CN103684855A - Intelligent control method for cluster routers - Google Patents
Intelligent control method for cluster routers Download PDFInfo
- Publication number
- CN103684855A CN103684855A CN201310619349.9A CN201310619349A CN103684855A CN 103684855 A CN103684855 A CN 103684855A CN 201310619349 A CN201310619349 A CN 201310619349A CN 103684855 A CN103684855 A CN 103684855A
- Authority
- CN
- China
- Prior art keywords
- router
- cluster
- alive
- keep
- control
- 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
Links
Images
Landscapes
- Data Exchanges In Wide-Area Networks (AREA)
Abstract
The invention discloses an intelligent control method for cluster routers. A master control router, a slave control router and edge routers are selected by intelligent boot negotiation or strategy control, functions of cluster router equipment numbering, global interface mapping, unified coding and the like are completed by agreed protocol interaction, an internal forwarding system of the cluster routers is established, a distributed internal connection structure is shielded for a conventional upper-layer application, and a virtualized flat network view is provided. Based on standard techniques such as BFD (bidirectional forwarding detection) and moderate reboot, disaster tolerance and destroy resistance capacities are provided by real-time maintenance and fault switching of the customized master and slave control routers and edge routers. The upper-layer application and the hardware forwarding system are shielded through complex internal connection relation of the cluster routers, seamless compatibility to existing router equipment software and hardware is supported, meanwhile, extraction of internal interconnection node relationship of the cluster routers and management channel establishment are realized, and independent network management capability is provided for child routing nodes of the cluster routers.
Description
technical field
The present invention relates to the method for a kind of cluster routers intelligence management and control.
Background technology
Cluster routers are combinations of distribution router and parallel router.Current existing cluster routers mostly are the interconnected cluster router structure of multimachine cabinet, the switching node using is generally to forward rack or traditional unit router, can realize the forwarding of each router of edge, but the interconnected cluster routers of multimachine cabinet need to be usingd special-purpose routing forwarding unit as node, interconnected with special purpose interface and agreement between node, the flexibility of its expansion is very restricted, and can not meet cluster routers in larger capacity, the demand of flexible configuration etc. more.
Summary of the invention
In order to overcome the above-mentioned shortcoming of prior art, the invention provides the method for a kind of cluster routers intelligence management and control, solve conventional cluster router and needed special purpose interface and the interconnected problem of agreement ability, can realize in the situation that substantially not changing the existing software of traditional route exchange device, fast by the organic cluster router system that is built into of the router of a plurality of different abilities, diverse location and different disposal platform; Many for cluster routers fringe node, the problem such as manageability not, that the present invention has realized is preferably virtual to node abstraction in cluster routers, unified management and disaster tolerance is anti-ruins, and makes edge respectively form router collaborative work and parallel forwarding well; Realize the extraction of cluster routers edge network interconnecting relation and nodal information and the foundation of management path, the network management of seamless compatibility to each sub-route node of cluster routers simultaneously.
The technical solution adopted for the present invention to solve the technical problems is: the method for a kind of cluster routers intelligence management and control, after cluster router system powers on, in system, each router all starts the notice timer that powers on, the cluster routers that this facility information is carried in the timed sending notification packet that powers on, is made as broadcast address by object MAC; In the notification packet that powers on that the initial active and standby control router of simultaneity factor appointment sends, carry active and standby control appliance sign;
Initial master control router is being received powering on after notification packet for control router, edge router, by the transmission notice response message that powers on, and store the facility information of this notification packet that powers on of receiving, start keep-alive simultaneously and detect timer and wait for the timing keep-alive declaration that regularly receives corresponding router; If surpass the keep-alive time, do not receive keep-alive declaration, master control router thinks that corresponding router rolls off the production line, delete the facility information of storage and offline information is synchronized to other routers in this cluster route system, upgrading, delete global fault's node of cluster router system in time;
Initial master control router is after receiving that powering on of initial standby control router noticed, also TCP Socket connection foundation will be initiated to synchronize with the standby data of controlling between router automatically, follow-uply by this connection, complete the synchronous of various information, send simultaneously and notice response message; The initial standby router of controlling will start keep-alive probe timer receiving to power on to notice after replying, and regularly to master control router, send keep-alive declaration information, wait for that the keep-alive that receives master control router declares to reply;
Edge router is after receiving the notice response message of master control router, to stop sending the notification packet that powers on, start keep-alive probe timer, timing sends keep-alive declaration information to active and standby control router simultaneously, and the keep-alive declaration of wait reception control router is replied; If surpass the keep-alive response detected that keep-alive time edge router is not all received active and standby control router, think and restarting the control router failure of current group system the notice timer that powers on and waiting for that new control router produces;
Master control router, knowing that new edge router reaches the standard grade after information, will distribute independently router-number for this router, by the source MAC in router advertisement, number information be sent to this router; This edge router is received after number information, first completes the establishment of router edge interface plate port number and its addressing keyword, obtains the inside addressing MAC that each external interface of router is corresponding; This edge router is offered master control router by information spinners such as router-number, interface board number, port numbers and addressing MAC bindings, the edge router interface information that the cluster management dissection process of master control router reports, mode Unified coding with flattening creates independently logic interfacing, then master control router by with cluster route system in the control interface of core switching device, two layers of binding relationship of addressing MAC and edge router are issued to device interior center switching equipment and set up its two-layer retransmitting table item.
Compared with prior art, good effect of the present invention is: the present invention is directed to the administrative skill in cluster routers architectural framework, proposing a kind of novel logic simplifies, extendible technical scheme, realization completes fast a plurality of different abilities in the situation that substantially not changing the existing software of traditional route exchange device, the router of position and platform is built to cluster router system, realization is abstract to multinode in cluster routers framework, virtual, unified management and disaster tolerance are anti-to be ruined, the cluster routers network interface view of flattening can be provided for upper layer network application program, make each edge router well collaborative work and parallel forwarding simultaneously, break through unit case routing device at exchange capacity, power consumption, the restriction of the aspects such as heat radiation, larger to realize capacity, organize route switching system flexibly.
Accompanying drawing explanation
Examples of the present invention will be described by way of reference to the accompanying drawings, wherein:
Fig. 1 is cluster routers intelligence management and control edge frame schematic diagram;
Fig. 2 is that cluster routers logic port physical mappings is related to schematic diagram.
Embodiment
As shown in Figure 1, cluster routers comprise active and standby control router and several edge routers, master control router is responsible for the transmitting-receiving processing of overall network protocol massages and the message repeating of external interface, and standby control router and each edge router are only responsible for the forwarding of procotol and service message.First the present invention selects active and standby control router and edge router by start Intelligent Negotiation or policy control, protocol interaction by agreement completes the functions such as cluster routers device numbering, interface global map and Unified coding, set up the forwarding system of cluster routers inside, for traditional upper layer application shields distributed internal structure and virtualized flat network topology view is provided.Based on standard techniques such as BFD, graceful restarts, active and standby control router, edge router by customization are safeguarded and fault real-time reversion, disaster tolerance capability to resist destruction is provided, realize the extraction of cluster routers internal network interconnecting relation and nodal information and the foundation of management path, the network management capabilities of seamless compatibility to each sub-route node of cluster routers.Concrete grammar is as follows:
After cluster router system powers on, each router all starts the notice timer that powers in system, and the cluster routers that this facility information is carried in the timed sending notification packet that powers on, is made as broadcast address by object MAC; Simultaneity factor is specified initial active and standby control router by management strategy, in the notification packet that powers on that the initial active and standby control router of appointment sends, carries active and standby control appliance sign clearly to identify the identity of oneself.
Initial master control router is being received powering on after notification packet for control router, edge router; by the transmission notice response message that powers on; and store the facility informations such as router source MAC of this notification packet that powers on of receiving, start keep-alive simultaneously and detect timer and wait for the timing keep-alive declaration that regularly receives corresponding router.If surpass the keep-alive time (system is set in advance), can not receive keep-alive declaration; control router and think that corresponding edge router rolls off the production line; delete the facility information of storage and offline information is synchronized to other routers in this cluster route system, upgrading, delete global fault's node of cluster router system in time.
Initial master control router is after receiving that powering on of initial standby control router noticed; also TCP Socket connection foundation will be initiated to synchronize with the standby data of controlling between router automatically; follow-uply by this TCP, connected the synchronous of the various information such as relevant device configuration, external interface; send simultaneously and notice response message; the initial standby router of controlling will start keep-alive probe timer receiving to power on to notice after replying; regularly to master control router, send keep-alive declaration information, wait for that the keep-alive declaration of reception master control router is replied.
Edge router is after receiving the notice response message of master control router; to stop sending the notification packet that powers on; start keep-alive probe timer, timing sends keep-alive declaration information to active and standby control router simultaneously, and the keep-alive declaration of wait reception control router is replied.If surpass the keep-alive response detected that keep-alive time edge router is not all received active and standby control router; think and restarting the control router failure of current group system the notice timer waiting system that powers on and specifying new control router to produce by management strategy.
Master control router is knowing that new edge router reaches the standard grade after information, and router distributes independently router-number for this reason, by source MAC in router advertisement, this number information is sent to this router.Edge router is received after number information, first complete the establishment of router edge interface plate port number and its addressing keyword (being that MAC is final addressing keyword as common Ethernet forwards), obtain the inside addressing MAC that each external interface of router is corresponding.This edge router is offered master control router by information spinners such as router-number, interface board number, port numbers and addressing MAC bindings, the edge router interface information that the cluster management dissection process of master control router reports, creates independently logic interfacing with the mode Unified coding of flattening.
Master control router is when setting up global map logic interfacing, for each logic interfacing is distributed mutually different interface name, the mapping relations of simultaneously setting up logic interfacing and router-number, interface board number, port numbers and addressing MAC binding, the global map unified addressing logic list item structure of foundation as shown in Figure 2.
After externally network interface overall situation virtual map and logic interfacing have been set up, master control router by with cluster route system in the control interface of core switching device, two layers of binding relationship of addressing MAC and edge router are issued to device interior center switching equipment and set up its two-layer retransmitting table item.
The inventive method has realized the inner complicated inside-connecting relation of cluster routers well to upper layer application and the system mask of core forwarding hardware, so existing software that can the existing conventional router equipment of seamless compatibility, and without hardware repeater system is done to any change, greatly reduced traditional routing device to time and cost in cluster routers upgradings, transition process.The inventive method can offer cluster router system inside-connecting relation the management system of manufacturer in the mode of outside tripartite's interface, so the equipment management system of manufacturer can be with the clear overall picture of seeing the inner inside-connecting relation of cluster router system of visual means, can set up internal managing path by this method simultaneously and realize the management and control to each sub-router equipment of internal system, the management function that realized modularization, becomes more meticulous supports.Concrete manifestation is as follows:
One, shielding cluster complex router inside-connecting relation, the application of seamless adaptation conventional router:
Overall situation interface virtual map is for controlling all kinds of Routing Protocols and the upper layer application protocol shielding distributed type colony router hardware interconnecting relation of router, being presented in face of upper layer application is a flattening logic interfacing collection that is modeled as conventional router, and the upper layer application such as Routing Protocol are the work such as Pattern completion route calculating of routing device routinely.
Based on overall virtual map and Unified coding, controlling all kinds of Routing Protocols that move on router completes after agreement exchange, renewal route table items is handed down to cluster routers management, cluster routers management is carried at logical AND physical addressing mapping relations in new route table items message and sends to each edge router, finally completing cluster router system, to transmit complete machine synchronous, realizes the route forwarding function of cluster routers.
Two, cluster routers inside-connecting relation presents and manages support:
By being distributed in the workflows such as collaborative work, source synchronous of the cluster routers intelligence management and control software of each subset in cluster router system, cluster routers intelligence management and control software can produce the network topology that comprises the inner annexation of cluster routers edge subset, finally on the master control borad of master control router, forms and gathers.By the self-defining protocol interface of tripartite, can very simply edge network topological connection relation be extracted and report cluster routers management system, thereby present the overall picture that cluster routers edge connects.
Presenting cluster routers management and control software and can realize the informations such as converting flow to each sub-router, interface load, report and to the real-time foundation of edge router management channels based on cluster routers inside-connecting relation, can be network management system and provide that visualization is high, the comprehensive cluster routers distributed node of information work real-time situation, realize the Network Management Function to cluster routers edge sub-router board, module level.
Three, active and standby control router redundancy:
In cluster routers administrative skill, for the cluster routers network function avoiding causing after single control router failure processed is unavailable and from long problem recovery time, the master-slave redundancy mechanism that can control router by support is solved.
After control router has powered on, active and standby control router has connected the source synchronous such as global information by TCP Socket, and is surveyed and known mutually operating state each other by real-time keep-alive.At master control router, by keep-alive, detect and find standbyly while controlling router when machine, can specify or select at random new standby control router by management strategy, carry out equally subsequently TCP establishment of connection and global information etc. between new active and standby control router and synchronize.
When master control router occurs that generic failure or abnormal conditions should not work on, the standby router of controlling can guarantee that by technology such as GR or NSR in not interruption of network situation, triggering active and standby control router changes, and for controlling router, select the new standby router of controlling by former, complete corresponding TCP connection foundation simultaneously and synchronize with data.Even in master control router burst catastrophe failure situation, can not accomplish to support the application of the technology such as GR, thereby cannot ensure the not active and standby switching under interrupt levels of network, but the standby router of controlling also can complete the convergence again of network route within the fastest time, network is damaged to influence degree and drop to minimum.
Claims (2)
1. the method for cluster routers intelligence management and control, it is characterized in that: after cluster router system powers on, each router all starts the notice timer that powers in system, and the cluster routers that this facility information is carried in the timed sending notification packet that powers on, is made as broadcast address by object MAC; In the notification packet that powers on that the initial active and standby control router of simultaneity factor appointment sends, carry active and standby control appliance sign;
Initial master control router is being received powering on after notification packet for control router, edge router, by the transmission notice response message that powers on, and store the facility information of this notification packet that powers on of receiving, start keep-alive simultaneously and detect timer and wait for the keep-alive declaration that regularly receives corresponding router; If surpass the keep-alive time, do not receive keep-alive declaration, control router and think that corresponding router rolls off the production line, delete the facility information of storage and offline information is synchronized to other routers in this cluster route system, to upgrade, to delete global fault's node of cluster router system in time;
Initial master control router is after receiving that powering on of initial standby control router noticed, also TCP Socket connection foundation will be initiated to synchronize with the standby data of controlling between router automatically, follow-uply by this connection, complete the synchronous of various information, send simultaneously and notice response message; The initial standby router of controlling will start keep-alive probe timer receiving to power on to notice after replying, and regularly to master control router, send keep-alive declaration information, wait for that the keep-alive that receives master control router declares to reply;
Edge router is after receiving the notice response message of master control router, to stop sending the notification packet that powers on, start keep-alive probe timer, timing sends keep-alive declaration information to active and standby control router simultaneously, and the keep-alive declaration of wait reception control router is replied; If surpass the keep-alive response detected that keep-alive time edge router is not all received active and standby control router, think and restarting the control router failure of current group system the notice timer that powers on and waiting for that new control router produces;
Master control router, knowing that new edge router reaches the standard grade after information, will distribute independently router-number for this router, by the source MAC in router advertisement, number information be sent to this router; This edge router is received after number information, first completes the establishment of router edge interface plate port number and its addressing keyword, obtains the inside addressing MAC that each external interface of router is corresponding; This edge router is offered master control router by information spinners such as router-number, interface board number, port numbers and addressing MAC bindings, the edge router interface information that the cluster management dissection process of master control router reports, mode Unified coding with flattening creates independently logic interfacing, then master control router by with cluster route system in the control interface of core switching device, two layers of binding relationship of addressing MAC and edge router are issued to device interior center switching equipment and set up its two-layer retransmitting table item.
2. the method for a kind of cluster routers intelligence according to claim 1 management and control, it is characterized in that: control router when the mode Unified coding with flattening creates independently logic interfacing, for each logic interfacing, distribute mutually different interface name, set up the mapping relations of logic interfacing and router-number, interface board number, port numbers and addressing MAC binding simultaneously.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201310619349.9A CN103684855A (en) | 2013-11-29 | 2013-11-29 | Intelligent control method for cluster routers |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201310619349.9A CN103684855A (en) | 2013-11-29 | 2013-11-29 | Intelligent control method for cluster routers |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN103684855A true CN103684855A (en) | 2014-03-26 |
Family
ID=50321236
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201310619349.9A Pending CN103684855A (en) | 2013-11-29 | 2013-11-29 | Intelligent control method for cluster routers |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN103684855A (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106603443A (en) * | 2016-12-14 | 2017-04-26 | 郑州云海信息技术有限公司 | Method for data exchange in cluster server system |
| CN108900392A (en) * | 2018-06-22 | 2018-11-27 | 四川斐讯信息技术有限公司 | A kind of method and system of smart home multiple gateway access |
| CN116208479A (en) * | 2022-12-09 | 2023-06-02 | 四川天邑康和通信股份有限公司 | Voice port matching method, system, electronic equipment and readable storage medium |
| CN116707616A (en) * | 2023-06-28 | 2023-09-05 | 四川大学 | Space network multi-protocol distributed datagram forwarding method |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101931587A (en) * | 2009-06-19 | 2010-12-29 | 华为技术有限公司 | Method and system of virtue cluster route |
| CN101938417A (en) * | 2010-09-01 | 2011-01-05 | 中兴通讯股份有限公司 | Method for realizing configuration of main and auxiliary board cards as well as board cards |
| CN102150400A (en) * | 2009-11-13 | 2011-08-10 | 华为技术有限公司 | Router and cluster router |
| CN102273150A (en) * | 2010-03-29 | 2011-12-07 | 华为技术有限公司 | Cluster router and cluster routing method |
-
2013
- 2013-11-29 CN CN201310619349.9A patent/CN103684855A/en active Pending
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101931587A (en) * | 2009-06-19 | 2010-12-29 | 华为技术有限公司 | Method and system of virtue cluster route |
| CN102150400A (en) * | 2009-11-13 | 2011-08-10 | 华为技术有限公司 | Router and cluster router |
| CN102273150A (en) * | 2010-03-29 | 2011-12-07 | 华为技术有限公司 | Cluster router and cluster routing method |
| CN101938417A (en) * | 2010-09-01 | 2011-01-05 | 中兴通讯股份有限公司 | Method for realizing configuration of main and auxiliary board cards as well as board cards |
Non-Patent Citations (1)
| Title |
|---|
| 吴捷、施彦媛: "一种新型集群路由器智能管理技术的研究", 《通信技术》 * |
Cited By (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106603443A (en) * | 2016-12-14 | 2017-04-26 | 郑州云海信息技术有限公司 | Method for data exchange in cluster server system |
| CN106603443B (en) * | 2016-12-14 | 2019-09-27 | 郑州云海信息技术有限公司 | A method of carrying out data exchange in aggregated server system |
| CN108900392A (en) * | 2018-06-22 | 2018-11-27 | 四川斐讯信息技术有限公司 | A kind of method and system of smart home multiple gateway access |
| CN116208479A (en) * | 2022-12-09 | 2023-06-02 | 四川天邑康和通信股份有限公司 | Voice port matching method, system, electronic equipment and readable storage medium |
| CN116208479B (en) * | 2022-12-09 | 2024-05-28 | 四川天邑康和通信股份有限公司 | Voice port matching method, system, electronic equipment and readable storage medium |
| CN116707616A (en) * | 2023-06-28 | 2023-09-05 | 四川大学 | Space network multi-protocol distributed datagram forwarding method |
| CN116707616B (en) * | 2023-06-28 | 2024-02-06 | 四川大学 | Space network multi-protocol distributed datagram forwarding method |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN101027873B (en) | Systems and methods for network management | |
| CN101478435B (en) | Topology collecting method for stacking system and dual control board equipment | |
| CN105429883B (en) | The long connection method of multiple smart machines and server apparatus and smart machine | |
| CN102355366B (en) | Member-stacking device and method for managing member-stacking device at split stacking moment | |
| CN102917084B (en) | Automatic allocation method of IP address of node inside fat tree structure networking data center | |
| CN103067206A (en) | Internet of things across-machine room distributed type full automatic hot standby server switching system | |
| CN102710457B (en) | A kind of N+1 backup method of cross-network segment and device | |
| CN102904818A (en) | Method and device for updating ARP (Address Resolution Protocol) information table | |
| CN101594383B (en) | Method for monitoring service and status of controllers of double-controller storage system | |
| CN101965691B (en) | Cable redundancy and networked system | |
| CN103856357B (en) | A kind of stacking system fault handling method and stacking system | |
| CN101431428B (en) | Security monitoring service recovery method and system | |
| CN103002065A (en) | Method and device for sharing internet protocol (IP) address by host device and standby device | |
| CN103684855A (en) | Intelligent control method for cluster routers | |
| CN102843259A (en) | Middleware self-management hot backup method and middleware self-management hot backup system in cluster | |
| CN110661641B (en) | Virtual network function VNF deployment method and device | |
| CN105577444B (en) | A kind of wireless controller management method and wireless controller | |
| CN104731727A (en) | Double control monitoring and management system and method for storage system | |
| CN104468633A (en) | SDN southing security proxy product | |
| CN102647304A (en) | Synchronizing method of address resolution protocols and device | |
| CN108616376A (en) | A kind of FC network system failures dynamic reconfiguration method | |
| CN110830324A (en) | Method and device for detecting network connectivity of data center and electronic equipment | |
| CN100428722C (en) | Distributed processing system and processing method for multi-generating tree protocol | |
| CN103036702A (en) | Network segment crossing N+1 backup method and network segment crossing N+1 backup device | |
| CN104753707A (en) | System maintenance method and network switching equipment |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| WD01 | Invention patent application deemed withdrawn after publication | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20140326 |