WO2014101708A1 - Data transmission method and network node in layer 2 network - Google Patents

Data transmission method and network node in layer 2 network Download PDF

Info

Publication number
WO2014101708A1
WO2014101708A1 PCT/CN2013/089936 CN2013089936W WO2014101708A1 WO 2014101708 A1 WO2014101708 A1 WO 2014101708A1 CN 2013089936 W CN2013089936 W CN 2013089936W WO 2014101708 A1 WO2014101708 A1 WO 2014101708A1
Authority
WO
WIPO (PCT)
Prior art keywords
interface
mac address
network node
network
group
Prior art date
Application number
PCT/CN2013/089936
Other languages
French (fr)
Chinese (zh)
Inventor
喻敬海
付振涛
韩辉
Original Assignee
中兴通讯股份有限公司
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Priority to CN201210574336.X priority Critical
Priority to CN201210574336.XA priority patent/CN103905325B/en
Application filed by 中兴通讯股份有限公司 filed Critical 中兴通讯股份有限公司
Publication of WO2014101708A1 publication Critical patent/WO2014101708A1/en

Links

Classifications

    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04LTRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L12/00Data switching networks
    • H04L12/28Data switching networks characterised by path configuration, e.g. local area networks [LAN], wide area networks [WAN]
    • H04L12/2854Wide area networks, e.g. public data networks
    • H04L12/2856Access arrangements, e.g. Internet access
    • H04L12/2863Arrangements for combining access network resources elements, e.g. channel bonding
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04LTRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L12/00Data switching networks
    • H04L12/02Details
    • H04L12/16Arrangements for providing special services to substations contains provisionally no documents
    • H04L12/18Arrangements for providing special services to substations contains provisionally no documents for broadcast or conference, e.g. multicast
    • H04L12/1863Arrangements for providing special services to substations contains provisionally no documents for broadcast or conference, e.g. multicast comprising mechanisms for improved reliability, e.g. status reports
    • H04L12/1868Measures taken after transmission, e.g. acknowledgments
    • H04L12/1872Measures taken after transmission, e.g. acknowledgments avoiding ACK or NACK implosion
    • YGENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y02TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02DCLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THIR OWN ENERGY USE
    • Y02D50/00Techniques for reducing energy consumption in wire-line communication networks
    • Y02D50/30Techniques for reducing energy consumption in wire-line communication networks by selective link activation in bundled links

Abstract

Embodiments of the present invention relate to a data transmission method and a network node in a layer 2 network. The method comprises: configuring, according to a network architecture, a balance group for a network node on the basis of a principle for avoiding a broadcast storm, one balance group comprising one or more interfaces; the network node receiving a packet from a receiving interface; and the network node forwarding the packet to an interface excluding the balance group the receiving interface is located.

Description

Layer network data transfer method and a network node

FIELD

The present invention relates to network communication technology, and more specifically, is a Layer 2 network data transmission method and a network node.

Background technique

With the increasingly wide range of Internet applications, data center applications are increasingly widespread. In the relevant standards organizations, and all relevant technical standards for data centers is also actively advancing. In order to achieve more convenient virtual machines dynamically migrate data center network, data center network requires a large Layer 2 networks become the basic requirement to build a data center.

Layer 2 network, in order to improve reliability and robustness of the network connection, the Internet is usually set in the redundancy, so when a failure occurs, traffic can be switched to the backup link. However, there are redundant links exist naturally lead to a loop in the network, which led to the emergence of network broadcast storms, causing various problems such as traffic congestion. STP (Spanning Tree Protocol, Spanning Tree Protocol) protocol and the subsequent improvement between network nodes may be avoided by changing the device node loop port state, while when the link failure, so that by changing the device port status data forwarding forwarded through the other link up to ensure continued network data transfer. But STP protocol and its subsequent improvement in data center applications with its fatal weakness, that is a serious lack of bandwidth utilization, a large number of link port is blocked because the protocol can not be used. To this end, the industry put forward several related technology to achieve non-STP large Layer 2 network, currently much discussed are IRF (Intelligent Resilient Framework, Intelligent Resilient Framework), VSS (Virtual Switching System, Virtual Switching System), VPC (virtual Port-channel, virtual channel technology), TRILL (transparent interconnection of Lots of links, multilink transparent interconnection) technology and SPB (shortest path bridging, shortest path bridging) technology.

IRF technology, VSS and VPC techniques are stack-based technical implementation. IRF technology is an application in the low-end switch device virtualization technology. It allows multiple low-end switches Taichung device (device member) is a high-end virtual switch device (virtual device). Meanwhile, the physical member ports of each device are bundled together as one logical port, similar to the single physical device in an actual network configuration and use, whereby the device can be cross-link aggregation function, which avoids ring the occurrence of the road, while increasing the robustness of the link. From the point of view of the data center networking diagram, the network structure after polymerization using IRF corresponds to a tree structure, the loop condition no longer occur. IRF technology need to enable IRF virtual control protocol on the member devices, and exchanging packets between the virtual member devices.

TRILL and SPB technology technology, technical aspects of the multi-path data forwarding are achieved. TRILL is a technology proposed by IETF "Ethernet frames multi-path forwarding" technology program, presented the technology to subvert the traditional way of forwarding the Ethernet frame. Because the multi-path technology has always been only three IP technology. Application of this technology in the TRILL switcher device, the switch can realize this function is called "routing bridge (Routing Bridge, referred RBridge)." Link-state control protocols (to Intermediate System Routing Protocol, Intermediate System to Intermediate System Routing Protocol Intermediate System) by a similar inter-IS-IS routing protocol RBridge TRILL (Transparent Interconnection of Lots of Links, translucent multi-link network) the IS -IS calculate the shortest path to each other and equivalent multipath. TRILL IS-IS topology is calculated only between RBridge, without concern for network topology between two hosts. When forwarding the Ethernet frame, the Ingress RBridge (inlet routing bridge) Encapsulation RBridge source identification and destination identification RBridge in the Ethernet frame, coupled with the outer VLAN identification, identification outer RBridge next hop, hop by hop to this egress RBridge (outlet routing bridge) node, egress RBridge the TRILL header stripped, to restore a conventional Ethernet frame, for the conventional Ethernet frame forwarding. By using the building data center network switch device having characteristics such technologies to achieve a large STP no no loop layer 2 network. But to achieve by this technique a large Layer 2 network switch equipment requirements must be able to support this technology, and the original switch is not a simple software upgrade to support. SUMMARY

Embodiments of the present invention to provide a data transmission method and a network Layer network node to address the network Layer prone to broadcast storms.

Embodiment of the present invention provides a Layer 2 network data transfer method, comprising:

The network architecture principles to avoid broadcast storm is configured as a balanced group of network nodes, a balancing group comprises one or more interfaces; ^ gen network node receives packets from the receiving interface; and

The network node forwards the packet to the interface is located outside the reception interface balancing group.

Embodiment of the present invention provides a network node, the network node comprising:

Several interfaces;

Balanced configuration unit, which is arranged according to the network architecture principles to avoid broadcast storm is configured as a balanced group of network nodes, a balancing group comprising one or more interfaces; and

Packet forwarding control module, when receiving from the interface to receive packets, which is arranged to the receive interface is located outside the interface to forward packets balancing group.

A method and a network node embodiment of the present invention varies according to the network architecture in advance to avoid broadcast storm principle configuration balanced group of network nodes, for packet forwarding, all the balancing group according to a preset equalization than group where the other reception interface interface to forward packets to avoid broadcast storm, solutions of the embodiments of the present invention only needs to be set lower requirements of node devices, and easy to implement. Bu the other evening when, in the need for MAC (Media Access Control, MAC) address learning and aging, the same balancing group interfaces according to the strategy to achieve a balanced sharing of MAC address learning and improve the performance of the system. Brief Description

Layer 1 of the present invention is a method of data transmission network 1 is a schematic diagram of the embodiment;

FIG 2 is a schematic view of a balanced set of configuration of an embodiment of the present invention;

FIG 3 is a schematic diagram of the Layer 2 network data transfer method of the present invention;

FIG 4 is a schematic of an embodiment of the present invention, the MAC processing;

FIG 5 is a schematic flow diagram of a specific MAC learning performed;

FIG. 6-8 is a schematic block configuration example of embodiment of the network node embodiment of the present invention;

9-11 are schematic application example of the present invention; wherein Fig 9 is an embodiment of the present invention in a schematic view MAC learning and forwarding the standard fat tree structure; FIG. 10 is a non-standard forwarding fat tree structure of the embodiment of the present invention and the MAC learning schematic; FIG. 11 is a forwarding and multicast MAC learning in one embodiment of the present invention where a schematic view. Preferred embodiment of the present invention.

DRAWINGS Examples and embodiments of the present invention will be described in the detailed description of the technical solutions implemented to enable those skilled in the art may better understand the embodiments of the invention and can be implemented, but not cited embodiments of the present invention as It defined. Incidentally, in the case of no conflict, embodiments and features of the embodiments of the present application can be combined with each other.

Example 1

Layer network data transfer method of the present invention, Example 1, shown in Figure 1, the method comprising: Step 101: The network architecture to avoid broadcast storm is the principle of equalization network node group configuration, a balanced group comprises one or more interfaces ;

Said network node embodiment of the present invention may be a network access device and the core network device or a polymerization apparatus, for better description of the embodiments of the present invention, the following concepts are introduced herein defined:

Uplink interface: data center network structure, the interface with the upper node interactive network, the network node of this layer uplink interface.

Downstream interface: data center network architecture, network nodes interact with the underlying interface layer for downstream network node interface.

Embodiment of the present invention, the balancing group within the interface for both the uplink and the upper interface or a network device interaction are lower node downstream interface for interacting with the network.

Balancing group: group may also be referred to as an interface or port group to avoid broadcast storm principle of the same type of interface (uplink or downlink interface or interfaces the uplink interface or downlink interface part of the interface), to bundle together, preventing the received set of interface to flow within the group to another interface to forward, at the same time share or traffic sharing MAC address learning, as well as traffic protection. Similarly link bundle, but it does not require protocol negotiation, and local alone, and need not have the same peer-balancing group.

Wherein the flow sharing by uniformly on the port learning peer MAC balancing group, the subsequent forwarding to the outside when the balancing group, natural look-up table, the basic traffic MAC balanced, reflecting the traffic balancing may be considered.

For a multi-link environment is determined Ethernet network nodes, or its interface as an uplink interface, or an interface downlink, uplink interface UP indicates, DP represents the downlink interface; downlink or uplink interface in the interface, according to We need to configure a balanced group. 2, the balancing group comprising all or uplink interface includes all of the downstream interfaces, the networking, the balancing group may include only a portion of the uplink interface or only a portion of the downlink interface. Wherein, if the network node is a network access device, and the polymerization apparatus, uplink interface can be configured as a balanced group; if the core network node is a network device, the configuration of the topology of the network where the polymerization apparatus core balancing device is connected group.

Step 102: the network interface receives packets from a node received;

Example packets of said embodiment of the present invention is mainly required flooding packet, broadcast packet, a broadcast packet or an unknown protocol packets. For unicast packets, you need to find forwarding, forwarding to the corresponding port will not be forwarded to all ports.

Step 103: the network node forwards the packet to the interface is located outside the reception interface balancing group.

Example 1 In this embodiment, the same balancing group of packets received by the interface, the reception interface belongs to no group other interfaces forward equalizer, the equalizer when a group includes all uplink interface, from the uplink interface received broadcast / multicast, or unknown unicast packets, does not forward the uplink interface to another. Due to the advance in accordance with the network architecture to prevent broadcast storms principle of a balanced set of configuration for the network node performs packet forwarding, based on a balanced set of preset receiving interface to the outside where the balancing group of forwarding packets to all other interfaces can effectively avoid the ring generating passage, i.e. to avoid broadcast storm, solutions of the embodiments of the present invention only needs to be set lower requirements of node devices, and easy to implement.

Example 2

Layer network data transfer method of the present invention, Example 2, as shown in Figure 3, the method includes: Step 301: The network architecture to avoid broadcast storm is the principle of equalization network node group configuration, a balanced group comprises one or more interfaces ;

Interfaces are set in the equalization of the upstream port to the upper node are exchanged in the network, or the lower node downstream interface for interacting with the network.

Step 302: the network interface receives packets from a node received;

Step 303: The network node MAC processing packets MAC (Media Access Control, media access control) information;

Specifically, as shown in FIG. 4, the MAC process comprising:

Step 401: the network node according to the packet information carried in MAC queries the local MAC address table;

Step 402: If no matching entry is found in the address, then the network node MAC address learning;

The MAC address learning step comprises: determining whether the balancing group receiving other interfaces where the interface;

If no other interface, the reception interface for MAC address learning;

If other interfaces, an interface executing the selected reception interface is located within the balancing group according to the MAC address learning load balancing policy.

Step 403: The address entries found to match, then the network node determines whether the MAC address update condition;

The MAC address update condition comprising:

Condition A: matching address entry interface with said receiving interface in the same balancing group; Condition B: matching address table entry interface and the receiving interface of equalization in different groups and are not of the same type of interface; or

Condition C, matching address table entry interface and the receiving interface of equalization in different groups, but belonging to the same type of interface, and the network node support virtual machine migration.

Step 404: MAC address update when determining compliance with the conditions of the network node MAC address update.

MAC address update condition is satisfied A, the network node receiving the selected load balancing policy in accordance with the MAC address of the interface to perform an interface where the updated set of equalization; MAC address update satisfies the condition B, and C, the receiving interface of the network node MAC address update (that is, after receiving a packet of learned MAC addresses into effect, delete the original MAC address learning of interfaces).

Step 304: the network node forwards the packet to another interface is located outside the reception interface balancing group. And an interface configured balancing group in Example 2, the network node in the above embodiment of the following rules:

A, the same balancing group of packets received by the interface, the interface to the receiver no longer resides on another interface forwarding the balancing group, if a balancing group includes all uplink interface, from the uplink interface received broadcast / multicast or unknown unicast packets, does not forward the uplink interface to another.

B, as has been the same type of interface to receive the packet interface learned MAC address, and where the network node does not support virtual machine migration, the interface packet interface does not learn MAC addresses.

C, as the interface has the same balancing group received message interface learned MAC address, you can make the appropriate interface (such as less load Interface) based on balanced learning strategies.

E, have different types of interfaces learned MAC, MAC address after receiving the message interface to learn to take effect, delete the original MAC address of the interface to learn.

Due to the advance in accordance with the network architecture to prevent broadcast storms principle of a balanced set of configuration for the network node performs packet forwarding, based on a balanced set of preset receiving interface to the outside where the balancing group of forwarding packets to all other interfaces can effectively avoid the ring generating passage, i.e. to avoid broadcast storm, solutions of the embodiments of the present invention only needs to be set lower requirements of node devices, and easy to implement. In addition, the need for MAC address learning and aging, the same set of interfaces to achieve a balanced sharing of MAC address learning based on a balanced strategy to improve the performance of the system.

In Examples 1 and 2, an interface failure (e.g., chain scission), this can be transferred to the MAC table interface and other interfaces of the same balancing group failed interface above embodiment.

Solutions of the embodiments of the present invention is based, when a network node receives Ethernet packets according to the packet carries the MAC (Media Access Control, media access control) information, and to find the MAC MAC learning;

The node first has to find out whether the local MAC table information according to the MAC information of the MAC packet, to find out whether there are other interfaces where the balancing group, to find out whether the interface has been learned information in the MAC-balancing group;

If the location balancing group no other interface, the received packet interface based on the MAC learning rule MAC learning or updating; if located balancing group other interface configured strategies MAC learning or processing according to balancing group; Furthermore, different types of interfaces to learn the MAC MAC, the MAC after the entry into force of learning, and delete the original interface to learn.

In the network node MAC learning specific processes shown in Figure 5, the present process is embodied in the above rules, comprising the steps as follows:

Step 501: the network node receives an Ethernet packet;

Step 502: the node analyzes the received packet, and acquire the MAC address information of an interface received the packet;

Step 503: the node MAC query this node information table;

Step 504: The MAC table information queried results to determine whether there has been a corresponding address entry in the MAC packet; if not, continue processing; otherwise, go to step 508;

Step 505: MAC of the node table entry corresponding to the address is not the MAC address, whether the analysis is balanced set of packets received other interfaces where the interface; if not, execute step 506; otherwise, go to step 507;

Step 506: the interface of the node learning packet received packets carried in a MAC address, the MAC address entry is added, the process ends;

Step 507: Select interface MAC address learning packets balancing group within the receiving interface resides carried Load balancing policy; completing learning flow ends;

Step 508: Analysis and Comparison existing MAC address entry with the received interface information packets interfaces in the same balancing group; if yes, executing step 509; otherwise, go to Step 511; Step 509: determining whether the two interfaces interfaces belonging to the same type, if not the same type of interface, then go to step 511, otherwise, perform step 510;

Step 510: determining whether the network node belongs to technical support virtual machine migration, if supported, step 511 is executed, otherwise, the MAC information already exists due to the original interface, not learn a new interface, the flow ends;

Step 511: according to the selected load balancing policy to the reception interface updates the MAC address table in the balancing group resides in the interface information. In order to achieve the above-described methods, embodiments of the present invention further provides a network node, as shown in FIG. 6, the network node comprising:

Several interfaces;

Balanced configuration unit 11, a network architecture according to the principle in order to avoid broadcast storm of network nodes configured balancing group, a balancing group comprises one or more interfaces; and

Interfaces are within the balancing group for the upper layer of the upstream port or a network device interaction are used for the downlink interface interaction and the lower-layer network device.

Packet forwarding control module 12, upon receiving from the interface to receive packets for the reception interface to other interfaces where the balancing group other than forward packets.

Example 2 corresponds to the embodiment shown in Figure 7, the network node further comprises:

MAC address lookup unit 13 for the local MAC address table according to a MAC message carries information query;

MAC address management unit 14, an address for the matched entry is not found, a corresponding control interface MAC address learning, and when determining compliance with MAC address update conditions, the control interface corresponding MAC address update;

MAC address update judgment unit 15, configured to, when the matching entry found in the address determines whether the MAC address update conditions.

Alternatively, as shown in FIG. 8, the network node further comprises a Group Policy equalization control unit 16 for controlling the balancing group within the interface-based load balancing policy balancing group according to the equalization of the configuration unit 11 is disposed, if the MAC address management unit 14 judges the equalized reception interface group located other interfaces, if no other interface, the reception interface for controlling the MAC address learning; otherwise, control the equalization control unit 16 selects the group policy the interface performs the MAC address learning, and to learn a different interface to perform the MAC address, to achieve the effect of balancing the flow, the flow reaches the respective interfaces opposite balancing purposes.

As described above, the MAC address update condition comprising:

Condition A: matching address entry interface with said receiving interface in the same balancing group; Condition B: matching address table entry interface and the receiving interface of equalization in different groups and are not of the same type of interface; C or condition, the matching address table entry interface and the receiving interface of equalization in different groups, but belonging to the same type of interface, and the network node support virtual machine migration.

In particular, to meet the MAC address update condition A, the MAC address of the interface perform balancing group within the MAC address management unit controls the balancing group management control unit 16 selects the policy of the reception interface updates located; MAC address update satisfies the conditions B, and C, the MAC address management unit 14 controls the management of the reception interface MAC address update.

As shown, the network node further comprises a set of fault equalization processing unit 17, for detecting faulty balancing group according to the group equalizer equalizing configuration unit 11 is disposed in the interface, and the interface of the balancing group 8 transfer management unit 14 notifies the MAC address of the failed interface MAC address of the receiving interface to other interfaces where the balancing group. Significant differentiation occurs when the flow rate of the balancing group for each interface, by the control unit 16 Policy Group equalization process.

Compared with the related art, embodiments of the present invention introduces the uplink interface, and the downlink interface balancing group, the network node that the complete optimization learning and forwarding processing for the MAC packet, so that, on the one hand to ensure the efficient high-speed packet processing, another aspect of the present invention embodiments are not required in the network switching device unified operation and control protocols to exchange control packets, retains the characteristics of the second floor, the configuration reduces the complexity and improves the robustness of the system, improving the system design flexibility. The following examples in conjunction with the application, the method embodiments of the present invention will be described. Application Example 1

Figure 9 shows an application example, embodiments of the present invention are mainly described in the standard fat tree architecture, the packet forwarding process in detail and the MAC learning;

9, which is an embodiment of the present invention under standard forwarding fat tree (Fat-Tree) and the structure of FIG MAC learning process, and its MAC learning process is as follows:

Step 901: acl node from which it receives direct ARP request message sent by the server, receiving the packet interface D-P1 (a first downstream interface Interface, hereinafter similar);

Step 902: source MAC address destination MAC information and analyzing the information received acl node carried in the packet, and the packet-receiving interface number D-P1; 903 Step: acl node information learned source MAC address information and the interface D-P1, the write MAC table; this application example, D-P1 learned the MAC address of the interface;

Step 904: Node-balancing group acl interfaces D-P1 where no other interfaces, acl node forwards the ARP request packet to all the other interfaces (U-P1 and U-P2);

Step 905: agl ag2 node and the node receives a message from the node via its interface acl D-P1;

Step 906: source MAC address destination MAC information and analyze the received information agl node carried in the packet, and the packet-receiving interface number D-P1;

Step 907: agl node learning source MAC address information and the interface D-P1 information, the MAC address table is written; this application example, D-P1 learned the MAC address of the interface;

Step 908: ag2 node according to the same process of learning the ARP message carried in a MAC address on its interface to D-P1;

Step 909: Because balancing group D-P1 interfaces where there is no other interface, all other interfaces (D-P2, U-P1 and U-P2) forwards the packet to the node agl. cl, c2 and ac2 forwarding node receives the ARP request packet Gen 4;

Step 910: ac2 node receives the ARP request packet later analyzed MAC information and interface information received packets;

Step 911: Because the U-P1 ac2 node interface as an uplink interface; the MAC learning process learning MAC process in accordance with the network node, the interface learned by the U-P1 In this application example;

Step 912: Because U-P2 belong to the same U-P1 with a balanced group, ac2 node does not forward the uplink interfaces to other U-P2.

Understandably, as well as ac2 node downstream interface then also forwarded to the downlink interface to "^ text.

Step 913: cl node D-P1 interface receives the ARP request packet, the interface is due to the balancing group where no other interfaces, and the learning process is acl ag l nodes similar omitted; Step 914: Repeat the previous steps, ARP request packet reaches the destination server B, each network node completes learning of MAC source server a; a response to the request destination server B, thereby forming a floor between the forwarding path AB. Application Example 2

Figure 10 shows the application example 2, the main embodiment of the invention described in non-standard fat tree structure, a process was described in detail Gen MAC forwarding and learning;

As shown in FIG. 10, this embodiment is forwarded in non-standard fat tree structure and MAC learning process of the present invention, FIG embodiment, and its MAC learning process is as follows:

Step 1001: AC1 directly connected node receiving the gratuitous ARP message server VM1 virtual machine from which it is transmitted, the packet receiving interface D-P1;

Step 1002: AC1 MAC source node analyzes the acquired received information carried in the packet, and the packet-receiving interface number D-P1;

Step 1003: AC1 node learning source MAC address information and the interface D-P1 information, the MAC address table is written; this application example, D-P1 learned the MAC address of the interface;

Step 1004: the node balancing group AC1 interfaces D-P1 where no other interfaces, AC1 free node forwards the ARP packets to all other interfaces (U-P1, U-P2 and U-P3);

Step 1005: AGG1 node, AGG 2 node and the node respectively AGG 3 receives a message from the node via its interface AC1 D-P1;

Step 1006: AGG1 MAC source node analyzes the acquired received information carried in the packet, and the packet-receiving interface number D-P1;

Step 1007: AGG1 node learning source MAC address information and the interface D-P1 information, the MAC address table is written; this application example, D-P1 learned the MAC address of the interface;

Step 1008: AGG2 AGG3 node and the node learning process in the same MAC address on its interface to D-P1;

Step 1009: Since balancing group AGG1 node interfaces D-P1 where no other interfaces, the interfaces to all the nodes AGG1 (D-P2, D-P3, U-P1 and U-P2) in addition to D-P1 forwarding interface message. Cl, C2, AC2 and AC3 node receiving the gratuitous ARP packet is forwarded;

Step 1010: After AC2 node receiving the gratuitous ARP packet, analyzes its MAC information and interface information received packets; step 1011: AC2 node MAC learning processing according to the network node MAC learning process in the present application example, AC2 node the U-P2 learn the MAC address, U-P1 interface is not the MAC address learning.

Step 1012: Since the U-P2, U-P3 belong to the same U-P1 and a balancing group, the same as U-P2 and U-P3 does not participate in the forwarding of the uplink interface.

Step 1013: the node D-P1 C1 interface receives a gratuitous ARP packet, since the interface is configured with a balanced group. Let the balancing group according to a group policy interface equilibrium study carried ARP packet address, in the present embodiment, the D-P1 still learn the MAC address;

Step 1014: Since the node C1 is configured with a plurality of different equalization groups, C1 node forwards the ARP packets received to an interface other than the group located equalization ARP packet to all interfaces. In this application example, it is necessary to forward the U-P4, U-P5 and U-P6 interfaces;

Step 1015: AGG4, AGG5 AGG6 and forwarded by a node receives a free node C1

ARP packets, and analyzes the acquired information and receiving interface information carried in packets; and sequentially processed in the preceding steps forward;

Step 1016: free ARP "^ Burgundy forwarded to the entire network all network nodes, each network node to complete the learning of MAC on the source server virtual machine VM1 is, so that each node has a forwarding destination MAC for the MAC address of VM1 MAC forwarding entry.

Application Example 3

Figure 11 shows the application example 3, the main embodiment of the invention described in the multicast service, the packet forwarding process in detail and the MAC learning;

11, this embodiment of the invention is a contribution in one multicast forwarding situation and the MAC learning process diagram in which a router (Router) network node is a multicast router control, and its MAC learning process is as follows:

Step 1101: Router node to all of its interfaces transmitting multicast query message; the message is a broadcast message;

Step 1102: After cl, c2, c3 and c4 network node receives a multicast query packet, in accordance with the MAC learning process learning the MAC; cl network node to the following scheme as an example and described; Step 1103: receiving the multicast query cl packet interface U-P1, the balancing group U-P1 interface resides no other interfaces, then cl multicast request packet to the U-P1 except for the interface to forward all interfaces;

Step 1104: agl, ag3 and ag5 network node receives cl node forwards multicast query messages; agl network node as an example to describe;

Step 1105: agl network node obtains the received multicast query packet information and packet interface for receiving U-P1;

Step 1106: agl network nodes The network node MAC MAC learning process learning process, in the present embodiment, learned by the interface U-P1; forwarded to all interfaces other than the U-P1-balancing group is located;

Step 1107: acl ac2 and the network node receives packets forwarded by the network node agl; acl network node to be described as an example;

Step 1108: acl MAC learning and forwarding network node processes and agl similar process step similar to step 1105- 1106, will not be described;

Step 1109: A server receives multicast from the network node acl query message, the server A in response to the multicast query packet; thus forming a multicast branch path; other multicast branch path forming process with such the same, not be described;

Step 1110: When the server responds to all finished, the multicast path is formed.

Compared to the related art, embodiments of the present invention is a method and a network node in advance by a network architecture principles to avoid broadcast storm of network nodes configured to balancing group, for packet forwarding, where the receiving interface of the balancing group according to a preset All other interface to forward packets other than the balancing group can effectively avoid broadcast storm, solutions of the embodiments of the present invention only needs to be set lower requirements of node devices, and easy to implement. In addition, the need for MAC address learning and aging, the same set of interfaces to achieve a balanced sharing of MAC address learning based on a balanced strategy to improve the performance of the system.

Those of ordinary skill in the art will be appreciated that the above method may be all or part of the steps by a program instructing relevant hardware, the program may be stored in a computer-readable storage medium, such as a read-only memory, magnetic or optical disk. Alternatively, all or part of the steps of the above-described embodiments may be implemented using one or more integrated circuits. Accordingly, each module / unit in the above-described embodiments may be implemented preclude the form of hardware, and also can be implemented in the form of software function module. The present invention is not limited to any specific combination of hardware and software form.

The system described in the embodiments of the present invention / apparatus / device module only an example will be divided according to their function, it is understood, in the case of the system / device / apparatus performing the same function, those skilled in the art may be given one or more other functional division manner, wherein when the particular application may be any one or more functional modules preclude functional entity implemented using a device or unit, undeniable, the above conversion method are within the scope of protection of the present application Inside.

Obviously, the described embodiments are merely part of embodiments of the present invention, not all the embodiments of Fig. Based on the embodiments of the present invention, all other embodiments of ordinary skill in the art without creative efforts shall fall within the scope of the present invention.

Industrial Applicability

A method and a network node embodiment of the present invention varies according to the network architecture in advance to avoid broadcast storm principle configuration balanced group of network nodes, for packet forwarding, all the balancing group according to a preset equalization than group where the other reception interface interface to forward packets to avoid broadcast storm, solutions of the embodiments of the present invention only needs to be set lower requirements of node devices, and easy to implement. In addition, the need for MAC address learning and aging, the same set of interfaces to achieve a balanced sharing of MAC address learning based on a balanced strategy to improve the performance of the system.

Claims

Claims
A Layer 2 network data transfer method, comprising:
The network architecture principles to avoid broadcast storm is configured as a balanced group of network nodes, a balancing group comprising one or more interfaces;
^ Gen network node receives packets from the receiving interface; and
The network node forwards the packet to the interface is located outside the reception interface balancing group.
2. The method as claimed in claim 1, wherein: the interface within the group were equalized for the upper layer uplink interface device to interact with a network interface interaction, or both for the downlink and the lower-layer network device.
3. The method according to claim 1, wherein: the network node receives the packet, the method further comprising:
MAC control information of the network node queries the local media access according to the MAC address table packet carried;
If no match is found in the address entry, then the network node MAC address learning; found as matching address entry, then the network node determines whether the MAC address update condition; and
MAC address update when determining compliance with the conditions of the network node MAC address update.
4. The method as claimed in claim 3, wherein: the node MAC address learning network, comprising: determining whether the balancing group receiving other interfaces where the interface;
If no other interface, the reception interface for learning a MAC address; and
If other interfaces, an interface executing the selected reception interface is located within the balancing group according to the MAC address learning load balancing policy.
5. The method as claimed in claim 4, wherein: the MAC address update conditions include: matching address entry interface with said receiving interface in the same balancing group;
Matching address table entry interface and the receiving interface of equalization in different groups and are not of the same type of interface; or matching address table entry interface and the receiving interface of equalization in different groups, but the same type of the interface and the network node support virtual machine migration.
6. The method as claimed in claim 5, wherein,
Interface performs the MAC address in the MAC address update meet the conditions for matching address entry interface when said interface is to receive in the same balancing group, the network node where the receiving interface selected load balancing policy balancing group according to the updated ; as well as
Meet the conditions for matching MAC address update address table entry interface and the receiving interface of equalization in different groups and are not of the same type of interface; or an interface address of the matching entry is located in a different reception interface balancing group , but belonging to the same type of interface, and the network node support virtual machine migration, performing MAC address of the receiving interface of the network node update.
7. The method as claimed in claim 1, wherein: when a network node interface fails, the network node according to a load balancing policy to the MAC address table transfer to the interface of the failed interface of the failed interface resides balancing group.
8. A network node, comprising:
Several interfaces;
Balanced configuration unit, which is arranged according to the network architecture principles to avoid broadcast storm is configured as a balanced group of network nodes, a balancing group comprising one or more interfaces; and
Packet forwarding control module, when receiving from the interface to receive packets, which is arranged to the receive interface is located outside the interface to forward packets balancing group. 9. The network node according to claim 8, wherein: the interface within the group were equalized for the upper layer uplink interface device to interact with a network interface interaction, or both for the downlink and the lower-layer network device.
10. The network node as claimed in claim 8, wherein: said network node further comprising: a Media Access Control MAC address lookup means arranged to local MAC address table based on a MAC message carries information query;
MAC address management unit, which is arranged at the matching address entry is not found, a corresponding control interface MAC address learning and updating the MAC address determined in line with the conditions, the corresponding control interface MAC address update; and MAC address update determination unit, which is arranged at the address found in the matching table entry, it is determined whether the MAC address update conditions.
11. The network node according to claim 10, wherein: the network node further comprises a balancing group policy control means arranged to load balancing policy balancing group within the interface of the balancing group according to the equalization configuration unit configured control, whether the balancing group MAC address management unit judges that the other reception interface where the interface, if no other interface, the reception interface for controlling the MAC address learning; if any other interfaces, the equalization control group policy control unit selects the interface to perform MAC address learning.
12. The network node according to claim 11, wherein: said MAC address update condition comprising:
Matching address entry interface with said receiving interface in the same balancing group;
Interface and the address table entry matches the receiving interface is in a different balancing group and is not the same type of interface; or
Matching address table entry interface and the receiving interface of equalization in different groups, but belonging to the same type of interface, and the network node support virtual machine migration.
13. The network node as claimed in claim 12, wherein,
Meet the conditions for matching MAC address update address entry interface when said interface is to receive in the same balancing group, the MAC address management unit controls the balancing group management policy control means selects said reception interface located equalization interface performs the MAC address updates in the group; and a MAC address update meet the condition of the interface address table matching reception interfaces are located in different balancing group and is not the same type of interface; or matching address entry interface and said receiving interface equalization in different groups, but belonging to the same type of interface, and the network node support virtual machine migration, managing the MAC address management unit controls the reception interface MAC address update.
14. The network node as claimed in claim 8, wherein: the network node further comprises a set of fault equalization processing unit arranged to detect balancing group according to the equalization balancing group configuration unit arranged in the interface, and the equalization group transfer interface when a failure occurs, the MAC address management unit notifies the fault interface MAC address table of said reception interface to other interfaces where the balancing group.
PCT/CN2013/089936 2012-12-26 2013-12-19 Data transmission method and network node in layer 2 network WO2014101708A1 (en)

Priority Applications (2)

Application Number Priority Date Filing Date Title
CN201210574336.X 2012-12-26
CN201210574336.XA CN103905325B (en) 2012-12-26 2012-12-26 Layer network data transfer method and a network node

Publications (1)

Publication Number Publication Date
WO2014101708A1 true WO2014101708A1 (en) 2014-07-03

Family

ID=50996480

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2013/089936 WO2014101708A1 (en) 2012-12-26 2013-12-19 Data transmission method and network node in layer 2 network

Country Status (2)

Country Link
CN (1) CN103905325B (en)
WO (1) WO2014101708A1 (en)

Families Citing this family (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104936253A (en) * 2015-06-26 2015-09-23 濮斌 Hardware architecture of tree topology based on 802.11 and protocol processing
CN105721314A (en) * 2016-02-05 2016-06-29 华为技术有限公司 Method and equipment for sending data packet

Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101155109A (en) * 2006-09-30 2008-04-02 华为技术有限公司 Ethernet switching system and equipment
CN101330370A (en) * 2007-05-17 2008-12-24 日本电气株式会社 Node, communication method, and program for node
US20110206058A1 (en) * 2010-02-24 2011-08-25 Cisco Technology, Inc., A Corporation Of California Automatic Determination of Groupings of Communications Interfaces

Family Cites Families (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2005104494A2 (en) * 2004-04-16 2005-11-03 Cesura, Inc Distributed computing environment and methods for managing and controlling the same
CN102355417A (en) * 2011-10-08 2012-02-15 杭州华三通信技术有限公司 Data center two-layer interconnection method and device

Patent Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101155109A (en) * 2006-09-30 2008-04-02 华为技术有限公司 Ethernet switching system and equipment
CN101330370A (en) * 2007-05-17 2008-12-24 日本电气株式会社 Node, communication method, and program for node
US20110206058A1 (en) * 2010-02-24 2011-08-25 Cisco Technology, Inc., A Corporation Of California Automatic Determination of Groupings of Communications Interfaces

Also Published As

Publication number Publication date
CN103905325B (en) 2018-12-11
CN103905325A (en) 2014-07-02

Similar Documents

Publication Publication Date Title
US8442064B2 (en) Virtual link aggregation of network traffic in an aggregation switch
US8995272B2 (en) Link aggregation in software-defined networks
EP2829031B1 (en) Virtual router terminating an overlay tunnel in a storage area network
US6115378A (en) Multi-layer distributed network element
US9025468B1 (en) Custom routing decisions
US8913489B2 (en) System and method for virtual fabric link failure recovery
US8873551B2 (en) Multi-destination forwarding in network clouds which include emulated switches
EP2264949B1 (en) Forwarding frames in a computer network using shortest path bridging
US9544224B2 (en) Virtual port channel bounce in overlay network
US9270572B2 (en) Layer-3 support in TRILL networks
US20150117256A1 (en) Extended ethernet fabric switches
US8767735B2 (en) System and method for multi-chassis link aggregation
US8166187B2 (en) Distributed IP gateway based on sharing a MAC address and IP address concurrently between a first network switching device and a second network switching device
US8948181B2 (en) System and method for optimizing next-hop table space in a dual-homed network environment
US20080068985A1 (en) Network redundancy method and middle switch apparatus
US8472447B2 (en) IP multicast snooping and routing with multi-chassis link aggregation
US8509087B2 (en) Per-graph link cost assignment in layer 2 multipath networks
US20120163164A1 (en) Method and system for remote load balancing in high-availability networks
US10097372B2 (en) Method for resource optimized network virtualization overlay transport in virtualized data center environments
US8767738B2 (en) MAC learning in a TRILL network
US8654680B2 (en) Packet forwarding using multiple stacked chassis
CN102415065B (en) Method and apparatus redundant routing network connected to the host
CN101548511B (en) System for providing both traditional traffic engineering enabled services and traffic engineering enabled services
EP2462723B1 (en) Providing a logical aggregated point-to-point data link incorporating a multi-point link
US8134922B2 (en) Reducing flooding in a bridged network

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 13869792

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase in:

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 13869792

Country of ref document: EP

Kind code of ref document: A1