CN103124235A - MS-ring Ethernet redundancy access algorithm - Google Patents

MS-ring Ethernet redundancy access algorithm Download PDF

Info

Publication number
CN103124235A
CN103124235A CN2013100185639A CN201310018563A CN103124235A CN 103124235 A CN103124235 A CN 103124235A CN 2013100185639 A CN2013100185639 A CN 2013100185639A CN 201310018563 A CN201310018563 A CN 201310018563A CN 103124235 A CN103124235 A CN 103124235A
Authority
CN
China
Prior art keywords
network
switch
port
ring
networks
Prior art date
Legal status (The legal status 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 status listed.)
Pending
Application number
CN2013100185639A
Other languages
Chinese (zh)
Inventor
周厚明
崔磊
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
WUHAN MAIWEI SHIDA SOFTWARE CO Ltd
Original Assignee
WUHAN MAIWEI SHIDA SOFTWARE CO Ltd
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
Application filed by WUHAN MAIWEI SHIDA SOFTWARE CO Ltd filed Critical WUHAN MAIWEI SHIDA SOFTWARE CO Ltd
Priority to CN2013100185639A priority Critical patent/CN103124235A/en
Publication of CN103124235A publication Critical patent/CN103124235A/en
Pending legal-status Critical Current

Links

Images

Landscapes

  • Small-Scale Networks (AREA)

Abstract

The invention discloses an MS-ring Ethernet redundancy access algorithm and relates to a technique to expand existing networks in an industrial Ethernet and maintain redundant link backup between the existing networks and newly-accessed networks. The technique is mainly applied to a two-layer Ethernet switch to enable two physical connections to exist between two Ethernets, one is main connection, another is backup connection, by default, communication of the two Ethernets is performed through the main connection, the backup connection serving as a backup of the main connection does not perform data transmission, and when the main connection fails, the backup connection can immediately execute functions of the main connection after the main connection is detected to be in a failure. All protocol messages designed by the MS-ring Ethernet redundancy access algorithm are transmitted only in one Ethernet, the newly-accessed networks and basic networks can independently execute ring network protocol algorithms of their own, and each ring network protocol algorithm is executed only in an own network and does not influence other networks.

Description

MS-ring ethernet redundancy Access Algorithm
Technical field
The present invention relates in Industrial Ethernet, existing network be expanded, keep simultaneously a kind of technology of redundant link backup between existing network and the network that newly adds.This technology mainly is used on layer 2 ethernet switch, make two physical connections are arranged between two Ethernets, one is main connection, one is that backup connects, in the time of main connection fault, backup connects can carry out the function of main connection at once, thereby guarantees the reliability that connects between two networks.
Background technology
Carry out stage by stage step by step in the very large situation of the construction of Ethernet, normally basic backbone network of model, then expand on basic network, thereby progressively build a large Ethernet.Generally only have a link between new network and basic network, when this link breaks down, cause two networks to lose connection, can't guarantee the reliability of communicating by letter.To and the finishing sector of this situation in generally can two links be arranged between two networks, wherein one is redundant link, only has to occur connecting fault when other one, this redundancy link is taken over the work of main connection.
Above-mentioned solution need to be carried out the RSTP agreement simultaneously between new network and basic network, if in basic network, equipment is not supported the RSTP agreement the such scheme inefficacy.Even two networks are all supported the RSTP agreement simultaneously, when sending fault, the self-healing time of RSTP is approximately about 20 seconds, and this self-healing time can not satisfy the very high industrial communications field of time requirement.This is proposed MS-ring ethernet redundancy Access Algorithm, it can guarantee that two networks have two links to be connected, this algorithm is only carried out this algorithm in the network terminal of new access, and basic network is had no requirement, and completes in the millisecond time in the conversion of primary link and backup link simultaneously.
Summary of the invention
Each port of Ethernet switch has two kinds of dissimilar states: connection status, port status.Wherein connection status is the physical state of switch ports themselves, shows whether the line of this port is normal, and this state is that a read states has two kinds: UP state and DOWN state.Port status is the port logic state of switch, and the port that it has determined switch can be received and dispatched the message of processing those particular types, determines also whether this port has address learning function simultaneously.The value of port status has multiple, but generally all comprises such four kinds: forwarding state, blocked state, learning state, disabled status.Their explanation as shown in Figure 1.
The execution of MS-ring algorithm need to be used the realization of BPDU frame.The BPDU Frame is to operate in the very special message of data link layer, and this BPDU frame can be received and dispatched at port blocked, substantially not this ability of normal data frame.Specify simultaneously the port that sends of message by the BPDU frame, also can obtain the entry port of message when receiving BPDU.
Can divide three kinds of situations according to the access way of new network and basic network with the topological structure of new access network discusses: actual application can equivalence becomes a kind of in these three kinds of situations.
The first situation: connect two lines of basic network all on same switch.The line road of serving as theme is set, article one, circuit is spare line, the connection status of main line is forwarding state under initial situation, spare line is disabled status, the communication of such two networks is all undertaken by main line, then the switch connection status of circulation timing scanning main line in a short period of time, if find that the main line port connects DOWN, the port status of main line port is set to disabled status so, simultaneously the port status of spare line is arranged forwarding state, then empty the address table of switch.The data that flowed from main line in this case will flow from spare line.And if then in the scanning of the port status of back, the circuit UP of discovery master port changes, the port status of spare line is set to disabled status, and then the port status of main line is set to forwarding state, empties at last address table.Above these operations be placed on operation always in the daemon on a backstage, just can guarantee the reliability of communicating by letter below this simple link.
The second situation: connect two lines of basic network on different switches.Need in this case to use the BPDU Frame, the form of this Frame as shown in Figure 2.This situation has further carefully in two kinds of situation: new access network is non-ring topology, is non-ring topology on the topology of new access network.Two kinds of topological structures of the below's minute new access network of situation discussion.
New access network is non-ring topology, the switch of new access network is divided into Three role here, main edge switch, and from edge switch, intermediary switch.Main edge switch: be used for connecting external network, its edge port is the data transmission channel of external network and own network.From edge switch: be used for connecting external network, its edge port is the transfer of data backup path of external network and own network.Intermediary switch: be used for the link of the inner respective switch of New-deployed Network.The port of edge switch (comprise main edge switch and from edge switch) has can divide two kinds of roles: edge port and common port.Switch role and port role explanation are as shown in Figure 3.The whole execution flow process of algorithm is: at first main edge switch in New-deployed Network is set and from edge switch and their edge port.Then carry out different flow processs according to role's difference of switch.the execution flow process of main edge switch is: it is forwarding state that main edge switch starts the port status that edge port is set, then the linking status of timing scan edge switch, if find that edge port connects DOWN, edge port is set to disabled status, send out link down message by the BPDU frame, continue to send this message, until receive the link down ack message of sending from edge switch, then after main edge switch detects edge connection up variation, send link up message, continue to send this message, until receive the link up ack message of sending from edge switch.Then the state of its main edge port is set to forwarding state, then empties the address table of switch.Main edge switch can send hello message every a set time simultaneously, this message be mainly tell from edge switch master edge switch working properly at present, if the edge port fault of main edge switch or connection down have changed, stop immediately sending hello message, can be according to whether receiving that hello packet decides the port status of the edge port of oneself from edge switch.The port state shift figure of host exchange as shown in Figure 4.More complicated than the state variation of the edge port of main edge switch from the state variation of the edge port of edge switch.Be initially located in disabled status from the edge port of edge switch, then it monitors the message that main edge switch sends, if it can not receive hello message in a period of time, it oneself state is set to forwarding state.Perhaps after it received link down message, the state of own edge port was set to forwarding state, empties address table, sent link down ack message as the response to link down message.After it received link up message, edge port was set to disabled status, sent link up ack message and responded, and emptied simultaneously address table.From the state diagram of edge switch as shown in Figure 5.Attention is sent because all message all is encapsulated in the BPDU frame, and the BPDU frame can be specified the port that sends of message, and the transmitting-receiving of the message of regulation edge switch all can not be carried out through edge port, just carries out at non-edge port.Such benefit is to prevent that these messages are on the impact of basic network.
Above-mentioned algorithm implementation has determined that becoming forwarding state from the edge port fault of main edge switch to the edge port from edge switch begins communication, time in the middle of this depends mainly on the time of main edge switch scanning Edge detected port, and smaller to the transmission time expand from edge switch from main edge switch, it is relevant in the processing mode of intermediary switch with message simultaneously.In the processing scheme of simplifying, intermediary switch can be carried out the MS-ring algorithm, intermediary switch is received message, the target MAC (Media Access Control) address of message is the address of BPDU, it is generally multicast address, the mode of switch default treatment multicast is the export-oriented all of the port forwarding of port that enters except message, and message can send to the another one edge switch from an edge switch like this.From the time expand edge port of the port status of edge switch switching time=edge Wiring faults self-healing time ≈ 0.5* host exchange detection time+transmission.Calculation is sent out if intermediary switch is carried out MS-ring, the Centronics port that intermediary switch connects two edge switch is set simultaneously, receive message from a Centronics port, forward from the another one Centronics port, here can be by adding static multicasting address (multicast address that MS-ring message adopts) and port (2 Centronics ports and a cpu port) vector in address table.Intermediary switch is receiving that MS-ring message will be transmitted to cpu and another Centronics port like this.This intermediary switch is when carrying out the MS-ring algorithm, detect simultaneously the connection status of oneself 2 Centronics port, if find to connect down, send so link down message, this message is received from edge switch, its edge port is set to forwarding state, empties simultaneously address table, sends link down ack message.Doing like this is port in order to ensure intermediary switch when breaking down, and network is cut into two.The network of new access is not looped network, itself does not have redundancy network, just lost contact with basic network with this part network that is connected from edge switch after middle circuit down falls, the switch that falls in the urgent need to linkdown from edge switch this moment is given notice message to processing from edge switch.So this processing mode of intermediary switch is more outstanding than simplifying the version processing mode.Certainly the simplification processing mode of intermediary switch has own advantage, be exactly intermediary switch can be that the very common switch that there is no management function serves as, it is not necessarily the same with host exchange is the switch of the same execution MS-ring of producer algorithm.Can reduce expenses like this.Even intermediary switch is not carried out the MS-ring algorithm, from edge switch equally owing to can not receive for a long time hello message, thereby the state of edge port is set to forwarding state and recovers communication automatically, recovers only in this case the consuming time long of communication.Fig. 6 is presented at 4 switch networkings message flow cardon under normal circumstances.What Fig. 7 showed is that host exchange edge port down is the message flow situation.Fig. 8 shows when host exchange edge port up changes, the message flow situation.Fig. 9 shows that the Centronics port down of intermediary switch is, message flow situation, Figure 10 show the message flow cardon that intermediary switch up changes.
New access network is ring topology, and the network of new access itself has been ring topology, and wherein 2 switches have two ports to be connected with basic network as edge port as edge switch.In this case, this network can be carried out the looped network agreement of oneself, guarantees reliability and the redundancy of own internal network.The processing of these two kinds of topological structures is basically identical, and the difference of their maximums is the processing of intermediary switch, there is no difference if intermediary switch is not carried out the MS-ring algorithm.If carry out the MS-ring algorithm, need to not send link up or link down message after detecting the connection variation so, its reliability is mainly guaranteed by the looped network agreement of oneself moving.Topmost lime light is: the network that newly adds selects the port of looped network agreement operation not comprise that the edge port of edge switch gets final product in operation looped network agreement.The network topology schematic diagram as shown in figure 11.One of them standing procedure is to carry out the RSTP algorithm in the network of new access, but this algorithm does not comprise the edge port of edge switch, allow the redundancy of new access network be guaranteed by the RSTP algorithm, the redundant reliability of this network and outside basic network is guaranteed by the MS-ring redundant arithmetic of edge switch.
 
Description of drawings
The state description of Fig. 1 switch ports themselves
Fig. 2 BPDU message format
Structural representation in Fig. 3 MS-ring algorithm other than ring type topology
The state diagram of main edge switch in Fig. 4 other than ring type topology
In Fig. 5 other than ring type topology from the state diagram of edge switch
Normal condition initial message flow process in Fig. 6 other than ring type topology
The edge port down message flow of main edge switch in Fig. 7 other than ring type topology
The edge port up message flow of main edge switch in Fig. 8 other than ring type topology
Intermediary switch down message flow in Fig. 9 other than ring type topology
Intermediary switch up message flow in Figure 10 other than ring type topology
Topology of networks figure in Figure 11 ring topology
The pie graph of Figure 12 MS-ring algoritic module
Embodiment
Comprise three major parts: user interface and diagnosis, protocol algorithm specific implementation, hardware relevant portion.Three parts are independent mutually, are convenient to transplant on different platforms and hardware.MS-ring consists of as shown in figure 12.
User interface and diagnosis are mainly opened, forbidding MS-ring, port status and the connection status that can check switch simultaneously.
Protocol algorithm realizes part: the specific implementation of agreement partly as far as possible and concrete hardware platform and system platform irrelevant, mainly be responsible for the transmitting-receiving of protocol message, process, safeguard the state machine of realizing switch ports themselves, it is this algorithm core.At first at operating system backstage long-play, secondly this agreement has utilized the BPDU frame of exchange chip to realize being operated in data link layer as a daemon in the realization of this algorithm.The frame structure that this algorithm relates to is consistent in switch outside, but particularly the frame structure of data link layer is relevant with exchange chip in switch inside, and exchange chip processing to label in treatments B PDU is discrepant.So its transmitting-receiving processes from essence or hardware is relevant, is just consistent beyond switch, can be general in different platform.
The hardware relevant portion: the hardware relevant portion is relevant with the exchange chip of operating system and employing, comprises two submodules: network interface card drive part, and switch ports themselves operation part.The below says that mainly protocol algorithm realizes part and hardware relevant portion.Drive the label field that needs special processing BPDU at network interface card, because the label that MS-ring needs exchange chip automatically to add, common upper strata uses need not need to delete this label basically, so driver is wanted can accurately to distinguish MS-ring frame and normal data frame and to do different processing, this can be simply target MAC (Media Access Control) address by frame judge.Different with normal frames when receiving the MS-ring Frame, it will keep the label that exchange chip adds, and the point of addition of this label may be different in different chips.Simultaneously for convenience of the upper procedure reception ﹠ disposal, it is immediately following 2 bytes in the source MAC back for the protocol fields that the needs interpolation is two layers, sometimes in order to prevent from washing out the byte of back, generally that Extended Spaces is vacated 2 bytes in this position, 2 bytes vacating are added specific value, and then upper layer application is bundled on this particular value and receives.

Claims (3)

1.MS-ring the ethernet redundancy Access Algorithm is in Industrial Ethernet, existing network to be expanded, while is at existing basic network and newly add a kind of technology that keeps the redundant link backup between network, it is characterized in that: be used on layer 2 ethernet switch, make two physical connections are arranged between two Ethernets, one is main connection, one is that backup connects, in the time of main connection fault, backup connects can carry out the function of main connection at once, thereby guarantees the reliability that connects between two networks.
2. according to claim 1, main connection be connected that connect can be at a switch, also can be at two switches, it is characterized in that: newly add network and basic network can independently carry out the looped network protocol algorithm of oneself, while looped network protocol algorithm separately can only be carried out in the network of oneself does not affect other network.
3. according to claim 1, newly add topology of networks to have multiple, it is characterized in that: newly add network topology structure to support loop configuration, also support the other than ring type topological structure.
CN2013100185639A 2013-01-18 2013-01-18 MS-ring Ethernet redundancy access algorithm Pending CN103124235A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN2013100185639A CN103124235A (en) 2013-01-18 2013-01-18 MS-ring Ethernet redundancy access algorithm

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN2013100185639A CN103124235A (en) 2013-01-18 2013-01-18 MS-ring Ethernet redundancy access algorithm

Publications (1)

Publication Number Publication Date
CN103124235A true CN103124235A (en) 2013-05-29

Family

ID=48455095

Family Applications (1)

Application Number Title Priority Date Filing Date
CN2013100185639A Pending CN103124235A (en) 2013-01-18 2013-01-18 MS-ring Ethernet redundancy access algorithm

Country Status (1)

Country Link
CN (1) CN103124235A (en)

Cited By (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103546383A (en) * 2013-11-08 2014-01-29 南车株洲电力机车研究所有限公司 Redundancy transmission method, device and system for looped network
CN103685079A (en) * 2014-01-02 2014-03-26 武汉迈威实达软件有限公司 Self-adaptive Ethernet single-loop network algorithm
CN111698157A (en) * 2020-07-23 2020-09-22 迈普通信技术股份有限公司 Link management method, board card and switch
CN115189986A (en) * 2022-05-31 2022-10-14 武汉迈威通信股份有限公司 Ring network redundancy backup method and system
CN115473792A (en) * 2022-08-01 2022-12-13 上海微波技术研究所(中国电子科技集团公司第五十研究所) Flow table-based dual-channel high-reliability communication control system and method

Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101079781A (en) * 2007-02-01 2007-11-28 北京东土科技股份有限公司 An implementation method for industrial Ethernet fast-speed redundancy
CN101262402A (en) * 2007-11-08 2008-09-10 北京东土科技股份有限公司 A method for realizing redundant backup between loop networks

Patent Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101079781A (en) * 2007-02-01 2007-11-28 北京东土科技股份有限公司 An implementation method for industrial Ethernet fast-speed redundancy
CN101262402A (en) * 2007-11-08 2008-09-10 北京东土科技股份有限公司 A method for realizing redundant backup between loop networks

Cited By (9)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103546383A (en) * 2013-11-08 2014-01-29 南车株洲电力机车研究所有限公司 Redundancy transmission method, device and system for looped network
CN103546383B (en) * 2013-11-08 2015-08-05 南车株洲电力机车研究所有限公司 Redundant transmission method, equipment and system on a kind of looped network
CN103685079A (en) * 2014-01-02 2014-03-26 武汉迈威实达软件有限公司 Self-adaptive Ethernet single-loop network algorithm
CN111698157A (en) * 2020-07-23 2020-09-22 迈普通信技术股份有限公司 Link management method, board card and switch
CN111698157B (en) * 2020-07-23 2022-05-24 迈普通信技术股份有限公司 Link management method, board card and switch
CN115189986A (en) * 2022-05-31 2022-10-14 武汉迈威通信股份有限公司 Ring network redundancy backup method and system
CN115189986B (en) * 2022-05-31 2024-02-27 武汉迈威通信股份有限公司 Ring network redundancy backup method and system
CN115473792A (en) * 2022-08-01 2022-12-13 上海微波技术研究所(中国电子科技集团公司第五十研究所) Flow table-based dual-channel high-reliability communication control system and method
CN115473792B (en) * 2022-08-01 2023-09-22 上海微波技术研究所(中国电子科技集团公司第五十研究所) Dual-channel communication control system and method based on flow table

Similar Documents

Publication Publication Date Title
CN100417094C (en) Network failure recovery method with redundancy port
US8488475B2 (en) Fault processing method, system and exchanging device based on industry ethernet network
EP2352253B1 (en) Method and apparatus for protecting link aggregation group of ethernet ring
US8139510B2 (en) Optimizations and enhancements to the IEEE RSTP 802.1w implementation
CN101652963B (en) Method for reconfiguring a communications network
US9218230B2 (en) Method for transmitting messages in a redundantly operable industrial communication network and communication device for the redundantly operable industrial communication network
CN104104570A (en) Aggregation processing method in IRF (Intelligent Resilient Framework) system and device
US20140071810A1 (en) Communication system and processing method therefor
CN101702663A (en) Method for updating ring network topology information, device and system thereof
CN103124235A (en) MS-ring Ethernet redundancy access algorithm
CN103607293B (en) A kind of flow rate protecting method and equipment
CN103581164A (en) Method for transmitting messages in redundantly operable industrial communication network and communication device for redundantly operable industrial communication network
JP2005130049A (en) Node
CN103107940B (en) For the redundant gateway system of device level looped network
CN101753397A (en) System, method and device for rapidly self-healing Ethernet ring
CN101674240A (en) Method and device for realizing single-node dual home for distributed system
WO2010031295A1 (en) Control method for ethernet failure recovery
CN104427012B (en) Port negotiation method and apparatus
CN101436975A (en) Method, apparatus and system for implementing rapid convergence in looped network
CN103731303A (en) Method and device of vertical converged framework for achieving multi-active detection
CN102420737B (en) Method for processing failure in MRP (Media Redundancy Protocol) looped network and MRP looped network
CN111343019A (en) Reconnection method, reconnection device and storage medium
CN102244589B (en) Method and opposite terminal apparatus for processing link fault in virtual switch unit system
CN102158383A (en) Data transmission method, device and system for E1 bidirectional ring network
CN103441940A (en) Method and device for switching network device from stand-by state to main state and network device

Legal Events

Date Code Title Description
C06 Publication
PB01 Publication
C10 Entry into substantive examination
SE01 Entry into force of request for substantive examination
C02 Deemed withdrawal of patent application after publication (patent law 2001)
WD01 Invention patent application deemed withdrawn after publication

Application publication date: 20130529