WO2012083654A1

WO2012083654A1 - Method and distributed system for processing internet protocol packet fragments

Info

Publication number: WO2012083654A1
Application number: PCT/CN2011/075923
Authority: WO
Inventors: 刘巍巍; 吴刚; 孙伟
Original assignee: 中兴通讯股份有限公司
Priority date: 2010-12-22
Filing date: 2011-06-20
Publication date: 2012-06-28
Also published as: CN102065012A; CN102065012B

Abstract

The present invention discloses a method, a line card and a distributed system for processing Internet Protocol (IP) packet fragments. The method includes: the line card receives the packet fragments (301); when receiving all the fragments of the packet, the line card determines whether the line card is the destination line card (302); when the line card is not the destination line card, the line card transmits all the fragments of the packet to the destination line card; when the line card is the destination line card, the line card reassembles all the fragments of the packet (303).

Description

IP message fragmentation processing method and distributed system

Technical field

The invention relates to the field of network communication, and specifically relates to a method and a distributed system for processing IP fragments.

Background technique

In the IP network, when the length of the packet exceeds the MTU (Maximum Transmission Unit), the packet needs to be fragmented. After receiving the fragmented packet, the destination device reassembles according to the offset of each fragment. .

Distributed systems typically include centralized control line cards, line cards, and interface line cards that communicate over the internal bus. Figure 1 shows a simplified distributed system composition. Interface line card, externally connected to the IP network. Service line card, to achieve the processing of the upper layer business. The centralized control line card, as the master control node of the system, is responsible for generating the routing table and interface table, and synchronizing to all other line cards.

In the prior art, a distributed system usually configures a centralized reassembly module on one or more service line cards or interface line cards, and reassembles the fragmentation message of the system.

Applying the Chinese patent CN200710097246 on April 28, 2007, the principle of selecting a reconfigurable interface line card is proposed. The fragments of different packets are sent to different interface line cards for reorganization, so as to share the reorganization to some extent. The load on the online card of the module increases the speed of reorganization. If a fragmented packet is received from one or more interface line cards, it will be sent to the online card of the centralized reassembly module for processing. This will result in the packet being on the interface line card, the centralized reassembly module, the online card, and the service line card. Forwarded multiple times. Moreover, due to the limitation of the message channel between the line cards, the mediation between the centralized and reassembled messages may be re-sliced again.

In addition, as shown in FIG. 2, even if the same interface line card receives all the fragmented packets, and the fragmented and non-fragmented packets are not cross-transmitted, the prior art implements the centralized reassembly module by sending the fragmented packets. The online card processing causes the fragmented packet (message 1) and the non-fragmented packet (the packet 2) to be transmitted as different packets to be transmitted in different paths in the system, and finally the packets received by the upper layer service are out of order. The problem is that in some business applications in the communication field, messages of the same stream are not allowed to be out of order. Summary of the invention

The technical problem to be solved by the present invention is to provide a method for processing IP packet fragmentation and a distributed system to avoid multiple retransmissions of packets.

To solve the above problem, the method for processing the IP packet fragmentation provided by the present invention includes: receiving, by the line card, the packet fragmentation;

When the line card receives all the fragments of the message, it determines whether the line card is a destination line card; the card; when the line card is the destination line card, the line card reorganizes all the fragments of the message .

Optionally, the step of the line card receiving the message fragmentation includes:

When receiving the first fragment of the packet, create a virtual reassembly queue;

When receiving the subsequent fragment of the packet, the fragment is written into the virtual reassembly queue at the location in the original packet;

Determine whether all fragments of the packet are received according to the write status of the virtual reassembly queue.

Optionally, the method further includes: when the virtual reassembly queue is created, the timer is initialized. If the line card is an interface line card, and all the fragments of the packet are not received within the timeout period, perform the following operations. :

The interface line card determines a line card where the centralized reassembly module is located;

If the line card where the centralized reassembly module is located is the local line card, the received message fragment is discarded. If the line card where the centralized reassembly module is located is not the local line card, the interface line card will be received. The message fragment is sent to the line card where the centralized reassembly module is located.

Optionally, the method further includes: when the virtual reassembly queue is created, the timer is initialized, and if the line card is a service line card, and all the fragments of the packet are not received within the time limit, the discarding has been performed. The received message is fragmented.

Optionally, the step of determining whether the line card is a destination line card comprises: determining, by the line card, whether the line card is a destination line card by using a query.

Optionally, when the line card is a service line card, the step of determining whether the line card is a destination line card The method includes: determining whether the line card is a destination line card by determining whether the line card is a line card where the centralized reassembly module is located, and if the line card is not located in the centralized reassembly module, the line card is the destination line card; When the line card of the reassembly module is centralized, and the line card is determined to be the destination line card, the service line card directly performs packet reassembly according to the received message fragment.

Optionally, the method further includes: if the currently received packet fragment overlaps or overlaps with the received packet fragment, discarding the currently received packet fragment.

In order to solve the above technical problem, the present invention further provides a line card, where the line card includes: a receiving module, configured to: receive a message fragment;

The destination line card judging module is configured to: determine whether the line card is a destination line card when receiving all the fragments of the message;

a forwarding module, configured to: when receiving all the fragments of the packet and the local line card is not the destination line card, send all the fragments of the packet to the destination line card;

The message reassembly module is configured to: when all the fragments of the message are received and the line card is the destination line card, all the fragments of the text are reorganized.

Optionally, the receiving module includes:

a receiving submodule, which is configured to: receive a message fragment;

a virtual reassembly queue management sub-module, configured to: when receiving the first fragment of the packet, create a virtual reassembly queue; and when receiving the subsequent fragmentation of the packet, according to the location of the fragment in the original text Written into the virtual reassembly queue;

The fragmentation determination sub-module is configured to: determine whether all fragments of the message are received according to the write status of the virtual reassembly queue.

Optionally, the virtual reassembly queue management sub-module is further configured to: when the virtual re-queue queue is created, initialize a timer; when the line card is an interface line card, the line card further includes a centralized reassembly module line card. The determining module is configured to: determine whether the line card is centralized with the line card where the module is located when all the fragments of the packet are not received within the time limit; the forwarding module is further configured to: When the line card is the line card, the received message fragment is discarded. When the line card where the centralized reassembly module is located is not the line card, the received message fragment is sent to the centralized reassembly module. Line card. Optionally, the virtual reassembly queue management sub-module is further configured to: when the virtual re-queue queue is created, initialize a timer; when the line card is a service line card, the forwarding module is further configured to: When all fragments of the packet are not received, the received packet fragment is discarded.

Optionally, the destination line card determining module is configured to determine whether the line card is a destination line card by querying a sub-publishing or routing.

Optionally, when the line card is a service line card, the destination line card determining module is configured to: when receiving the message fragment, determine whether the line card is a line card where the centralized reassembly module is located. Whether the line card is the destination line card. If the line card is not the centralized reassembly module, the line card is the destination line card. The message reassembly module is configured to directly perform message reassembly according to the received message fragmentation.

In order to solve the above technical problems, the present invention also provides a distributed system comprising a plurality of line cards as described above.

In the processing method of the IP packet fragmentation, the line card and the distributed system of the present invention, the line card determines whether the fragmentation of the full packet has been received, and forwards to the destination line card when the reception is complete, and the destination line card implements the message. Reorganization, so as to avoid multiple forwarding and reassembly of packet fragments, saving system resources.

BRIEF abstract

The drawings are intended to provide a further understanding of the invention, and are intended to be illustrative of the invention. In the drawing:

1 is a schematic diagram of a simplified distributed system in an existing communication technology;

2, in the prior art, the fragmented packet and the non-fragmented packet are not transmitted on the same interface board, but the packet is out of order due to the line card where the fragmented packet is sent to the centralized reassembly module;

3 is a schematic diagram of a method for processing an IP packet fragmentation according to an embodiment of the present invention; FIG. 4 is a schematic diagram of a process for receiving packet fragmentation according to an embodiment of the present invention;

5 is a flowchart of processing a packet fragmentation by an interface line card according to an embodiment of the present invention; FIG. 6 is a flowchart of processing a packet fragmentation by a service line card according to an embodiment of the present invention; FIG. 7 is a schematic diagram of avoiding packet disorder after using a virtual reassembly queue according to an embodiment of the present invention; FIG. 8 is a schematic diagram showing a module structure of a line card according to an embodiment of the present invention.

Preferred embodiment of the invention

The main method of processing the IP packet fragmentation method and the distributed system of the present invention is that the line card of the distributed system judges whether the fragmentation of the full packet has been received, and forwards to the destination line card when the reception is complete, and the destination line card The message is reorganized to avoid multiple forwarding and reassembly of packet fragments, saving system resources and avoiding out-of-order packets.

As shown in FIG. 3, the processing method of the IP fragment of the present invention includes:

Step 301: The line card receives the packet fragmentation.

Step 302: When the line card receives all the fragments of the packet, determine whether the line card is a destination line card;

Step 303: When the line card is not the destination line card, the line card sends all the fragments of the message to the destination line card; when the line card is the destination line card, the line card will All shards were reorganized.

As shown in FIG. 4, step 301, the line card receiving the packet fragmentation specifically includes the following steps: Step 3011: When receiving the first fragment of the packet, create a virtual reassembly queue; Step 3012: Receive the When the subsequent fragmentation of the packet is performed, the fragment is written into the virtual reassembly queue in the original packet.

Step 3013: Determine whether all the fragments of the packet are received according to the write status of the virtual reassembly queue.

To ensure the time-dependent control of the received packet fragmentation, when the virtual reassembly queue is created, the timer is initialized, if the line card is an interface line card, and all the fragments of the packet are not received within the time limit. , do the following:

If the line card where the centralized reassembly module is located is the local line card, the received message fragment is discarded. If the line card where the centralized reassembly module is located is not the local line card, the interface line card will be received. Newspaper The fragment is sent to the line card where the centralized reassembly module is located.

If the line card is a service line card and does not receive all the fragments of the packet within the time limit, the received message fragment is discarded.

The line card determines whether the line card is the destination service line card by querying the sub-publishing or routing.

If the currently received packet fragment overlaps or overlaps with the received packet fragment, the currently received packet fragment is discarded.

When the line card is a service line card, it is determined whether the line card is a destination line card by determining whether the line card is a line card where the centralized reassembly module is located. If the line card is not located in the centralized reassembly module, the line card is The destination line card is determined by determining the line card in which the centralized reassembly module is located, and the service line card directly performs packet reassembly according to the received message fragmentation.

The present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be noted that the embodiments in the present application and the features in the embodiments may be combined with each other without conflict. It is understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention. Embodiment 1

Refer to Figure 5 for the complete workflow of fragment packet reassembly on the interface line card, including the following steps:

5501, receiving a message fragment;

5502, determining whether the received packet fragment is the first arriving fragment, and if yes, proceeding to S503; otherwise, transferring to S504;

5503, create a virtual reassembly queue, initialize a timeout timer, and transfer 507;

5504, determining whether the received packet fragmentation is repeated with the received packet fragment, and if yes, proceeding to S506; otherwise, proceeding to S505;

5505, determining whether the received packet fragment overlaps with the accepted packet fragment, and if yes, proceeds to S506; otherwise, proceeds to S507;

5506, discarding the currently received fragment;

5507, according to the fragment offset, insert the fragment into the virtual reassembly queue; 5508, determining whether the tail piece has arrived, and if so, then proceeds to S309; otherwise, proceeds to S517;

5509, determining whether the first piece has arrived, and if so, then transferring to S310; otherwise, transferring to S517;

5510, determining whether there is a hole in the virtual reorganization queue, if yes, then transferring to S517; otherwise, transferring to S511;

Steps 508 through 510 do not have a strict sequence, mainly to determine whether all fragments of the message have been received.

5511, judging whether there is a message sub-publication, if it exists, then transferring to S512; otherwise, transferring to S513;

5512, extract source information, destination address, source port, destination port, protocol type and other information from the first piece, find the sub-publish, obtain the destination line card, and transfer to S514;

S513. Find a routing table to obtain a destination line card.

5514, determine whether the line card is a destination line card, and if so, then proceeds to S515; otherwise, proceeds to S516;

5515, the message is reorganized successfully, and the line card service is sent to the service, and the process ends;

5516: All the fragments in the virtual reassembly queue are sent to the destination line card for processing, and the process ends;

5517. Determine whether the timer expires. If yes, go to S518; otherwise, go to S501 and continue to receive the packet fragmentation;

5518, extracting source information, destination address, protocol type and other information from the message to perform hash operation, and obtaining the destination line card of the centralized reorganization module;

5519, judge whether the line card should be centralized to reorganize the module's online card, and if so, go to S520; otherwise, turn to S521;

S520, releasing the virtual reorganization queue, the reorganization fails, and the process ends;

S521: All the fragmented packets in the virtual reassembly queue are distributed to the online card processing of the centralized reassembly module, and the process ends.

According to the above process, it can be seen whether the real reorganization is performed, and the centralized reorganization module is not related to the line card. There is only one principle: the line card is the destination line card, then the real reorganization; otherwise, the virtual reorganization. That is to say, if the online card of the centralized reorganization module is not the destination line card, then the online card performs virtual reorganization, otherwise the real reorganization is performed, and the second card is not forwarded and reorganized. According to the above process, if the interface line card does not collect all the fragmented packets within the reassembly timeout period, it performs hash operation according to the source address, destination address, and protocol type of the packet, and obtains centralized reorganization. The line card where the module is located. If the online card of the centralized reorganization module is the local line card, the virtual reassembly queue is released, and the reorganization fails; otherwise, the "^艮 shards in the 4 bar virtual reassembly queue are sent to the centralized reassembly module for reorganization on the online card.

If all the fragmented packets are received within the reassembly timeout period, the source address, the destination address, the source port, the destination port, and the protocol type are extracted from the first packet, and the information is searched or reported according to the destination address of the packet. Determine whether the message is the line card for business processing or sent to the destination line card for business processing. If the line card performs service processing, the packets in the virtual reassembly queue are successfully reassembled. Otherwise, each fragment packet in the virtual reassembly queue is sent to the destination line card for processing.

Embodiment 2

Refer to Figure 6 for the complete workflow of fragment packet reassembly on the service line card, including the following steps:

S601, receiving a packet fragmentation;

5602, determining whether the line card is a online card of the centralized reorganization module, if not, then transferring to S603; if yes, proceeding to S604;

S601 and S602 do not have a strict sequence.

5603, the message is reorganized, and the process ends;

Generally, if S603 is executed, it indicates that the current line card is not the line card where the message reassembly module is located, and it must be the destination line card. The message fragment is derived from S516 of the process of Figure 5, indicating that the message received by the same interface line card is received. For the fragmentation, the transmission path is from the interface line card to the service line card as the destination line card. Specifically, the message can be reorganized according to the prior art. For example, when the message is reorganized, a timer is set, and the timer is collected. For all fragments, the real message reassembly is performed. Otherwise, the message reorganization fails, and the received >3⁄4 fragments are discarded.

5604, determining whether the received packet fragmentation is the first arriving fragment, and if yes, proceeding to S605; otherwise, proceeding to S606;

Execution S604, indicating that the current line card is the line card where the centralized reassembly module is located, except for the packet fragmentation. S516, which can be derived from the flow of FIG. 5, may also be derived from S521 of the flow of FIG. 5, indicating that the packets received by different interface line cards are fragmented, and the transmission path is fragmented by a different interface line card. Converged to the line card where the centralized reassembly module is located.

S605, creating a virtual reassembly queue, initializing a timeout timer, and transferring to S609;

S606, determining whether the received packet fragmentation is repeated with the received packet fragment, and if yes, proceeding to 608; otherwise, proceeding to S607;

5607, determining whether the received message fragment overlaps with the accepted message fragment, if yes, then proceeds to S608; otherwise, proceeds to S609;

5608, discarding the currently received fragment;

S609. Insert the fragment into the virtual reassembly queue according to the fragment offset, that is, perform an orderly buffer according to the location of the packet fragment in the original packet.

5610, determine whether the tail piece has arrived, and if so, then proceeds to S611 otherwise, and proceeds to S619;

5611, to determine whether the first film has arrived, and if so, then proceeds to S612; otherwise, proceeds to S619;

5612. Determine whether the virtual reassembly queue has a hole. If yes, go to S619; otherwise, go to S613;

Steps 610 through 612 do not have a strict sequence, mainly to determine whether all fragments of the message have been received.

5613, to determine whether there is a sub-publication. If yes, go to S614; otherwise, go to S615;

5614, extract the source address, destination address, source port, destination port, protocol type and other information from the first slice, find the sub-publish, get the destination line card, and transfer to S616.

5615, find the route, and get the destination line card;

5616, determine whether the line card is the destination line card, and if yes, go to S617; otherwise, go to S618;

5617, the text is reorganized successfully, and the line card service is sent to the line, and the process ends;

Execution S617 indicates that the current line card is both the line card where the centralized reassembly module is located, and the destination line card, for real "^ text reorganization.

5618: All the fragmented packets in the virtual reassembly queue are distributed to the destination line card for processing, and the process ends; executing S618 identifies that the current line card is only the line card where the centralized reassembly module is located, instead of the destination line. The card is virtual reorganized according to the virtual reassembly queue, sent to the destination line card, and then the real line message is reassembled by the destination line card.

S619. Determine whether the timer expires. If yes, go to S620; otherwise, go to S601 and continue to receive message fragments;

S620: The virtual reassembly queue is released, the packet reorganization fails, and the process ends.

The above process is a preferred implementation manner of the service line card. When the S602 is judged, the current line card is not the line card where the message reassembly module is located but the destination line card is simplified, and the virtual reassembly queue does not need to be created. Alternatively, step S602 may be performed after receiving the fragmented message without performing steps S602 and S603.

According to the above process, for the line card:

If all the shards are not collected within the reassembly timeout period, the shards in the virtual reassembly queue are released, and the reorganization fails.

If all the fragmented messages are collected within the reassembly timeout period. Extract the information such as the source address, the destination address, the source port, the destination port, and the protocol type from the first packet, check the distribution or check the route according to the destination address of the packet, and determine whether the packet is the line card for service processing or delivery to the destination line. The card performs business processing. If the line card performs service processing, the packets in the virtual reassembly queue are successfully reassembled. Otherwise, each fragment packet in the virtual reassembly queue is sent to the destination line card for processing.

The invention only realizes real message reorganization in the destination line card, can avoid secondary reorganization, save system resources, and improve system performance.

In the case that the current line card is not the destination line card, the virtual reassembly queue can be used to forward the packets in the virtual reassembly queue of a certain packet in a centralized manner, so that the packets of different packets are relatively fragmented. Packets that are independent of other packets are fragmented. The packets are isolated from each other to prevent the packets received from the upper-layer services from being out of order.

For the situation shown in Figure 2, if the interface line card, the service line card as the destination line card, and the line card where the >3⁄4 text reassembly module is located are different line cards, the message transmission path is as shown in Figure 7. After the virtual reassembly of the virtual reassembly queue on the interface line card 1 is performed, the fragment 1 and the fragment 2 of the packet 1 are sent to the service line card without going through the line card where the centralized reassembly module is located, and The transmission path of the packet 2 is the same, thereby avoiding the disorder between the message 1 and the message 2. In order to implement the above method, the present invention further provides a line card. As shown in FIG. 8, the line card includes:

a receiving module 80, configured to receive a message fragment;

The destination line card judging module 81 is configured to: determine whether the line card is a destination line card when receiving all the fragments of the message;

The forwarding module 82 is configured to: when all the fragments of the packet are received and the local line card is not the destination line card, send all the fragments of the packet to the destination line card;

The message reassembly module 83 is configured to: when all the fragments of the message are received and the line card is the destination line card, all the fragments of the text are reorganized.

Preferably, the receiving module 80 includes:

a receiving submodule 801, configured to receive a message fragment;

a virtual reassembly queue management sub-module 802, configured to: when receiving the first fragment of the packet, create a virtual reassembly queue; and when receiving the subsequent fragment of the packet, according to the fragment in the original text The location is written into the virtual reassembly queue;

The fragmentation determining sub-module 803 is configured to determine whether all fragments of the message are received according to the writing condition of the virtual reassembly queue.

Preferably, the virtual reassembly queue management sub-module 802 is further configured to initialize a timer when the virtual reassembly queue is created, and when the line card is an interface line card, further include a centralized reassembly module line card judging module ( If not shown in the figure, it is set as follows: When all the fragments of the packet are not received within the time limit, it is determined whether the line card is centralized or not, and the forwarding module 82 is further set in the centralized When the line card where the reassembly module is located is the line card, the received packet fragments are discarded. When the line card where the centralized reassembly module is located is not the line card, the received message fragments are sent to the centralized reassembly. The line card where the module is located.

Preferably, the virtual reassembly queue management sub-module 802 is further configured to: when the virtual re-queue queue is created, the timer is initialized, and when the line card is a service line card, the forwarding module 82 is further set to be in timing. When all the fragments of the packet are not received within the time, the received packet fragment is discarded.

The destination line card judging module 81 is configured to determine the line card by querying a sub-publishing or routing. Whether it is the destination line card.

Corresponding to the flow of FIG. 6 , when the line card is a service line card, the destination line card determining module 81 is configured to: when receiving the message fragment, determine whether the line card is a centralized reassembly module. The line card determines whether the line card is the destination line card. If it is not the line card where the centralized reassembly module is located, the line card is the destination line card. The line card is determined by determining the line card where the centralized reassembly module is located. In the case of the line card, the message reassembly module 83 is configured to directly perform message reassembly according to the received message fragmentation, including: if all the fragmentation is received within the time limit, the reorganization is successful, otherwise the reorganization fails and is discarded. The received message is fragmented.

In order to implement the above method, the present invention also provides a distributed system comprising a plurality of line cards as described above.

The method, the device and the system of the present invention enable the fragmented message to be reorganized only on the line card (ie, the destination line card) that is processed by the upper layer service. Otherwise, only the virtual reorganization is performed, thereby avoiding the traditional implementation of sending the fragmented message to the centralized device. The drawbacks caused by the real reorganization and redistribution of the online card of the reorganization module, such as the centralized reorganization module, are easy to become the system bottleneck. The retransmission of the message between the online cards may lead to secondary fragmentation and reorganization, as well as packet disorder. The traditional telecom field implementation is a brain that sends fragmented packets to the centralized card processing of the centralized reassembly module. The present invention can improve the performance of the reorganization performance in order to avoid excessive load on the board and affect business operations. It avoids the performance loss caused by unnecessary re-segmentation and re-reorganization after the message is reorganized, and improves the efficiency of message processing. At the same time, when the fragmented message and the non-fragmented message are not cross-transmitted and received by the same interface line card, the message can be saved because the message does not need to be sent to the centralized reassembly module to reorganize the online card. The system's virtual reorganization can be enabled or disabled through background configuration. In addition, the present invention resides on a plurality of interface line cards or service line cards to reside in a centralized reassembly module, thereby realizing the load balancing of the virtual reorganization on the online card of the centralized reassembly module.

The above description is only the preferred embodiment of the present invention, and is not intended to limit the present invention, and various modifications and changes can be made to the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and scope of the present invention are intended to be included within the scope of the present invention.

One of ordinary skill in the art can understand that all or part of the above steps can be completed by a program to instruct related hardware, and the program can be stored in a computer readable storage medium, such as read only. Memory, disk or disc, etc. Alternatively, all or part of the steps of the above embodiments may also be implemented using one or more integrated circuits. Correspondingly, each module/unit in the foregoing embodiment may be implemented in the form of hardware, or may be implemented in the form of a software function module. The invention is not limited to any specific form of combination of hardware and software.

Industrial applicability

In the present invention, the line card determines whether the fragmentation of the full message has been received, and forwards the packet to the destination line card when the reception is complete, and the destination line card reconciles the message, thereby avoiding multiple forwarding and reassembly of the packet fragmentation. Save system resources.

Claims

Claim

1. A method for processing IP 艮分分 , , , , , ,

The line card receives the packet fragmentation;

When the line card receives all the fragments of the message, it determines whether the line card is a destination line card; when the line card is the destination line card, the line card reorganizes all the fragments of the message.

The method of claim 1, wherein the step of the line card receiving the message fragmentation comprises: when receiving the first fragment of the message, creating a virtual reassembly queue;

3. The method of claim 2, further comprising: initializing a timer when the virtual reassembly queue is created, if the line card is an interface line card, and not receiving all of the messages within a time limit When sharding, do the following:

The method of claim 2, further comprising: initializing a timer when the virtual reassembly queue is created, if the line card is a service line card, and not receiving all the packets within the time limit When fragmentation, discard the received message fragments.

The method of claim 1, wherein the step of determining whether the line card is a destination line card comprises: determining, by the line card, whether the line card is a destination line card by querying a distribution or routing.

The method of claim 1, wherein, when the line card is a service line card, the step of determining whether the line card is a destination line card comprises: determining whether the line card is a centralized reassembly module In the line card, to determine whether the line card is the destination line card, if it is not the line card where the centralized reassembly module is located, the line card is the destination line card; determine the line by judging the way of the line card where the centralized reorganization module is located When the card is the destination line card, the service line card directly performs message reassembly according to the received message fragment.

The method of claim 1, the method further comprising: discarding the currently received packet fragment if the currently received packet fragment overlaps or overlaps with the received packet fragment.

8. A line card, the line card comprising:

a receiving module, configured to: receive a message fragment;

9. The line card of claim 8, wherein the receiving module comprises:

a receiving submodule, which is configured to: receive a message fragment;

10. The line card of claim 8, wherein the virtual reassembly queue management sub-module is further configured to: initialize a timer when the virtual reassembly queue is created;

When the line card is an interface line card, the line card further includes a centralized reassembly module line card judging module, which is configured to: determine whether the line card is reorganized or not when all the fragments of the message are not received within the time limit. The line card where the module is located; the forwarding module is further configured to: where the centralized reassembly module is located When the line card is the line card, the received message fragment is discarded. When the line card where the centralized reassembly module is located is not the line card, the received message fragment is sent to the line where the centralized reassembly module is located. card.

The line card according to claim 8, wherein the virtual reassembly queue management sub-module is further configured to: when the virtual re-queue queue is created, initialize a timer; when the line card is a service line card, The forwarding module is further configured to: when all the fragments of the packet are not received within the timeout period, discard the received packet fragment.

The line card according to claim 8, wherein the destination line card judging module is configured to determine whether the line card is a destination line card by querying a sub-publishing or routing.

The line card according to claim 8, wherein, when the line card is a service line card, the destination line card determining module is configured to: when receiving the message fragment, determine whether the line card is To determine whether the line card is the destination line card for the line card where the reassembly module is located, if the line card is not the centralized reassembly module, the line card is the destination line card; the message reassembly module is set to receive according to the The message fragmentation directly performs message reorganization.

A distributed system comprising a plurality of line cards as claimed in any one of claims 8-13.