CN107947950B

CN107947950B - Hardware multicast output interface table item issuing method and NP (network processor) line card board

Info

Publication number: CN107947950B
Application number: CN201711171928.6A
Authority: CN
Inventors: 徐艳
Original assignee: Hangzhou H3C Technologies Co Ltd
Current assignee: Hangzhou H3C Technologies Co Ltd
Priority date: 2017-11-22
Filing date: 2017-11-22
Publication date: 2020-09-08
Anticipated expiration: 2037-11-22
Also published as: CN107947950A

Abstract

The application provides a method and a device for issuing a hardware multicast output interface table entry, the method can be used for an NP line card board, and the method comprises the following steps: when a global port joins a multicast group, determining whether a physical port associated with the global port exists in a board; if yes, issuing a hardware multicast output interface table item to an NP chip where the physical port is located and a FAP chip of a mesh access processor corresponding to the NP chip; if not, the hardware group broadcasting interface list item is not sent to the NP chip and the FAP chip in the board. The method issues a hardware multicast output interface according to a chip where an actual physical port associated with a global port is located, so that when an FAP chip performs in-board multicast message copying, the FAP chip only copies the message to an NP chip where the actual physical port associated with the global port is located, thereby reducing the number of copies of the message by the FAP chip, reducing the packet loss rate on the FAP chip and the NP chip, and reducing the bandwidth waste between the FAP chip and the NP chip.

Description

Hardware multicast output interface table item issuing method and NP (network processor) line card board

Technical Field

The present application relates to the field of communications technologies, and in particular, to a method for issuing a hardware multicast egress interface table entry and an NP line card.

Background

Multicast is a communication mode parallel to unicast and broadcast, and can effectively solve the problems of single-point transmission and multipoint reception, thereby realizing the high-efficiency data transmission of point to multipoint in the network, saving a large amount of network bandwidth and reducing the network load.

In order to improve the forwarding performance, network products in the market may use a high-performance network processor as a forwarding engine to implement hardware forwarding of messages. The multicast feature of the product can be realized by the cooperation of an NP chip of an NP (Network processor) line card board, an FAP (Fabric Access processor) chip and an FE (Fabric element) chip. The NP chip is responsible for message forwarding, the FAP chip is responsible for in-board multicast message copying, the FE chip is responsible for inter-board multicast message copying, and the NP line card board is responsible for issuing an in-board hardware multicast output interface table entry. Regarding the issuing mode of the hardware multicast output interface table entry, the issuing mode can be divided into the following two types from the perspective of the output interface type:

1) if the output interface of the hardware multicast interface table entry is a physical port, the hardware multicast output interface table entry is issued to an NP chip where the physical port is located and an FAP chip corresponding to the NP chip;

2) if the output interface of the hardware multicast output interface table entry is a global port, the hardware multicast output interface table entry is issued to all NP chips and FAP chips of all NP line card boards of the whole machine.

Disclosure of Invention

In view of this, the present application provides a method for issuing a hardware multicast output interface table entry and an NP line card board, which are used to reduce useless bandwidth waste between an NP chip and an FAP chip and improve the duplication performance of the FAP chip and the forwarding performance of the NP chip when a global port exists in a multicast output interface.

Specifically, the method is realized through the following technical scheme:

in a first aspect of the present application, a method for issuing a hardware multicast egress interface entry is provided, where the method is applied to an NP line card, and the method includes:

when a global port joins a multicast group, determining whether a physical port associated with the global port exists in a board;

if yes, issuing a hardware multicast output interface table item to an NP chip where the physical port is located and a FAP chip of a mesh access processor corresponding to the NP chip;

if not, the hardware group broadcasting interface list item is not sent to the NP chip and the FAP chip in the board.

In a second aspect of the present application, an NP cable card board is provided, which has a function of implementing the above method. The functions can be realized by hardware, and the functions can also be realized by executing corresponding software by hardware. The hardware or software includes one or more modules or units corresponding to the above functions.

In one possible implementation, the NP wire card board includes:

a physical port determining unit, configured to determine whether a physical port associated with a global port exists in a board when the global port joins a multicast group;

the table item processing unit is used for issuing hardware multicast output interface table items to an NP chip where the physical port is located and a mesh access processor FAP chip corresponding to the NP chip if the physical port associated with the global port exists; if not, the hardware group broadcasting interface list item is not sent to the NP chip and the FAP chip in the board.

In another possible implementation manner, the NP cable card board includes a memory, a bus system, and a processor, and the memory and the processor are connected through the bus system; wherein

The memory stores machine readable instructions, and the processor executes any one of the methods of the first aspect by calling the machine readable instructions.

In a third aspect of the present application, a communication device is provided, which includes an FE board and at least one NP wire card board as described in the first aspect, where the FE board is connected to the NP wire card board.

According to the technical scheme, the hardware multicast output interface is issued according to the chip where the actual physical port associated with the global port is located, so that when the FAP chip performs in-board multicast message copying, the message is only copied to the NP chip where the actual physical port associated with the global port is located, the number of copies of the message by the FAP chip is reduced, the packet loss rate on the FAP chip and the NP chip is reduced, and the bandwidth waste between the FAP chip and the NP chip is reduced.

Drawings

Fig. 1 is a schematic diagram of multicast forwarding based on an NP cable card board in the prior art;

FIG. 2 is a flow chart of a method provided by an embodiment of the present application;

FIG. 3.1 is a schematic diagram of a linked list provided by an embodiment of the present application;

fig. 3.2 is a schematic diagram of a linked list after adding members of a global port according to the embodiment of the present application;

FIG. 3.3 is a schematic diagram of a link table after a global port member is deleted according to an embodiment of the present application;

fig. 3.4 is a schematic diagram of a linked list after a global port exits a part of multicast groups according to the embodiment of the present application;

fig. 3.5 is a schematic diagram of a linked list after a global port exits from all multicast groups according to the embodiment of the present application;

fig. 4 is a schematic diagram of multicast forwarding based on an NP cable card board according to an embodiment of the present application;

fig. 5 is a functional block diagram of an NP cable card board provided in an embodiment of the present application;

fig. 6 is a hardware structure diagram of an NP cable card board provided in an embodiment of the present application.

Detailed Description

Reference will now be made in detail to the exemplary embodiments, examples of which are illustrated in the accompanying drawings. When the following description refers to the accompanying drawings, like numbers in different drawings represent the same or similar elements unless otherwise indicated. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with the present application. Rather, they are merely examples of apparatus and methods consistent with certain aspects of the present application, as detailed in the appended claims.

The terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting of the application. As used in this application and the appended claims, the singular forms "a", "an", and "the" are intended to include the plural forms as well, unless the context clearly indicates otherwise. It should also be understood that the term "and/or" as used herein refers to and encompasses any and all possible combinations of one or more of the associated listed items.

It is to be understood that although the terms first, second, third, etc. may be used herein to describe various information, such information should not be limited to these terms. These terms are only used to distinguish one type of information from another. For example, first information may also be referred to as second information, and similarly, second information may also be referred to as first information, without departing from the scope of the present application. The word "if" as used herein may be interpreted as "at … …" or "when … …" or "in response to a determination", depending on the context.

The output interface of the hardware multicast interface table entry can be a physical port or a global port; the global port, i.e. the logical port, includes at least one member port, and the member port included in the global port may be a physical port or another global port. Next, how to implement hardware forwarding of multicast packets in an NP line card board when a multicast output interface is a global interface in the prior art is described with reference to fig. 1.

Referring to fig. 1, an actual physical port included in a global port Route-Aggregation1 is on an NP7 chip of an NP cable card board 1, and when the global port joins a multicast group to form a group broadcast interface, the conventional implementation flow is as follows:

1) when the global port joins the multicast group, the main control board (not shown in fig. 1, which may be a board independent from each NP line card board) applies for CUD (Copy _ Unique _ Data, multicast replication index of multicast packet) in the reserved CUD resource for the global port, and then issues the application result (i.e., the CUD allocated in each NP chip of each NP line card board when the global port is used as the outgoing interface of the multicast group) to each NP line card board through the software information structure.

For the NP chip, the CUD may be used as a key (i.e., a matching item) for multicasting the interface entry by the hardware of the NP chip, and is used to query the interface information and perform hardware forwarding; for the FAP chip, the CUD may be combined with an NP chip identification number and an MID (Multicast ID) to instruct the copy and forward of the Multicast packet in the board. The CUD is corresponding to the hardware multicast output interface table entry, so the CUD resource can be regarded as a hardware resource.

In the prior art, a part of CUD resources are generally reserved in advance as CUD resources of a global port, for example, [0, N ] can be reserved as CUD resources of the global port, that is, CUD resources within a range of 0-N of all NP chips of the whole machine are reserved for the global port to use, and a physical port cannot be used. The purpose of this is to ensure that the CUDs allocated in each NP chip may have the same value when the global port is used as an outgoing interface of a certain multicast group, so that the main control board can issue the application result of the CUD conveniently.

2) And each NP line card board acquires the CUD through the software information structure body, and generates a hardware multicast output interface table item based on the CUD and sends the hardware multicast output interface table item to all NP chips and FAP chips of the NP line card board.

In fig. 1, FAP0-FAP1 and NP0-NP3 chips of NP line card 0, FAP2-FAP3 chips of NP line card 1 and NP4-NP7 chips all receive the hardware multicast output interface table entry related to the global port. Taking the NP cable card 1 in fig. 1 as an example, assuming that the multicast ID of the multicast group added to the global port is MID _1, and the CUD applied by the main control board on all NP chips for the global port is CUD _1, the hardware multicast output interface table items issued by the NP cable card 1 to the NP4-NP7 and FA2-FAP3 may be as shown in table 1.

TABLE 1

3) And the NP line card board copies and forwards the message according to the forwarding table items such as the hardware multicast output interface table items.

In fig. 1, a message enters from an NP0 chip of an NP line card board 0, a hardware multicast entry is searched in an NP0 chip through multicast group information, and if the hardware multicast entry is hit, the message is forwarded by hardware with MID information recorded in the hit hardware multicast entry, and is sent to an FAP0 chip (see an arrow 1 in fig. 1). The FAP0 chip sends a message directly to the FE chip (see arrow 2 in fig. 1). The FE chip broadcasts the message between boards (see arrow 3 in fig. 1), and the FAP0-FAP3 chips all receive the broadcasted message. Since the hardware multicast output interface table entry related to the global port is issued globally, the FAP0-FAP3 chips will copy one packet to each connected NP chip (see arrow 4 in fig. 1), and carry the CUD applied in advance when copying. After receiving the message, the NP0-NP7 chip checks the hardware multicast output interface table entry according to the CUD carried in the message, and finds that the output interface is a global port, further determines whether there is an actual physical port in the chip (for example, when the global port is an aggregation port, it may be confirmed by checking the aggregation port table entry), if so, sends the message from the actual physical port (see arrow 5 in fig. 1), otherwise, discards the message.

It can be seen from the above process that, when the multicast output interface is a global interface, the hardware multicast output interface table entry related to the global interface is issued to NP chips of all NP line cards and FAP chips, so that the message copied between FE chips is copied in-board through all FAP chips and then sent to all NP chips; however, because all NP chips do not have actual physical ports included in the global port, the NP chip without the actual physical port discards the message copied from the FAP chip; the redundant copied messages waste the performance of the NP chip and the FAP chip and the bandwidth between the NP chip and the FAP chip, and the more NP line card boards of the whole machine, the more NP chips on the NP line card boards, the larger the waste. When the FAP chip copies the message to the NP chip, on one hand, due to hardware limitation, when the number of copies of the message is large, the packet loss occurs; on the other hand, because the copying capability of the FAP chip is relatively strong, the copied message can be rapidly sent to the NP chip, and when the number of copies is large, the performance of the NP chip is limited, and packet loss can also occur.

According to the scheme, the hardware multicast output interface table items are issued according to the chip where the actual physical port associated with the global port is located, so that the number of copies of the message of the FAP chip is reduced, the bandwidth waste between the NP chip and the FAP chip is reduced, and the forwarding performance and the utilization rate of hardware resources are improved.

The technical scheme of the application is described in the following with the accompanying drawings and various embodiments of the specification.

Referring to fig. 2, in an embodiment, when any global port joins any multicast group to become an outgoing interface of the multicast group, each NP line card board performs the following steps in the operating process:

step 201: when the global port joins the multicast group, the NP line card board determines whether a physical port associated with the global port exists in the board.

Optionally, when a certain global port joins in the multicast group, the NP line card may obtain the stored corresponding relationship between the global port and the physical port and the corresponding relationship between the physical port and the NP line card to which the physical port belongs, and then determine whether a physical port associated with the global port exists in the board according to the two corresponding relationships.

Step 202: if the physical port exists, the NP line card board issues a hardware multicast output interface table item to the NP chip where the determined physical port is located and the FAP chip corresponding to the NP chip.

Step 203: if not, the NP line card board does not send hardware group broadcast interface list items to the NP chip and the FAP chip in the board.

Before step 202 is executed, the NP line card board needs to apply for a CUD corresponding to the multicast group in an NP chip of the board for a global port joining the multicast group, and in subsequent step 202, a hardware multicast output interface entry to be issued to the NP chip and the FAP chip is generated based on the applied CUD. In the application, a global port and a multicast group ID (identification) uniquely correspond to a group of CUDs, namely, the CUDs allocated to the same global port corresponding to different multicast groups are different; different global ports of the same multicast group are allocated with different CUDs.

In the prior art, the CUDs of the global ports are uniformly allocated in the reserved CUD resources by the main control board. This CUD allocation has the following drawbacks: firstly, when no global port exists, the reserved CUD resource is in an idle state but cannot be used by a physical port, and hardware resources are wasted; second, when there are more global ports, the reserved portion of CUD resources may be insufficient.

In view of this, a new CUD allocation scheme is provided as follows, in this scheme, the CUD of the global port is allocated from the real-time idle CUD resources by each NP line card board, that is, the CUD resources are not reserved for the global port any more, all the CUD resources are shared by the physical port and the global port, and when the global port joins the multicast group, the NP line card board applies for the idle CUD corresponding to the multicast group in the NP chip of the board for the global port joining the multicast group in real time, so that the CUD resources can be saved, and the waste of hardware resources is avoided. Two specific CUD allocation schemes are provided herein.

The first allocation mode is that when a global port joins a multicast group, an idle CUD is respectively applied for the global port in all NP chips of the board. The distribution mode is that the global port respectively pre-occupies a CUD in each NP chip, and the pre-occupying CUD indicates that a hardware table item resource is reserved, so that the condition that the CUD cannot be applied when a new member port is added into the global port can be avoided.

The second allocation method is that when a global port joins a multicast group, an idle CUD is respectively applied for the global port in an NP chip where a physical port associated with the global port of the board is located. If the physical port associated with the global port does not exist in the board, the NP line card board does not need to apply for the CUD for the global port. The allocation formula really realizes instant allocation of CUD resources, and maximizes the utilization of hardware resources.

It should be noted that, no matter which CUD allocation manner is adopted, CUDs allocated to the same global port in each NP chip are unique in the NP chip, but CUDs allocated to the same global port in different NP chips may be the same or different.

Based on the applied CUD, the NP line card board can generate the hardware multicast output interface table items to be issued to the NP chip and the FAP chip. Compared with the prior art, the method and the device change the issuing quantity of the hardware multicast output interface table entries, but do not change the content of the hardware multicast output interface table entries.

Specifically, a key (matching item, also referred to as a keyword) issued to an interface entry of the hardware multicast of the NP chip is a CUD corresponding to a multicast group added to the global port, which is applied by the global port in the NP chip, and a value (matching result, also referred to as a result area) is used to indicate that the interface is the global port; the key issued to the hardware multicast output interface entry of the FAP chip is the multicast ID of the multicast group added by the global port, the value is used for indicating the FAP chip to copy the message to the NP chip under the FAP chip where the physical port associated with the global port is located, and the CUD corresponding to the multicast group and applied by the global port in the NP chip is carried when the message is copied.

Here, as illustrated in fig. 1, according to the technical solution of the present application, since the global port only has an actual physical port on the NP7 chip, the NP cable card board 1 only needs to send the hardware group play interface table entry to the NP7 chip and the FAP3 chip, and the sent hardware group play interface table entry may refer to the 4 th row and the 8 th row in table 1.

After the NP line cardboard is global mouthful and applies for the CUD in the NP chip of this board, can save following information: the MID of the multicast group added to the global port, the CUD applied to the global port in the NP chip at this time, and the NP chip identification number of the actual physical port of the global port in the board; for convenience of description, these pieces of information will be collectively referred to as global port information hereinafter. It should be understood that the global port information may also include: the FAP chip identification number corresponding to the NP chip where the actual physical port of the global port is located in the board, and the identification number of the NP line card board where the global port is located. Each identification number may be a number, a serial number composed of numbers, or a character string composed of numbers, letters, and symbols, and the embodiment of the present application is not particularly limited.

As an implementation, the NP line card board may use a linked list to store the global port information. In the chain table in the application, the global port can be used as an index, and each node of the chain table records the NP chip identification number of the actual physical port of the global port in the plate, the MID of the multicast group added by the global port, and the CUD and other information distributed in the NP chip when the global port is used as an output interface of different multicast groups. When the physical port included by the global port changes or the global port exits from a multicast group, the NP line card board can update the related global port information; when the global port exits from all multicast groups, the NP line card board can remove the node corresponding to the global port in the linked list. The following examples are given.

Suppose that global port 1 and global port 2 join the multicast group; assuming that the global port 1 has an actual physical port on the NP7 chip of the NP cable card 1, the CUD allocated by the global port 1 within the NP7 chip is 1 for the multicast group MID1, and the CUD allocated by the global port 1 within the NP7 chip is 2 for the multicast group MID2, then the linked list for storing the global port information on the NP cable card 1 may be as shown in fig. 3.1.

If the actual physical port of the NP4 chip of the NP cable card 1 is added for the member port of the global port 1, the CUD allocated by the global port 1 in the NP4 chip is 3 for the multicast group MID1, and the CUD allocated by the global port 1 in the NP4 chip is 2 for the multicast group MID2, then the linked list for storing the global port information on the NP cable card 1 may be as shown in fig. 3.2.

If the actual physical port on the NP7 chip exits global port 1 for the member port of global port 1 shown in fig. 3.2, the linked list on the NP cable card 1 for storing global port information may be as shown in fig. 3.3.

If global port 1 exits the multicast group MID1 for global port 1 shown in fig. 3.3, the linked list on the NP linecard 1 for holding global port information may be as shown in fig. 3.4.

If for global port 1 shown in fig. 3.4, global port 1 exits the multicast group MID2 again, and then global port 1 exits all multicast groups, the linked list for storing global port information on NP linecard 1 may be as shown in fig. 3.5.

The method and the device can also adopt the bitmap to store the information of the physical port owned by the global port in the board. A bitmap is a bit (bit) used to mark a certain state of a certain data. For each global port, the NP cable card board may define a character string (or an array) with a length of N according to the number N of NP chips included in the board, and use each bit of the character string (or the array) to represent one NP chip, and specify that when the bit is 1, it indicates that there is an actual physical port on the NP chip represented by the bit, and when the bit is 0, it indicates that there is no actual physical port on the NP chip represented by the bit.

Still taking fig. 1 as an example, in fig. 1, each of the NP line card board 0 and the NP line card board 1 includes 4 NP chips, the global port has an actual physical port only on the NP7 chip, assuming that the lower bit of the character string represents an NP chip with a preceding identification number, and the upper bit of the character string represents an NP chip with a succeeding identification number, the NP line card board 0 may represent an actual physical port without the global port in the board by the bit map of 0000, and the NP line card board 1 may represent an actual physical port with the global port in the board by 1000 and the actual physical port is on the NP7 chip of the board.

When a certain global port becomes an output interface of a multicast group for the first time, the NP line card board can calculate a bitmap according to an NP chip where a physical port associated with the global port in the NP line card board is located; when the global port becomes the output interface of other multicast groups again, the NP line card board can directly determine the NP chip where the physical port associated with the global port in the board is located according to the created bitmap.

It should be noted that the global port includes member ports that are not uniform, and new member ports may be added or existing member ports may be deleted. For member port add and member port exit events, the NP line card board may respond as follows:

1) when a member port adding event aiming at a certain global port is received, the NP cable card board can judge whether the newly added member port has a corresponding physical port in the NP cable card board. If not, no processing is carried out; if yes, generating a hardware multicast output interface table item based on the CUD which is applied by the global port in the NP chip where the newly-added member port is located and corresponds to the multicast group added by the global port, and issuing the generated hardware multicast output interface table item to the NP chip where the newly-added member port is located and the FAP chip corresponding to the NP chip.

If the NP line card board adopts the first CUD distribution mode, namely, when a global port joins in a multicast group, a spare CUD is respectively applied for the global port in each NP chip of the board in advance, and then a hardware multicast outlet interface list item is generated based on the CUD applied for the global port in advance in the NP chip where the newly-added member port is located; if the NP line card board adopts the second CUD allocation manner, that is, when a global port joins a multicast group, only applying for CUD in an NP chip where a physical port associated with the global port is located, then, when it is determined that a corresponding physical port exists in the board for a newly added member port, it is necessary to apply for CUD corresponding to the multicast group joined by the global port in the NP chip where the newly added member port is located in real time for the global port, and then generate a hardware multicast output interface entry based on the CUD applied in real time.

2) When a member port exit event for a certain global port is received, the NP cable card board can judge whether the exiting member port has a corresponding physical port in the NP cable card board. If not, no processing is carried out; if yes, determining a CUD corresponding to a multicast group added by the global port, which is previously applied by the global port in an NP chip where the withdrawn member port is located, indicating the NP chip where the withdrawn member port is located to delete a hardware multicast output interface table item taking the determined CUD as key, and indicating a FAP chip deletion value corresponding to the NP chip to delete a hardware multicast output interface table item including the determined CUD and a chip identification number of the NP chip where the withdrawn member port is located.

As an implementation manner, if the NP line card board adopts the second CUD allocation manner, after the NP chip and the FAP chip are instructed to delete the relevant hardware group play-out interface entry, the CUD corresponding to the multicast group added by the global port, which is applied by the global port in the NP chip where the exited member port is located, may also be released in real time. If the NP line card board adopts the first CUD distribution mode, the CUD corresponding to the quitted multicast group, which is applied by the global port in the NP chip of the board, is released again when the global port quits the multicast group.

Specifically, when the global port exits from a multicast group, the NP line card may perform the following steps: when the global port exits the multicast group, releasing the CUD corresponding to the multicast group, which is applied by the global port in an NP chip of the board, and instructing the NP chip to delete the hardware multicast output interface table entry taking the CUD released by the NP chip as key, and instructing the FAP chip corresponding to the NP chip to delete the value including the chip identification number of the NP chip and the hardware multicast output interface table entry of the CUD released by the NP chip. Therefore, the CUD resources distributed aiming at the global port can not be occupied when the global port exits, the CUD resources distributed aiming at the global port are used as idle CUD resources again for subsequent use, the CUD resources are saved, and the waste of hardware resources is avoided.

It can be seen from the above description that, in the present application, when a global port exists in a multicast output interface, a hardware multicast output interface is issued according to a chip where an actual physical port associated with the global port is located, so that when an FAP chip performs on-board multicast packet replication, a packet is replicated only to an NP chip where the actual physical port associated with the global port is located, thereby reducing the number of copies of the packet by the FAP chip, reducing packet loss rates on the FAP chip and the NP chip, and reducing bandwidth waste between the FAP chip and the NP chip. In addition, the CUD of the global port is selected from real-time idle CUD resources by the NP cable clamping board based on the NP chip, the CUD resources are not reserved for the global port any more, the CUD resources can be saved, and waste of hardware resources is avoided.

The embodiment of the application can also update and send the hardware component broadcasting interface table entry to the correct corresponding NP chip and FAP chip according to the change (addition or deletion) of the global port member, and can realize the correct forwarding of the message.

The embodiment of the application can also avoid occupying the CUD resources distributed aiming at the global port when the global port exits, so that the CUD resources distributed aiming at the global port are used as idle CUD resources again for subsequent use, the CUD resources are saved, and the waste of hardware resources is avoided.

In order to make it clear and obvious for a person skilled in the art, the technical solution of the present application is described below with reference to fig. 4.

Referring to fig. 4, the global port Route-Aggregation1 includes an actual physical port named Ten-gigabit ethernet1/0/1 on the NP7 chip of the NP line card 1.

The first condition is as follows: when the global port Route-Aggregation1 joins a multicast group (S, G) to become a multicast output interface, the implementation flow is as follows:

on NP wire card 0:

and S11, if the global port Route-Aggregation1 is used as the multicast output interface for the first time, calculating and storing a bitmap according to the NP chip where the physical port associated with the global port is located in the board. The member port Ten-gigabit Ethernet1/0/1 of Route-Aggregation1 is a physical port but not on the board, so NP line card board 0 records the bitmap of Route-Aggregation1 as 0000.

If the global port Route-Aggregation1 does not join the multicast group for the first time, the bitmap of the global port already exists, and S12 is directly executed.

S12, the NP cable card board 0 applies for CUD corresponding to the multicast group (S, G) in each NP chip of the board for the global port, and the application result may be as follows:

multicast group (S, G): NP 0(CUD _1), NP 1(CUD _1), NP 2(CUD _5), NP 3(CUD _ 10).

And S13, acquiring the bitmap saved in S11, wherein the bitmap is 0000, which indicates that the hardware multicast output interface table entry does not need to be issued.

On the NP wire card 1:

and S21, if the global port Route-Aggregation1 is used as the multicast output interface for the first time, calculating and storing a bitmap according to the NP chip where the physical port associated with the global port is located in the board. The member port Ten-gigabit Ethernet1/0/1 of Route-Aggregation1 is a physical port and is on the NP7 chip of the board, so the NP line card board 1 records the bitmap of Route-Aggregation1 as 1000.

If the global port Route-Aggregation1 does not join the multicast group for the first time, the bitmap of the global port already exists, and S22 is directly executed.

S22, the NP cable card 1 applies for CUD corresponding to the multicast group (S, G) in each NP chip of the board for the global port, and the application result may be as follows:

multicast group (S, G): NP 4(CUD _9), NP 5(CUD _11), NP 6(CUD _8), NP 7(CUD _ 5).

And S23, acquiring the bitmap saved in S21, wherein the bitmap is 1000, and 1000 indicates that the NP7 needs to issue the hardware multicast output interface table entry, thereby acquiring the CUD _5 pre-occupied by the NP7 in S22, and generating a hardware multicast interface table entry to issue to an NP7 chip and an FAP3 chip. Assuming that the multicast ID of the multicast group added to the global port Route-Aggregation1 is MID _1, the hardware multicast interface entry issued may refer to table 2 below.

TABLE 2

In fig. 4, a message enters from an NP0 chip of the NP line card 0, a hardware multicast entry is searched in an NP0 chip through multicast group information, and if the hardware multicast entry is hit, the message is forwarded by hardware with MID information recorded in the hit hardware multicast entry, and is sent to an FAP0 chip (see an arrow 1 in fig. 4). The FAP0 chip sends a message directly to the FE chip (see arrow 2 in fig. 4). The FE chip broadcasts the message between boards (see arrow 3 in fig. 4), and the FAP0-FAP3 chips all receive the broadcasted message. Because the hardware multicast output interface table entry related to the global port is only issued to the FAP3 chip, only the FAP3 chip copies a message (see arrow 4 in fig. 4) to the NP7 chip connected to the FAP3 chip, and the copy carries the CUD applied in advance. After receiving the message, the NP7 checks the hardware multicast output interface table entry and aggregation interface table entry according to the CUD carried by the message, and sends the message from Ten-gigabit ethernet1/0/1 (see arrow 5 in fig. 4).

Case two: when a new member port Ten-gigabit Ethernet1/0/5 is added to the global port Route-Aggregation1, the Ten-gigabit Ethernet1/0/5 is implemented on an NP4 chip of the NP line card 1, and the process is as follows:

on NP wire card 0:

s31, responding the event added by the member port, analyzing the newly added member port as Ten-gigabit Ethernet 1/0/5.

And S32, determining that the new member port is a physical port but not on the board, and directly returning.

On the NP wire card 1:

s41, responding the event added by the member port, analyzing the newly added member port as Ten-gigabit Ethernet 1/0/5.

And S42, determining that the new member port is a physical port and is on the NP4 chip of the board, updating the bitmap stored in the S21 from 1000 to 1001, and changing the bit representing the NP4 chip in the bitmap from 0 to 1.

S43, after the bitmap of the global port changes, all the multicast groups referring to the global port Route-Aggregation1 are found, the CUD corresponding to each multicast group pre-occupied by the global port in the NP chip of the board is obtained, and the hardware multicast output interface table items are generated and sent to the NP4 chip and the FAP2 chip.

Case two: when the global port Route-Aggregation1 leaves a member port Ten-gigabit Ethernet1/0/1, a Ten-gigabit Ethernet1/0/1 is implemented on an NP7 chip of the NP line card 1, and the process is as follows:

on NP wire card 0:

s51, responding the exit event of the member port, resolving the exit member port as Ten-Gigabitethernet 1/0/1.

And S52, determining that the exit member port is a physical port but not on the board, and directly returning.

On the NP wire card 1:

s61, responding the exit event of the member port, resolving the exit member port as Ten-Gigabitethernet 1/0/1.

And S62, determining that the quitted member port is a physical port and is on the NP7 chip of the board, so that the bitmap updated in S42 is updated from 1001 to 0001 again, and the bit representing the NP7 chip in the bitmap is changed from 1 to 0.

S63, after the bitmap of the global port changes, finding all the multicast groups which refer to the global port Route-Aggregation1, acquiring CUDs which are pre-occupied by the global port in the NP chip of the board and correspond to the multicast groups, and then instructing the NP7 chip and the FAP3 chip to delete the hardware output interface table items related to the acquired CUDs.

The embodiment of the present application further provides a packet forwarding method, which is applied to a communication device including an FE board and at least one NP line card board, where the FE board is connected to the NP line card board, and the NP line card board executes the method for issuing the hardware multicast output interface table entry as described above. In a specific example, with reference to fig. 4 and the foregoing description, it is known that the packet forwarding method provided in the embodiment of the present application can issue a hardware multicast output interface entry according to a chip where an actual physical port associated with a global port is located, so as to reduce the number of copies of packets by an FAP chip, thereby reducing bandwidth waste between an NP chip and the FAP chip, and improving forwarding performance and utilization rate of hardware resources.

The methods provided herein are described above. The apparatus provided in the present application is described below.

Referring to fig. 5, which is a block diagram of functional modules of an NP cable card board provided in an embodiment of the present application, the functional modules may include:

a physical port determining unit 501, configured to determine whether a physical port associated with a global port exists in a board when the global port joins a multicast group.

An entry processing unit 502, configured to issue a hardware multicast egress interface entry to an NP chip where the physical port is located and a mesh access processor FAP chip corresponding to the NP chip if a physical port associated with the global port exists; if not, the hardware group broadcasting interface list item is not sent to the NP chip and the FAP chip in the board.

In one embodiment, the apparatus may further include:

and a CUD application unit, configured to apply for the global port to apply for a CUD corresponding to the multicast group in an NP chip of the board.

The entry processing unit 502 is further configured to generate, based on the applied CUD, a hardware multicast output interface entry to be issued to an NP chip where the physical port is located and an FAP chip corresponding to the NP chip.

In one case, the CUD application unit is specifically configured to apply for an idle CUD in each of all NP chips of the board for the global port.

Correspondingly, the physical port determining unit 501 is further configured to, when a member port adding event for the global port is received, determine whether a corresponding physical port exists in the board for a newly added member port.

The table item processing unit 502 is further configured to not process the newly added member port when the physical port determining unit 501 determines that there is no corresponding physical port in the board; when the physical port determining unit 501 determines that a newly added member port has a corresponding physical port in the board, a hardware multicast output interface entry is generated based on the CUD corresponding to the multicast group, which is applied in advance in the NP chip where the newly added member port is located for the global port, and the generated hardware multicast output interface entry is issued to the NP chip where the newly added member port is located and the FAP chip corresponding to the NP chip.

In another case, the CUD application unit is specifically configured to apply for the global port an idle CUD in an NP chip where a physical port of the board associated with the global port is located.

The table item processing unit 502 is further configured to not process the newly added member port when the physical port determining unit 501 determines that there is no corresponding physical port in the board; when the physical port determining unit 501 determines that the newly added member port does not have a corresponding physical port in the board, it applies for the global port to apply for a CUD corresponding to the multicast group in an NP chip where the newly added member port is located, generates a hardware multicast output interface entry based on the applied CUD, and issues the generated hardware multicast output interface entry to the NP chip where the newly added member port is located and the FAP chip corresponding to the NP chip.

In one embodiment, the matching item of the hardware multicast output interface table item issued to the NP chip is a CUD corresponding to the multicast group and applied for the global port in the NP chip, and the matching result is used to indicate that the interface is the global port; and the matching item of the hardware multicast output interface table item issued to the FAP chip is the multicast ID of the multicast group, the matching result is used for indicating the FAP chip to copy the message to an NP chip under the FAP chip, wherein the NP chip is located by a physical port associated with the global port, and the CUD corresponding to the multicast group and applied by the global port in the NP chip is carried when the message is copied.

In one embodiment, the physical port determining unit 501 is further configured to, when receiving a member port exit event for the global port, determine whether a corresponding physical port exists in the exit member port in the local board.

The table item processing unit 502 is further configured to not process the exit member port when the physical port determining unit 501 determines that the exit member port does not have a corresponding physical port in the board; when the physical port determining unit 501 determines that the exiting member port has a corresponding physical port in the board, it determines the CUD corresponding to the multicast group, which is applied by the global port in the NP chip where the exiting member port is located, and instructs the NP chip where the exiting member port is located to delete the hardware multicast output interface entry which uses the determined CUD as a matching entry, and instructs the FAP chip corresponding to the FAP chip to delete the hardware multicast output interface entry which includes the determined CUD and the chip identification number of the NP chip where the exiting member port is located in the matching result.

In one embodiment, the entry processing unit 502 is further configured to release, when the global port exits the multicast group, a CUD corresponding to the multicast group, which is applied by the global port in an NP chip of the board, instruct the NP chip to delete a hardware multicast output interface entry using the CUD released by the NP chip as a matching entry, and instruct an FAP chip corresponding to the NP chip to delete a matching result that includes a chip identification number of the NP chip and a hardware multicast output interface entry of the CUD released by the NP chip.

The implementation process of the functions and actions of each module in the above device is specifically described in the implementation process of the corresponding step in the above method, and is not described herein again.

It should be noted that the division of the unit in the embodiment of the present invention is schematic, and is only a logic function division, and there may be another division manner in actual implementation. The functional units in the embodiments of the present application may be integrated into one processing unit, or each unit may exist alone physically, or two or more units are integrated into one unit. The integrated unit can be realized in a form of hardware, and can also be realized in a form of a software functional unit.

The functions, if implemented in the form of software functional units and sold or used as a stand-alone product, may be stored in a computer readable storage medium. Based on such understanding, the technical solution of the present application or portions thereof that substantially contribute to the prior art may be embodied in the form of a software product stored in a storage medium and including instructions for causing a computer device (which may be a personal computer, a server, or a network device) to execute all or part of the steps of the method according to the embodiments of the present application. And the aforementioned storage medium includes: a U-disk, a removable hard disk, a Read-Only Memory (ROM), a Random Access Memory (RAM), a magnetic disk or an optical disk, and other various media capable of storing program codes.

Referring to fig. 6, the present application further provides an NP cable card board, which includes a memory 601, a bus system 602, and a processor 603, wherein the memory 601 and the processor 603 are connected via the bus system 602.

The memory 601 stores machine-readable instructions, and the processor executes the above-described method for issuing a hardware multicast egress interface table entry by calling the machine-readable instructions.

The embodiment of the application also provides a communication device, which comprises an FE board and at least one NP line clamping board as described above, wherein the FE board is connected with the NP line clamping board. In a specific example, with reference to fig. 4 and the foregoing description, it is known that the communication device provided in the embodiment of the present application can issue a hardware multicast output interface entry according to a chip where an actual physical port associated with a global port is located, so as to reduce the number of copies of a packet by an FAP chip, thereby reducing bandwidth waste between an NP chip and the FAP chip, and improving forwarding performance and a utilization rate of hardware resources.

The above description is only exemplary of the present application and should not be taken as limiting the present application, as any modification, equivalent replacement, or improvement made within the spirit and principle of the present application should be included in the scope of protection of the present application.

Claims

1. A method for issuing hardware multicast output interface table entries is applied to a network processor NP line card board, and comprises the following steps:

if yes, issuing a hardware multicast output interface table item to an NP chip where the physical port is located and a FAP chip of a mesh access processor corresponding to the NP chip; wherein, the output interface of the hardware multicast interface table entry is a global interface; after the NP chip receives the message, searching a hardware multicast output interface table item based on a CUD carried by the message, and forwarding the message based on an output interface in the hardware multicast output interface table item;

if not, the hardware group broadcasting interface list item is not sent to the NP chip and the FAP chip in the board;

before issuing a hardware multicast output interface table item to an NP chip where the physical port is located and an FAP chip corresponding to the NP chip, the method further comprises the following steps:

applying for the global port a multicast replication index CUD of the multicast message corresponding to the multicast group in an NP chip of the board;

and generating a hardware multicast output interface table item to be issued to an NP chip where the physical port is located and an FAP chip corresponding to the NP chip based on the applied CUD.

2. The method of claim 1, wherein applying for the CUD corresponding to the multicast group in an NP chip of a native board for the global port comprises:

respectively applying for an idle CUD in all NP chips of the board for the global port;

the method further comprises the following steps:

when a member port adding event aiming at the global port is received, judging whether the newly added member port has a corresponding physical port in the board;

if not, no processing is carried out;

if yes, generating a hardware multicast output interface table item based on the CUD corresponding to the multicast group and applied in the NP chip where the newly-added member port is located in advance for the global port, and issuing the generated hardware multicast output interface table item to the NP chip where the newly-added member port is located and the FAP chip corresponding to the NP chip.

3. The method of claim 1, wherein applying for the CUD corresponding to the multicast group in an NP chip of a native board for the global port comprises:

respectively applying for the global port an idle CUD in an NP chip where a physical port of the board, which is associated with the global port, is located;

the method further comprises the following steps:

if not, no processing is carried out;

and if so, applying for the global port to a CUD corresponding to the multicast group in an NP chip where the newly-added member port is located, generating a hardware multicast output interface table item based on the applied CUD, and issuing the generated hardware multicast output interface table item to the NP chip where the newly-added member port is located and an FAP chip corresponding to the NP chip.

4. The method according to any of claims 1-3, wherein the matching item of the hardware multicast output interface table entry issued to the NP chip is a CUD corresponding to the multicast group applied in the NP chip for the global port, and the matching result is used to indicate that the interface is the global port; and the matching item of the hardware multicast output interface table item issued to the FAP chip is the multicast ID of the multicast group, the matching result is used for indicating the FAP chip to copy the message to an NP chip under the FAP chip, wherein the NP chip is located by a physical port associated with the global port, and the CUD corresponding to the multicast group and applied by the global port in the NP chip is carried when the message is copied.

5. The method of claim 4,

the method further comprises the following steps:

when a member port exit event aiming at the global port is received, judging whether the exit member port has a corresponding physical port in the board;

if not, no processing is carried out;

if yes, determining the global port applies for the CUD corresponding to the multicast group in the NP chip where the exited member port is located, instructing the NP chip where the exited member port is located to delete the hardware multicast output interface table entry taking the determined CUD as a matching entry, and instructing the FAP chip corresponding to the NP chip to delete the hardware multicast output interface table entry comprising the determined CUD and the chip identification number of the NP chip where the exited member port is located in the matching result.

6. The method of claim 4,

the method further comprises the following steps:

when the global port exits the multicast group, releasing the CUD corresponding to the multicast group, which is applied by the global port in an NP chip of the board, instructing the NP chip to delete the hardware multicast output interface table entry taking the CUD released by the NP chip as a matching entry, and instructing the FAP chip corresponding to the NP chip to delete the matching result including the chip identification number of the NP chip and the hardware multicast output interface table entry of the CUD released by the NP chip.

7. A Network Processor (NP) cable card, comprising:

the table item processing unit is used for issuing hardware multicast output interface table items to an NP chip where the physical port is located and a mesh access processor FAP chip corresponding to the NP chip if the physical port associated with the global port exists; if not, the hardware group broadcasting interface list item is not sent to the NP chip and the FAP chip in the board; wherein, the output interface of the hardware multicast interface table entry is a global interface; after the NP chip receives the message, searching a hardware multicast output interface table item based on a CUD carried by the message, and forwarding the message based on an output interface in the hardware multicast output interface table item;

the NP line cardboard still includes:

a CUD application unit of multicast replication index of multicast message, configured to apply for the global port a CUD corresponding to the multicast group in an NP chip of the board;

and the table item processing unit is further configured to generate a hardware multicast output interface table item to be issued to an NP chip where the physical port is located and an FAP chip corresponding to the NP chip based on the applied CUD.

8. The NP wire clamp of claim 7,

the CUD application unit is specifically used for respectively applying for an idle CUD in all NP chips of the board for the global port;

the physical port determining unit is further configured to determine whether a corresponding physical port exists in the board for a newly added member port when a member port addition event for the global port is received;

the table item processing unit is further configured to not process the newly added member port when the physical port determining unit determines that there is no corresponding physical port in the board; and when the physical port determining unit determines that the newly added member port has a corresponding physical port in the board, generating a hardware multicast output interface table item based on a CUD (compute unified device description) corresponding to the multicast group, which is applied in advance in an NP (network processor) chip where the newly added member port is located for the global port, and issuing the generated hardware multicast output interface table item to the NP chip where the newly added member port is located and the FAP chip corresponding to the NP chip.

9. The NP wire clamp of claim 7,

the CUD application unit is specifically configured to apply for an idle CUD for the global port in an NP chip where a physical port of the board associated with the global port is located;

the table item processing unit is further configured to not process the newly added member port when the physical port determining unit determines that there is no corresponding physical port in the board; and when the physical port determining unit determines that the newly added member port does not have a corresponding physical port in the board, applying for the global port a CUD corresponding to the multicast group in an NP chip where the newly added member port is located, generating a hardware multicast output interface table item based on the applied CUD, and issuing the generated hardware multicast output interface table item to the NP chip where the newly added member port is located and the FAP chip corresponding to the NP chip.

10. The NP cable card of any one of claims 7 to 9, wherein the matching entry of the hardware multicast output interface entry issued to the NP chip is a CUD corresponding to the multicast group and applied in the NP chip for the global port, and the matching result is used to indicate that the interface is the global port; and the matching item of the hardware multicast output interface table item issued to the FAP chip is the multicast ID of the multicast group, the matching result is used for indicating the FAP chip to copy the message to an NP chip under the FAP chip, wherein the NP chip is located by a physical port associated with the global port, and the CUD corresponding to the multicast group and applied by the global port in the NP chip is carried when the message is copied.

11. The NP wire clamp of claim 10,

the physical port determining unit is further configured to determine whether a corresponding physical port exists in the board for a member port that exits when the member port exit event for the global port is received;

the table item processing unit is further configured to not process the member port determined by the physical port determining unit to exit when the corresponding physical port is not present in the board; when the physical port determining unit determines that the exited member port has a corresponding physical port in the board, determining the CUD corresponding to the multicast group, which is applied by the global port in an NP chip where the exited member port is located, and instructing an NP chip where the exited member port is located to delete a hardware multicast output interface table item which takes the determined CUD as a matching item, and instructing an FAP chip corresponding to the NP chip to delete a hardware multicast output interface table item which comprises the determined CUD and a chip identification number of the NP chip where the exited member port is located in a matching result.

12. The NP wire clamp of claim 10,

the table item processing unit is further configured to release the CUD corresponding to the multicast group, which is applied by the global port in the NP chip of the board, when the global port exits the multicast group, instruct the NP chip to delete the hardware multicast output interface table item using the CUD released by the NP chip as a matching item, and instruct the FAP chip corresponding to the NP chip to delete the hardware multicast output interface table item including the chip identification number of the NP chip and the CUD released by the NP chip in the matching result.

13. A network processor NP line card board is characterized by comprising a memory, a bus system and a processor, wherein the memory and the processor are connected through the bus system;

the memory has stored therein machine-readable instructions, the processor executing the method of any of claims 1 to 6 by calling the machine-readable instructions.