CN101188564B - A Realization Method of Multiple Spanning Tree Protocol - Google Patents

Abstract

The invention discloses a method for realizing multiple spanning tree protocol, the method includes: enabling multiple spanning tree protocol MSTP and multiple spanning tree protocol iMSTP based on QINQ or Layer 2 VPN on a switch; setting spanning tree sub-instances to obtain Spanning tree sub-instance topology; the egress switch outputs user data forwarded through the spanning tree sub-instance topology. In order to obtain the topology of the sub-instance of the spanning tree, the iMSTP mode must first be judged according to the iMSTP ID. The method of the invention not only solves the problem that the existing MSTP protocol does not support the QINQ protocol and the two-layer VPN protocol, but also can distinguish different VLAN services under the two-layer protocol, and enhances the scalability of the VLAN service.

Description

A Realization Method of Multiple Spanning Tree Protocol

technical field

The invention relates to the field of data communication, in particular to a method for realizing multiple spanning tree protocols.

Background technique

From the moment STP (Spanning-Tree Protocol, Spanning Tree Protocol) was born, it has determined its characteristics as a "small" protocol and a "big" protocol. The so-called "size" here refers to the protocol itself, but the spanning tree protocol is much more important than any routing protocol and MPLS (Multiprotocol Label Switching) protocol.

The spanning tree protocol has gone through three stages in its historical development: the first generation spanning tree protocol: SSTP/RSTP (simple spanning tree protocol/rapid spanning tree protocol); the second generation spanning tree protocol: PVST/PVST+ (VLAN (virtual LAN) Spanning Tree Protocol); the third generation of Spanning Tree Protocol: MSTP (Multi Spanning-Tree Protocol-Multiple Spanning Tree Protocol).

Although the first-generation spanning tree protocol is simple and fast, but today the IEEE 802.1Q protocol is popular, because the protocol mainly involves the realization of VLAN, SSTP or RSTP cannot distinguish VLAN instances, nor can they operate on services belonging to different VLANs. The first generation spanning tree protocol can no longer meet the requirements of the IEEE 802.1Q protocol.

In order to solve the situation of being unable to do anything to VLAN services, the second-generation spanning tree protocol came into being. However, with this spanning tree protocol, each VLAN must maintain a spanning tree instance, so that the number of spanning tree instances must correspond to the VLAN one-to-one. Lack of flexibility, and when there are many VLANs, the number of spanning tree instances that need to be maintained will increase accordingly, which will bring a great burden to the switch and greatly affect the performance of the switch.

The third-generation spanning tree protocol solves the problems faced by the previous two generations of protocols. It is a multi-spanning tree instance that can distinguish VLANs and make multiple combinations of VLANs. Each spanning tree instance represents a series of VLAN combinations. In this way, MSTP can well solve all the problems of the 802.1Q protocol.

With the increasingly IP-based trend in the data communication field, it is necessary to support different VLAN services. Under this trend, the QINQ (or SVLAN, VLAN stacking) protocol and the Layer 2 VPN (Virtual Private Network) protocol were born. The idea of using Layer 2 protocol to distinguish different services will inevitably require corresponding countermeasures from Spanning Tree Protocol, but the development of Spanning Tree Protocol to MSTP protocol has stagnated. Judging from the characteristics of current data communication, the existing MSTP protocol can only be distinguished according to the VLAN ID of the switch port participating in the multi-spanning tree calculation, and the spanning tree topology calculation is performed according to the mapping between different multi-spanning tree instances and VLAN IDs. The double-VLAN tagging protocol or multi-layer VLAN tagging protocol and Layer 2 VPN protocol involved in the QINQ protocol cannot support operations that need to distinguish between different VLAN services, and the scalability of VLAN services is poor, which does not provide the necessary space for future development.

Contents of the invention

The purpose of the present invention is to propose a method for implementing the Multiple Spanning Tree Protocol, so as to overcome the insufficient support of the MSTP protocol for the QINQ protocol and the Layer 2 VPN protocol in the prior art.

In order to realize the above-mentioned purpose of the invention, the present invention is specifically realized like this:

A method for implementing multiple spanning tree protocols, comprising the steps of:

Step 1, the data message that needs to be forwarded by the switch enters the switch, the switch enables the iMSTP multiple spanning tree protocol, adds the iMSTP protocol field in the protocol message, and sets the iMST spanning tree sub-instance according to the setting of the iMSTP protocol field;

Step 2, perform spanning tree sub-instance topology calculation through the fields added in the protocol message;

In step 3, the switch forwards and outputs the data packets forwarded through the spanning tree sub-instance topology.

Wherein said step 2, calculating spanning tree sub-instance topology includes the following steps:

Step 2a, judge the iMST mode;

Step 2b, conduct regional root election;

In step 2c, path cost calculation is performed.

Wherein in said step 2a, carry out iMSTP pattern judgment according to the iMSTP ID that is set in the iMSTP agreement field of described setting;

The value of the iMSTP ID is determined by the VLAN networking mode;

If the QINQ network is used, the iMSTP ID value is the VLAN ID of the inner VLAN tag of QINQ;

If Layer 2 VPN is used for networking, the value of iMSTP ID is the trunk value configured on the edge interface of Layer 2 VPN or the VCID value assigned to different services in Layer 2 VPN.

In the implementation method of the multiple spanning tree protocol, in the step 1, before the switch opens the iMSTP multiple spanning tree protocol, the MSTP multiple spanning tree protocol is first opened.

In step 1, if no iMST spanning tree sub-instance is set, MST multiple spanning tree instance topology calculation is performed, and the switch forwards the output data message according to the multiple spanning tree instance topology.

Adopting the realization method of the Multiple Spanning Tree Protocol of the present invention not only solves the problem that the existing MSTP protocol supports insufficient QINQ protocol and Layer 2 VPN protocol, but also can distinguish different VLAN services under the Layer 2 protocol, and also enhances the VLAN service scalability.

Description of drawings

The accompanying drawings described here are used to provide a further understanding of the present invention and constitute a part of the application. The schematic embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute improper limitations to the present invention.

Fig. 1 is the structural block diagram of iMST under the QINQ pattern among the present invention;

Fig. 2 is the structural block diagram of iMST under VPLS VLAN mode among the present invention;

Fig. 3 is the structural block diagram of iMST under VPLS VCID mode among the present invention;

Fig. 4 is the processing flowchart of the specific embodiment of the present invention;

Fig. 5 is a system structure diagram for performing a function verification test on the IP network equipment system formed by the method of the present invention.

Fig. 6 is a schematic diagram of the structure of the BPDU message of the existing MSTP protocol modified by the method of the present invention.

Detailed ways

Method of the present invention mainly comprises the steps:

Step 1, open the Multiple Spanning Tree Protocol MSTP on the switch, and the Multiple Spanning Tree Protocol iMSTP based on QINQ or Layer 2 VPN (the MSTP protocol with a spanning tree sub-instance); Data packets are forwarded directly according to the generated multiple spanning tree (MST) topology.

Step 2, configure the spanning tree sub-instance to obtain the topology of the spanning tree sub-instance;

When configuring the spanning tree sub-instance, the judgment of the iMSTP mode should be carried out first. The iMSTP mode is identified according to the iMSTP ID. Due to the different VLAN networking methods, the values of the iMSTP ID are also different.

If QINQ is used for networking, you need to set the iMSTP mode to QINQ mode, and specify the VLAN ID of the inner VLAN tag of QINQ as the iMSTP ID;

If the Layer 2 VPN is used for networking, it is necessary to distinguish between the trunk VLAN (the Layer 2 VPN edge interface is configured as a trunk VLAN) method and the VCID method:

If the trunk VLAN mode is used, assign the trunk value configured on the Layer 2 VPN edge interface to the iMSTP ID;

If the VCID method is adopted, assign the VCID value assigned to different services in each Layer 2 VPN to the iMSTP ID;

Step 3: The data packets that need to be forwarded by the switch enter the switch and are forwarded according to the topology of the spanning tree sub-instances. The topology status of each spanning tree sub-instance can be queried through special commands; the topology of different spanning tree sub-instances should be different.

In step 4, the egress switch outputs the user data forwarded through the spanning tree sub-instance topology.

Data packets belonging to different services can be forwarded according to different spanning tree sub-instance topologies, and then the correctness and integrity of the packets can be verified through packet analysis.

Specific embodiments of the present invention will be further described in detail below in conjunction with the accompanying drawings.

As shown in Figure 4, the iMSTP processing flow in the present invention includes the following steps:

Step 401, the switch receives the user data message;

Step 402, the switch judges whether to open the MSTP multiple spanning tree protocol, if so, then proceed to step 403, otherwise, proceed to step 408;

Step 403, the switch judges whether to open the iMSTP multiple spanning tree protocol, if so, then proceed to step 404, otherwise, proceed to step 406;

In the step 402 or step 403, enabling the Multiple Spanning Tree Protocol (MSTP or iMSTP) is implemented in such a way that even if the multiple spanning tree instance on the switch is enabled, in the switch global configuration mode, an enabling command (for example: enable spanning-tree); then execute this command to specify the spanning tree mode (MSTP or iMSTP) (for example: spanning-tree mode MSTP).

Step 404, increase the iMSTP protocol field in the iMSTP protocol message, according to the setting of the iMSTP protocol field, the switch judges whether to set the iMST spanning tree sub-instance, if so, then proceed to step 405, otherwise, proceed to step 406;

The setting iMST sub-instance refers to increasing the spanning tree type iMST in the switching network, iMST and MST, SST and RST are independent of each other, and then should increase the mode supported by iMST in the switch, i.e. the iMSTP mode, however, iMSTP and MSTP configurations (i.e. topologies) are independent of each other and can co-exist.

Step 405, enter the iMST multi-spanning tree negotiation mode, calculate the spanning tree sub-instance topology, and then turn to step 407;

The computing spanning tree sub-instance topology includes the following steps:

Step 5a, judge iMSTP mode;

Step 5b, perform regional root election;

In step 5c, path cost calculation is performed.

Step 406, enter the MST multi-spanning tree negotiation mode, and calculate the spanning tree instance topology;

The computing spanning tree topology includes the following steps:

Step 6a, perform regional root election;

In step 6b, path cost calculation is performed.

Step 407, the switch protocol layer notifies the driver layer to generate forwarding channels according to the calculated topology;

Step 408, the switch forwards the user data packets after the topology.

As shown in Figure 6, taking the QINQ protocol as an example, in order to provide iMSTP protocol support, for the QINQ interface, the BPDU (Bridge Protocol Data Unit) message of the existing MSTP protocol is transformed, and an iMSTP (Inside MSTP) message with a length of 18 bits is added. ) protocol field, which is embedded in the original MSTP protocol field, and identifies the iMST sub-instance in the MST instance to which the QINQ interface belongs. Spanning tree instance number) is specified as the iMSTP ID, and other parts of the iMSTP ID protocol field are the same as the existing MSTP ID field.

Before configuring the above iMST sub-instance, first add the spanning tree type iMST in the QINQ switching network. are independent of each other; then configure the added spanning tree sub-instance (the configuration of the added iMST spanning tree sub-instance is independent of the configuration of the existing MST spanning tree instance, but the two can coexist).

The iMST is called a "sub-instance", which is represented by an instance number as x.y, where x is the instance number and y is the sub-instance number. x represents the instance number of the MST when the spanning tree sub-instance is not set, which can correspond to a set of VLANs in a series (under QINQ conditions, that is, the set of outer VLAN tags); y represents the set of spanning tree In the case of a sub-instance, the iMST sub-instance number corresponds to the set of inner private network VLANs (under the condition of QINQ, that is, the set of inner VLAN tags).

For example, under the condition of QINQ, the outer VLAN is marked as VLAN10, and the inner VLAN is marked as VLAN20. It can be set that MST instance 1 corresponds to the outer VLAN tag VLAN10, and iMST instance 1.1 corresponds to the inner VLAN tag VLAN20. Then when iMST If enabled, when data packets interact, the topological environment will be determined by the inner VLAN20 instead of the outer VLAN10. Different iMST sub-instances can be linked to the same MST instance, as shown in Figure 1.

The iSMT sub-instance must be attached to the MST instance, but if you don’t know what the value of the public network VLAN ID is, or you don’t need to calculate the spanning tree according to the public network VLAN tag, you can set the MST instance 1 VLAN ID1-4094 (for the QINQ protocol, with In the data packet of the VLAN tag, there is a 12-bit VLAN ID field, which is used to store the VLAN ID value, and the value space of this value is from 1 to 4094). All iMST sub-instances are bound to MST instance 1, That is 1.y.

For the Layer 2 VPN protocol, take VPLS (Virtual Private Local Area Network) protocol as an example. In order to provide iMSTP protocol support, when configuring iMST, you should first solve a difficulty in iMST’s support for VPLS, that is, the inner VLAN tag of the private network. uncertainty. The so-called uncertainty of the inner label of the private network is aimed at switches (specifically, high-end routing switches with MPLS functions).

To solve the above difficulties, you should first obtain the encapsulation mode of the PE ingress port (the encapsulation mode of the PE ingress port determines the implementation of Layer 2 VPN memory labels). In the VPLS network, there are two types of xconnect interfaces on the PE side. Encapsulation method: access VLAN and trunk VLAN.

If the xconnect interface on the PE side has been configured in the trunk mode, due to the addition of the VPLS VLAN network in the iMST configuration mode, according to the characteristics of the VPLS packet, the VLAN tag encapsulated in the MPLS tag is used for spanning tree sub-instance matching. This method is similar to the QINQ method. iMSTP fills the iMSTP ID field in the BPDU message of the switch with the iSMT sub-instance matching the trunk encapsulation value of the xconnect interface on the PE side, and configures the P routing switches one by one. Before MPLS label distribution, After the spanning tree sub-instance topology is calculated, as shown in Figure 2, the rest can be implemented using the QINQ method.

If the xconnect interface on the PE side is configured as the access mode, for the access VLAN, the inner label of the private network will not carry the 802.1Q label, which will make it difficult to distinguish the private network. The method is randomly allocated from 16 to 1048575, and it is difficult to know in advance how the VPLS inner layer marks are allocated.

In order to solve the above problems, VCID configuration is performed on the xconnect interface on the PE side, and the VCID values assigned to different services in each Layer 2 VPN are assigned to the iMSTP ID, that is, the VPLS VCID function in the iMSTP configuration mode is added. The business is distinguished according to its own VCID. iMSTP configures the iMST sub-instance with the private network VPN VCID value and fills it into the iMSTPID field in the switch BPDU message, and configures it one by one on the P routing switch to calculate and establish a new VCID-based Topology, on this basis, perform TCP message handshake and establish MPLS label binding, as shown in Figure 3, and the rest can be implemented in the QINQ mode.

To sum up, for the Layer 2 VPN protocol, when calculating the topology of the spanning tree sub-instance, it is only necessary to determine the one-to-one correspondence between the inner label value and the iMSTP ID. If the memory label of the Layer 2 VPN is determined (ie One input parameter), then the solution of QINQ memory tag (that is, one input parameter) can be applied, and this input parameter is substituted into the topology calculation method to obtain the spanning tree sub-instance topology.

Fig. 5 is the system structural diagram that the present invention carries out functional verification test to IP network equipment, uses two ports of tester to be connected with two user access ports of tested equipment respectively, is denoted as port 1 and port 2 respectively, four The devices under test form a ring network.

If the iMSTP mode of the device under test adopts the QINQ mode networking, the specific test steps include:

Step 1. Configure all interconnection interfaces of the device under test as QINQ interfaces, and the two devices under test connected to the tester function as PEs, converting packets with a single 802.1Q label on the user side into packets with dual 802.1Q labels on the public network ;

Step 2, configure iMSTP on all devices under test, map iMSTP sub-instances for different VLANs on the user side, set the iMSTP mode to QINQ, and observe the establishment of the sub-spanning tree (iMSTP) after completing the iMSTP configuration;

Step 3: Ports 1 and 2 of the tester send user packets with various 802.1Q tags to each other, and then observe the flow of traffic on each device under test. You can try to disconnect some of the links to check whether the traffic will be interrupted.

If the iMSTP mode of the device under test adopts Layer 2 VPN mode networking, the specific test steps include:

Step 1, configure the VFI user interface on the device directly connected to the tester, and map the VCID with the iMSTP sub-instance; set the iMSTP mode to VPLS VCID; after completing the iMSTP configuration, you can observe the establishment of the sub-spanning tree (iMST);

Step 2, configure VPLS on the device, you can observe whether the VPLS label is assigned successfully and whether the PW pseudowire is successfully established;

Step 3: Port 1 and port 2 of the tester send various VFI user messages to each other, and then observe the flow of traffic on each device under test; you can try to disconnect some of the links to check whether the traffic will be interrupted;

Step 4, change the iMSTP mode to VPLS VLAN, repeat steps 2 and 3.

The IP network device under test needs to have the following functions, VLAN-related functions, QINQ functions, routing functions, and Layer 2 VPN functions;

The VLAN-related functions include user VLAN isolation and 802.1Q encapsulation. The QINQ function includes supporting Layer 2 data forwarding with double Q labels. The routing function includes generating routes through configuration and establishing L2VPN pseudowires. The Layer 2 VPN function includes supporting functions such as VPLS and VPWS (Virtual Private Wire Service, Virtual Private Wire Service).

The tester includes: configuration module: perform corresponding configuration according to the interface type of the docking device, and configure the traffic of the whole machine test according to the test requirements; sending module: send the configured test traffic; receive module: receive the device under test Send traffic and perform statistical calculations, etc.

The above-described embodiments are two preferred modes of the present invention, and any common changes and substitutions made by those skilled in the art within the scope of the technical solutions of the present invention should be included in the protection scope of the present invention.

Claims (4)

1. the implementation method of a Multiple Spanning Tree Protocol is characterized in that, comprises the steps:

Step 1, need to enter switch through the data message that switch is transmitted, switch is opened and is had the Multiple Spanning Tree Protocol iMSTP that generates the tree example, in protocol massages, increase the iMSTP protocol fields, and according to being provided with of iMSTP protocol fields obtained the setting that iMST generates the tree example, be provided with iMSTP ID in the described iMSTP protocol fields, the value of described iMSTP ID is to be determined by the VLAN networking mode, if adopt the networking of QINQ mode, then iMSTP ID value is the VLAN ID of the inner VLAN label of QINQ; If adopt the networking of two-layer VPN mode, then iMSTP ID value is to VCID value that different business distributed in the trunk value that disposed of two-layer VPN edge interface or the two-layer VPN;

Step 2 is carried out the iMSTP pattern according to the iMSTP ID that is provided with in the described iMSTP protocol fields and is judged, generates tree example topology and calculates;

Step 3, switch are transmitted output through generating the data message that tree example topology is transmitted.

2. the implementation method of Multiple Spanning Tree Protocol as claimed in claim 1 is characterized in that, in the described step 2, calculates and generates tree example topology, comprises the steps:

Step 2b carries out regional root election;

Step 2c carries out path cost and calculates.

3. the implementation method of Multiple Spanning Tree Protocol as claimed in claim 1 is characterized in that, in the described step 1,

Before carrying out switch unlatching iMSTP Multiple Spanning Tree Protocol, carry out the unlatching of MSTP Multiple Spanning Tree Protocol earlier.

4. the implementation method of Multiple Spanning Tree Protocol as claimed in claim 3 is characterized in that, in the described step 1,

Do not generate the tree example if iMST is set, then carry out MST Multiple Spanning Tree Instance topology and calculate, according to the Multiple Spanning Tree Instance topology, switch is transmitted the dateout message.

Images