed to improve the comprehensive replication capabilities of the equipment. The commonly used method is to use a higher-performance network processing chip, but this method will bring additional and higher costs, increase the overall power consumption of the equipment, and the need to replace the equipment as a whole. Summary of the invention

[0003] Purpose of the invention: The present invention provides a high-performance multi-user traffic replication system and method based on a combination of a switching chip and a network processor. Based on the original hardware of the diversion switch, the performance of multi-user traffic replication of the entire device is improved by utilizing the functional characteristics of the switching chip and the network processing chip and combining the principle of digital communication forwarding.

[0004] Technical solution: The present invention discloses a high-performance multi-user traffic replication system based on a combination of a switching chip and a network processor, comprising: a switching chip, a first network processor and a second network processor; the original traffic first enters the first network processor, the first network processor adds a private message header required by the switching chip to the traffic message, and then passes the traffic to the switching chip, the switching chip replicates and forwards the traffic according to the three-layer multicast information in the message header, the replicated traffic enters the second network processor for processing, the second network processor replicates the received traffic message again, and then restores all the replicated traffic to the original message data, and performs unicast forwarding in sequence, respectively forwarding to different exits of the switching chip or device, each exit corresponds to a different user, and each user obtains a portion of the traffic.

[0005] Furthermore, the system has a switching chip and two network processors, and the system traffic path is first entering the first network processor, then entering the switching chip, then entering the second network processor, then entering the switching chip, and finally output to the outside of the system or outside the device.

[0006] Furthermore, the first network processor needs to encapsulate a switching private message header with three-layer multicast information based on the original message, wherein the three-layer multicast information determines the number of copies of the traffic by the switching chip.

[0007] Furthermore, during the initialization of the switching chip, multiple layer 3 multicast groups are established, each multicast group adds a different number of layer 3 vlan interfaces, each vlan interface is bound to only one physical port connected to the second network processor, and different multicast groups replicate different amounts of traffic.

[0008] Furthermore, the second network processor needs to replicate again and forward the traffic to the switch chip in sequence by unicast.

[0009] Furthermore, the external ports of the switching chip can correspond to different users to perform final traffic distribution.

[0010] Furthermore, assuming that the replication capacity of the network processing chip is m, that is, the maximum number of copies allowed within a specific processing time is m, if the switching chip adds a three-layer multicast group and n three-layer interfaces are added to the multicast group, then the maximum replication capacity of the system within the specific processing time is n*m.

[0011] Accordingly, a high-performance multi-user traffic replication method based on a combination of a switching chip and a network processor includes the following steps:

[0012] Step 1, switching chip initialization stage;

[0013] Step 2, the first network processor adds information; the original traffic first enters the first network processor, the first network processor adds the private message header required by the switching chip to the traffic message, and then passes the traffic to the switching chip;

[0014] Step 3, switching chip layer 3 multicast replication stage: the switching chip replicates and forwards the layer 3 multicast information in the message header, and the replicated traffic enters the second network processor for processing;

[0015] Step 4, the second network processor replication and unicast forwarding stage; the second network processor replicates the received traffic message, then restores all the replicated traffic to the original message data, and performs unicast forwarding in sequence, forwarding them to different exits of the switching chip or device respectively. Each exit corresponds to a different user, and each user obtains a portion of the traffic.

[0016] Furthermore, in step 1, multiple multicast groups are created during the initialization phase of the switching chip, and a different number of three-layer VLAN interfaces are added to each multicast group. Each VLAN interface is bound to a physical port connected to the second network processor, ensuring that a three-layer VLAN interface only copies one copy of the traffic, and different multicast groups can copy different amounts of traffic.

[0017] Furthermore, in step 2, the first network processor needs to encapsulate a switching private message header with three-layer multicast information based on the original message, wherein the three-layer multicast information determines the number of copies of the traffic by the switching chip.

[0018] Furthermore, in step 3, the switching chip performs a layer 3 multicast query on the traffic entering from the first processor to determine whether there is a corresponding layer 3 multicast number. If not, the traffic is discarded or other forwarding processes are performed. If so, layer 3 multicast replication is performed and forwarded to the second network processor.

[0019] Furthermore, in step 4, the second network processor performs a certain number of replications again based on the traffic copied by the switching chip according to the number of users to reach the specified number of users, sorts all the traffic and restores the original information, and forwards it to the switching chip via unicast. The switching chip removes the private message header and forwards a copy of the traffic to each user from different ports.

[0020] Beneficial effect: Compared with the prior art, the present invention has the following significant advantages: The present invention is based on the original hardware of the diversion switch, and by utilizing the functional characteristics of the switching chip and the network processing chip, combined with the digital communication forwarding principle, the performance of multi-user traffic replication of the entire device is improved. There is no need to purchase additional hardware or chips, no need to replace equipment, and it can greatly improve the overall traffic replication capability and the number of multiple users of the device. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] Figure 1 It is a schematic diagram of the system structure of the present invention.

[0022] Figure 2 It is a schematic diagram of the method flow of the present invention. DETAILED DESCRIPTION

[0023] like Figure 1As shown, a high-performance multi-user traffic replication system based on a combination of a switching chip and a network processor includes: a switching chip, a first network processor and a second network processor; the original traffic first enters the first network processor, the first network processor adds a private message header required by the switching chip to the traffic message, and then passes the traffic to the switching chip, the switching chip replicates and forwards according to the three-layer multicast information in the message header, the replicated traffic enters the second network processor for processing, the second network processor replicates the received traffic message again, and then restores all the replicated traffic to the original message data, and performs unicast forwarding in sequence, and forwards them to different exits of the switching chip or device respectively, each exit corresponds to a different user, and each user obtains a portion of the traffic.

[0024] With one switching chip and two network processors, the system traffic path is to first enter the first network processor, then the switching chip, then the second network processor, then the switching chip, and finally output to the outside of the system or the outside of the device.

[0025] The first network processor needs to encapsulate a switching private message header with three-layer multicast information based on the original message, where the three-layer multicast information determines the number of copies of the traffic by the switching chip.

[0026] During the initialization of the switching chip, multiple Layer 3 multicast groups are established. A different number of Layer 3 VLAN interfaces are added to each multicast group. Each VLAN interface is bound to only one physical port connected to the second network processor. Different multicast groups have different amounts of replicated traffic.

[0027] The second network processor needs to replicate again and forward the traffic in sequence to the switch chip in unicast.

[0028] The external ports of the switching chip can correspond to different users for final traffic distribution.

[0029] Assume that the replication capability of the network processing chip is m, that is, the maximum number of copies allowed within a specific processing time is m. If the switch chip adds a Layer 3 multicast group and n Layer 3 interfaces are added to the multicast group, then a maximum of n*m copies can be replicated within a specific processing time.

[0030] like Figure 2 As shown, a high-performance multi-user traffic replication method based on a combination of a switching chip and a network processor includes the following steps:

[0031] Step 1: Initialize the switching chip and its basic functions; create multiple Layer 3 multicast groups, namely group1, group2, group3, ..., groupn, where n must be less than or equal to the maximum number of users allowed. Select a VLAN range for the Layer 3 VLAN interface and determine a basic VLAN. Assuming the basic VLAN value is V, it must satisfy 1 <= V <= 4095; add 1 Layer 3 VLAN interface to group1, with a VLAN value of V+1; add 2 Layer 3 VLAN interfaces to group2, namely V+1 and V+2; add 3 Layer 3 VLAN interfaces to group3, namely V+1, V+2, and V+3; similarly, add n Layer 3 VLAN interfaces to groupn, namely V+1, V+2, V+3, ..., V+n, where V+n <= 4095; thus, 1 copy of traffic can be replicated through group1, 2 copies of traffic can be replicated through group2, 3 copies of traffic can be replicated through group3, and similarly, n copies of traffic can be replicated through groupn. Bind each Layer 3 VLAN interface to only one physical port connected to the second network processor, so that each VLAN interface only copies one copy of the traffic and ensures that the copied traffic enters the second network processor.

[0032] Step 2, connect the original traffic to the first network processor; the first network processor adds a private message header used by the switching chip based on the original traffic; assuming that the number of users is N, add three-layer multicast information to the private message header according to the number of users N, that is, the three-layer multicast group groupi, where i<=N; the value of i can be adjusted according to the replication capability of the second network processor, so that the overall replication quantity and replication speed can achieve the best effect.

[0033] Step 3, forward the traffic processed by the first network processor to the switching chip; the switching chip queries the three-layer multicast table according to the three-layer multicast information in the private message header of the traffic, and determines whether there is a corresponding three-layer multicast group groupi; if the switching chip does not have the corresponding three-layer multicast group groupi, the traffic is discarded, or other forwarding processes are performed; if the switching chip has the corresponding three-layer multicast group groupi, three-layer multicast replication and forwarding are performed, and a total of i copies of traffic are copied and forwarded to the network processor 2; the copied traffic messages are respectively equipped with vlan tag labels V+1, V+2, V+3, ..., V+i; this helps the network processor to identify traffic and perform operations such as traffic sorting.

[0034] Step 4: The second network processor selects a portion of the traffic according to the number of users N and replicates it again. If j additional copies are made, i+j=N must be satisfied, and a total of N traffic is generated. The second network processor removes the additional VLAN tag in the N traffic, restores the original information, and modifies the private message header information of the switching chip, changing it to unicast forwarding. The forwarding target of each traffic is N different user ports of the switching chip. Each user port receives a copy of the traffic, and the switching chip removes the private message header of the switching chip and outputs it. Finally, each user port outputs a copy of the traffic that is the same as the original traffic.

[0035] The present invention improves the performance of multi-user traffic replication of the whole device by utilizing the functional characteristics of the switching chip and the network processing chip, combined with the principle of digital communication forwarding. No additional hardware or chips need to be purchased, no equipment needs to be replaced, and the overall traffic replication capability and the number of multiple users of the device can be greatly improved.

Claims

1. A high-performance multi-user traffic replication system based on a combination of a switching chip and a network processor, characterized in that: include: a switch chip, a first network processor, and a second network processor; The original traffic first enters the first network processor. The first network processor adds the private message header required by the switching chip to the traffic message, and then passes the traffic to the switching chip. The switching chip copies and forwards the traffic according to the three-layer multicast information in the message header. The copied traffic enters the second network processor for processing. The second network processor copies the received traffic message again, and then restores all the copied traffic to the original message data, and performs unicast forwarding in turn, forwarding it to different exits of the switching chip or device. Each exit corresponds to a different user, and each user receives a portion of the traffic.

2. The high-performance multi-user traffic replication system based on a combination of a switching chip and a network processor as claimed in claim 1, characterized in that: The first network processor needs to encapsulate a switching private message header with three-layer multicast information based on the original message, where the three-layer multicast information determines the number of copies of the traffic by the switching chip.

3. The high-performance multi-user traffic replication system based on a combination of a switching chip and a network processor as claimed in claim 1, characterized in that: During the initialization of the switching chip, multiple Layer 3 multicast groups are established. A different number of Layer 3 VLAN interfaces are added to each multicast group. Each VLAN interface is bound to only one physical port connected to the second network processor. Different multicast groups have different amounts of replicated traffic.

4. The high-performance multi-user traffic replication system based on a combination of a switching chip and a network processor as claimed in claim 1, characterized in that: The second network processor needs to replicate again and forward the traffic in sequence to the switch chip in unicast.

5. The high-performance multi-user traffic replication system based on a combination of a switching chip and a network processor as claimed in claim 1, characterized in that: Assume that the replication capacity of the network processing chip is m, that is, the maximum number of copies allowed within a specific processing time is m. If the switching chip adds a three-layer multicast group and n three-layer interfaces are added to the multicast group, the maximum replication capacity of the system within the specific processing time is n*m.

6. A method for a high-performance multi-user traffic replication system based on a combination of a switching chip and a network processor as claimed in claim 1, characterized in that: The steps include: Step 1, switching chip initialization stage; Step 2, the first network processor adds information; the original traffic first enters the first network processor, the first network processor adds the private message header required by the switching chip to the traffic message, and then passes the traffic to the switching chip; Step 3, switching chip layer 3 multicast replication stage: the switching chip replicates and forwards the layer 3 multicast information in the traffic message, and the replicated traffic enters the second network processor for processing; Step 4: Second network processor replication and unicast forwarding phase; The second network processor copies the received traffic message, then restores all the copied traffic to the original message data, and forwards it in unicast in sequence, forwarding it to different exits of the switching chip or device respectively. Each exit corresponds to a different user, and each user receives a share of the traffic.

7. The high-performance multi-user traffic replication method based on a combination of a switching chip and a network processor as claimed in claim 6, characterized in that: In step 1, multiple multicast groups are created during the initialization phase of the switching chip. A different number of Layer 3 VLAN interfaces are added to each multicast group. Each VLAN interface is bound to a physical port connected to the second network processor, ensuring that only one Layer 3 VLAN interface copies one copy of the traffic, and different multicast groups copy different amounts of traffic.

8. The high-performance multi-user traffic replication method based on a combination of a switching chip and a network processor as claimed in claim 6, characterized in that: In step 2, the first network processor needs to encapsulate a switching private message header with three-layer multicast information based on the original message, wherein the three-layer multicast information determines the number of copies of the traffic by the switching chip.

9. The high-performance multi-user traffic replication method based on a combination of a switching chip and a network processor as claimed in claim 6, characterized in that: In step 3, the switching chip performs a layer 3 multicast query on the traffic entering from the first network processor to determine whether there is a corresponding layer 3 multicast number. If not, the traffic is discarded or other forwarding processes are performed. If so, layer 3 multicast replication is performed and forwarded to the second network processor.

10. The high-performance multi-user traffic replication method based on a combination of a switching chip and a network processor as claimed in claim 6, characterized in that: In step 4, the second network processor makes a certain number of copies again based on the traffic copied by the switching chip according to the number of users to reach the specified number of users, sorts all the traffic and restores the original information, and forwards it to the switching chip via unicast. The switching chip removes the private message header and forwards a copy of the traffic to each user from different ports.

Images