CN1710878A - Method for realizing reliability in network with independent loading control layer - Google Patents
Method for realizing reliability in network with independent loading control layer Download PDFInfo
- Publication number
- CN1710878A CN1710878A CNA2004100487714A CN200410048771A CN1710878A CN 1710878 A CN1710878 A CN 1710878A CN A2004100487714 A CNA2004100487714 A CN A2004100487714A CN 200410048771 A CN200410048771 A CN 200410048771A CN 1710878 A CN1710878 A CN 1710878A
- Authority
- CN
- China
- Prior art keywords
- network
- entity
- ownership
- functional
- functional entity
- 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 disclosed method is applicable to network of possessing independent load control layer. The method includes steps: setting up more than one functional entities with identical function characteristics as multiple attributive functional entities; step A that is to say in procedure of running network, in real time determining whether attributive functional entities are needed to rearrange; if yes, current normal operational attributive functional entity takes over current operation; otherwise, carrying out step A continuously. When one or more attributive functional entities go out of order, other attributive functional entities are able to take over current operation so as to guarantee continued service data stream. The invention raises reliability of end-to-end QoS architecture and ability for balancing load so that network planning and flow control are convenient. The method is suitable to network in any scale without limitation in network structure.
Description
Technical field
The present invention relates to quality of service (Qos) control technology end to end, be meant especially a kind of in having the network of independent bearing control layer the implementation method of reliability.
Background technology
Along with the continuous expansion of internet (Internet) scale, various network services are fallen over each other to emerge in large numbers, and advanced multimedia system also emerges in an endless stream.Because real time business is comparatively responsive to characteristics such as network transfer delay, delay jitters, when sudden high file transfer (FTP) being arranged on the network or contain the business such as hypertext transmission (HTTP) of image file, real time business will be subjected to very big influence; In addition, because multimedia service will take a large amount of bandwidth, so also will make and need guaranteed key business to be difficult to be transmitted reliably in the existing network.So, to be transmitted reliably for guaranteeing key business, various service quality (QoS, Quality of Service) technology is just arisen at the historic moment.The Internet engineering duty group (IETF, Internet Engineering Task Force) has proposed a lot of service models and mechanism, to satisfy the demand of QoS.What industry was relatively approved at present is to use integrated service (Int-Serv, Integrated Service) model at the access or the edge of network, uses differentiated service (Diff-serv, Differentiated Service) model at server.
The Diff-serv model only ensures QoS by the measure of setting priority level, though this model has the high characteristics of line efficiency, concrete effect is difficult to prediction.Therefore, industry is that the Diff-Serv model of backbone network has been introduced an independently bearer control layer, set up the special Diff-Serv qos signaling mechanism of a cover, and set up a resource management layer specially for the Diff-Serv network, the topological resource of supervising the network, this resource management Diff-Serv mode is called as the Diff-Serv of independent bearing control layer model.Fig. 1 is the schematic diagram of this model, and wherein, 101 is service server, belongs to service control layer, can realize functions such as soft switch, as Call Agent (CA); 102 is load network resource manager, belongs to bearer control layer; 103 is edge router (ER, Edge Router), and 104 is core router, and 103 and 104 all belong to bearer network.In this model, load network resource manager is responsible for configuration management rule and network topology, is client's service bandwidth application Resources allocation.Transmit client's service bandwidth application request and result between the load network resource manager of each management domain by signaling, and each load network resource manager is the routing information of professional application distribution etc.When the service bandwidth application of bearer control layer process user, will determine the path of customer service, load network resource manager can notify ER to transmit Business Stream according to the path of appointment.How bearer network realizes that according to the definite path of bearer control layer user traffic flow is by specifying routing forwarding, the existing technology of industry mainly is to utilize the MPLS technology at present, use the resource reservation mode to set up LSP, use the explicit route mechanism of RSVP-TE or CR-LDP to set up LSP end to end along the service flow path of bearer control layer appointment.
Current just simple connection of each key function inter-entity of network with independent bearing control layer, when certain functional entity, break down or during heavy traffic as resource control entity (RCF), functional entity with its associated, as edge router ER, business control function body SCF, can't be effectively under current architectural framework, operate as normal fully, therefore, in the prior art, in having the network of independent bearing control layer, when a functional entity broke down, the neither one reliability engineering guaranteed that current network normally moves.
At present, the method for the simplest assurance reliability is a cold standby.Cold standby is meant the fully backup of an entity as another entity.If A is an entity A, B is the backup entity of entity A.When entity A breaks down, utilize the backup entity B to substitute entity A fully.But for the backup entity B, realize that the precondition that substitutes fully is: carrying connects with professional the connection and all needs to rebuild.The method of this cold standby, at the networking initial stage, network size is very little, and professional real-time is less demanding, and because cold standby does not need to switch and smoothly, so cold standby is the most effective routing mechanism, and realize simple.Therefore, as long as traffic carrying capacity is little, allow to interrupt, cold standby always can be worked with flying colors.But along with the growth of traffic carrying capacity, the expansion of the demanding business of real-time, what the user wished to feel is that business is not interrupted, the method for cold standby is showed deficiency gradually.In this case, device fails needs interrupting service to rebuild, and has just become the shortcoming of cold standby maximum.So complexity, use cold standby in the demanding wide area network of real-time, can make a lot of business when the bearer network segment fault, be forced to interrupt rebuilding, it does not possess the multi-homing feature of key business node simultaneously, can't realize load balancing and network security.
Figure 2 shows that the bandwidth broker device model of QBone, this model also is a kind of Diff-Serv model scheme of independent bearing control layer.Referring to shown in Figure 2, Internet2 has defined the corresponding bandwidth proxy server for each Diff-Serv management domain specially, the bandwidth broker device is responsible for handling from subscriber's main station, perhaps service server, perhaps network maintenance staff's bandwidth application request, the bandwidth broker device determines whether to allow user's bandwidth application according to the resource reservation situation of current network and configured strategy and the professional SLA that signs with the user.All kinds of SLA configuration informations of record, the topology information of physical network, the configuration information and the policy information of router in the wide manager, a large amount of static and dynamic information such as user authentication information, current resource obligate information, network seizure condition information.Simultaneously, Bandwidth Broker also needs to write down routing iinformation, to establish user service flow path and cross-domain downstream bandwidth manager position.No reliability design.In the Bandwidth Broker model of Internet 2, the resource and the configuration information of the all-router in the direct management area of Bandwidth Broker exist topological sum to manage too complicated problems; Simultaneously, Bandwidth Broker needs the dynamic routing information in Minute book zone, has the frequent problem of routing table update, causes the instability of network reservation; The professional route that Bandwidth Broker is determined according to the dynamic routing information of one's respective area also is difficult to consistent with the forwarding route of Business Stream reality.No reliability design and key node multi-homing feature.Because a lot of implementation issues that the Bandwidth Broker model exists also do not have commercialization before this type order.
Figure 3 shows that the structural representation of the Rich QoS of Japanese NEC Corporation.In scheme shown in Figure 3 is as critical component by the QoS server.Its solution also comprises strategic server and LIST SERVER and the network management monitoring server that matches with the QoS server.When service server after the QoS server sends bandwidth request, the QoS server is noted down the resource request of this calling, and according to its qos requirement, and the current resource situation of current topological sum of bearer network is that service request is distributed the bearer path meet the demands, and distribution result is fed back service server.The QoS server sends the order of corresponding LSP strategy modification according to the bandwidth occupancy situation of business to strategic server, and strategic server just according to the order of QoS server, disposes the respective edges router.The explicit route technology that edge router will use MPLS LSP to set up according to the path of QoS server appointment, rebulids or adjusts LSP.
In this networking mode, what the QoS server was managed remains one than complicated bearer network, and router quantity is many.Though the professional control of carrying out is arranged, there are not reliability design and key node multi-homing feature, autgmentability is very poor, and network size is limited, can not adapt to the teleservice demand of a national public network.
Summary of the invention
In view of this, the object of the present invention is to provide a kind of in having the network of independent bearing control layer the implementation method of reliability,
In order to achieve the above object, the invention provides a kind of in having the network of independent bearing control layer the implementation method of reliability, this method comprises: the functional entity that be provided with more than, has the identical function characteristic for belonging to functional entitys more, and this method also comprises:
A. by the ownership functional entity according to service traffics by the strategy of the pro-rata of setting in advance with service distributing to a plurality of ownership functional entitys;
B. in network operation process, whether real-time judge needs to switch the ownership functional entity, and if desired, then current operation belongs to functional entity normally and takes over work at present, otherwise, continue execution in step B.
In described ownership functional entity, be provided for the heartbeat line of testing state between any two ownership functional entitys.
The data backup passage of backup the other side data is set between described two ownership functional entitys.
The described current state that detects any two ownership functional entitys by the heartbeat line in real time that is judged as, determine whether an ownership functional entity receives that another belongs to the heartbeat signal of functional entity in the predefined time, if receive, then do not need to switch execution in step A; Otherwise, switch.
The working method of described many ownership functional entitys is an active/standby mode, or the load sharing mode.
Described network with independent bearing control layer comprises business control function body SCF, resource control function entity RCF, edge/transmission route device entity ER/TR, Access Network AN.
The functional entity that be provided with more than one, has an identical function characteristic for the steps of many ownership functional entitys is: an AN or edge/transmission route device entity ER/TR is set to be connected with a plurality of RCF, and described AN or ER/TR and a plurality of RCF are in same resource management territory.
The functional entity that be provided with more than one, has an identical function characteristic for the steps of many ownership functional entitys is: a SCF is connected with a plurality of RCF.
The functional entity that be provided with more than one, has an identical function characteristic for the steps of many ownership functional entitys is: a RCF is connected with a plurality of SCF.
The functional entity that be provided with more than one, has an identical function characteristic for the steps of many ownership functional entitys is: a RCF is connected with a plurality of RCF in other resource management territory.
The present invention forms many attaching relations by two above identical functions entities being set as many ownership functional entitys between connected other functional entity of these functional entitys.Can be between many ownership functional entitys by configuration or by certain safe transmission channel intercommunication, knowing the other side mutually, thereby need determining whether switches.When one or several in the angelica functional entity broke down, other ownership functional entity can be taken over the work of faulty entity, guaranteed the continuity of business data flow, the operation of non-interrupting service.This method has improved the reliability and the load balancing ability of end-to-end QoS framework greatly, makes things convenient for the planning and the Flow-rate adjustment of network.At the different situations of miscellaneous service, and the concrete condition of network, different redundant home device can be set, satisfy the QoS demand of user to business.And the present invention to network configuration without limits, applicable to the network of any scale, realizes simply being easy to maintenance management.
Description of drawings
Fig. 1 is a bearer control layer network model independently in the prior art;
Fig. 2 is the SIBBS signalling diagram between the bandwidth broker device of Internet2;
Fig. 3 is the Rich QoS scenario-frame schematic diagram of Japanese NEC Corporation;
Fig. 4 is for realizing the networking schematic diagram of the many ownership of the present invention;
Fig. 5 is the connection diagram between any two the ownership functional entitys of the present invention.
Embodiment
For making the purpose, technical solutions and advantages of the present invention clearer, the present invention is described in further detail below in conjunction with accompanying drawing.
Core content of the present invention is: in having the network of independent bearing control layer, two above identical functions entities are set as many ownership functional entitys, connected other functional entity of these functional entitys, promptly by the ownership functional entity, between form many attaching relations, can adopt each other active/standby mode or load sharing mode to move.Quilt ownership functional entity arrives a plurality of ownership functional entitys by the pro rate strategy of setting in advance with service distributing according to service traffics; Can be between many ownership functional entitys by configuration or by certain safe transmission channel intercommunication, knowing the other side mutually, thereby need determining whether switches.When one or several in the angelica functional entity broke down, other ownership functional entity can be taken over the work of faulty entity, guaranteed the continuity of business data flow, the operation of non-interrupting service.Functional entity can be resource control entity, edge router and transmission route device etc.
Many ownership here actually are meant a kind of special connected mode, and promptly many ownership functional entitys link to each other with the another one functional entity.As shown in Figure 4, an access node (AN) can be connected on a plurality of edge routers (ER); ER or AN can be connected on a plurality of resource control entities (RCF) in same territory; A RCF can be connected on a plurality of operation controlling entity (SCF); A SCF can be connected on a plurality of RCF; A RCF can be connected on the RCF in other territory.When one or several in the angelica functional entity breaks down, as with opened by the link of ownership functional entity, hardware and software failure etc., other ownership functional entity can be taken over the work of faulty entity, guarantees the continuity of business data flow, the operation of non-interrupting service.Certainly, do not breaking down, can specify yet and switch by the third party.
Many ownership functional entitys can adopt active and standby each other working method, are called active and standby ownership functional entity.In course of normal operation, need to detect in real time the current state of active and standby ownership functional entity in many ownership functional entitys, judge whether to carry out active and standby switching.Here, the current state that described detection belongs to functional entity more can be that active and standby ownership functional entity detects mutually, or is directly detected by third party device.Detection mode can be by the heartbeat line between the active and standby ownership functional entity, and the signaling between the active and standby ownership functional entity transmits passage and realizes.
Referring to shown in Figure 5, A is an ownership functional entity, and B is the Another Place To Fall functional entity, has the inclination wire jumper between two ownership functional entitys, carries out the maintenance and the negotiation of state by the heartbeat line; Also has the data backup passage between two ownership functional entitys, to realize the non real-time automatic data backup or to come Backup Data by maintenance command.
Illustrate, it is such belonging to the detection that functional entity implements fully: if belong to the heartbeat signal that functional entity is not received main ownership functional entity at the appointed time fully, think that then main ownership functional entity breaks down, so change the state of self into " work ", expression can be carried out normal Signalling exchange by connected miscellaneous equipment.Whether overtime herein, proprietary protocol is generally adopted in the communication between the heartbeat line, and the value of fixed time can be pre-configured, and indication is set simultaneously when sends the heartbeat transmission timer of heartbeat signal and indication and receive heartbeat signal heartbeat timeout timer.
Certainly, belong to functional entity more and can also adopt the load sharing mode to work, in this case, detection mode also can detect the other side mutually by the heartbeat line and whether be in normal operating conditions.
Be that example illustrates technical scheme of the present invention with RCF as many ownership entity of ER below.
RCF is as many ownership entity of ER, be meant a plurality of RCF in an ER (ER1) and the territory (RCF1 ..., RCFn) be connected, i.e. the situation that ER is controlled by a plurality of RCF.
Its course of work is as follows:
Different user or professional QoS demand from SCF send to RCF1 respectively, ..., RCFn, RCF1, ..., RCFn determines required carrying route according to network topology and resource situation respectively, and route and qos parameter are all issued ER1, finishes the routing work of the different user that originates among the ER1 or professional carrying data flow.
Work as RCF1 ..., when RCFn receives the release resource request of SCF, RCF1 ..., RCFn will discharge corresponding QoS resource, and issue the bearing resource release command to ER1, the resource reservation of respective carrier data flow among the cancellation ER1.
If one or several among the RCF, as RCF1, RCF2 ..., break down, as with ER between be connected interruption, this moment, ER reported other the one or more of ownership functional entity that do not break down with affected business state information, as RCF3, RCF4, ..., take over the work of faulty entity by faulty entity not, guarantee professional not middle-end operation.
Be that example is come method of the present invention with ER as many ownership entity of RCF below.
ER as many ownership of RCF be meant a plurality of ER in a RCF (RCF1) and the territory (ER1 ..., Ern) be connected, a RCF controls the situation of a plurality of ER.
Its course of work is as follows:
Different user or professional QoS demand from SCF send to RCF1, RCF1 determines required carrying route according to professional start-stop, network topology and resource situation respectively, route and qos parameter are issued ER1 respectively, ..., ERn, finish originating from ER1 respectively ..., the routing work of the carrying data flow among the ERn.
When RCF1 received the release resource request of SCF, RCF1 discharged corresponding QoS resource, and issued the bearing resource release command to relevant ER according to professional start-stop parameter, the resource reservation of respective carrier data flow among the cancellation ER.
If one or several among the ownership functional entity ER, as ER1, ER2 ..., break down, this moment, RCF sent to the one or more of other ownership functional entity that does not break down with affected business state information, as ER3, ER4 ..., take over the work of faulty entity by faulty entity not, guarantee professional not middle-end operation.
Many ownership course of work of other functional entity can be analogized according to separately connection and functional characteristic.
In a word, the above is preferred embodiment of the present invention only, is not to be used to limit protection scope of the present invention.
Claims (10)
1, a kind of in having the network of independent bearing control layer the implementation method of reliability, be applied to have in the network of independent bearing control layer, it is characterized in that the functional entity that be provided with more than, has the identical function characteristic for belonging to functional entitys more, this method also comprises:
A. by the pro rate strategy of setting in advance service distributing is arrived a plurality of ownership functional entitys according to service traffics by the ownership functional entity;
B. in network operation process, whether real-time judge needs to switch the ownership functional entity, and if desired, then current operation belongs to functional entity normally and takes over work at present, otherwise, continue execution in step B.
2, method according to claim 1 is characterized in that, is provided for the heartbeat line of testing state in described ownership functional entity between any two ownership functional entitys.
3, method according to claim 2 is characterized in that, the data backup passage of backup the other side data is set between described two ownership functional entitys.
4, method according to claim 1, it is characterized in that, the described current state that detects any two ownership functional entitys by the heartbeat line in real time that is judged as, determine whether an ownership functional entity receives that another belongs to the heartbeat signal of functional entity in the predefined time, if receive, then do not need to switch execution in step A; Otherwise, switch.
5, method according to claim 1 is characterized in that, the working method of described many ownership functional entitys is an active/standby mode, or the load sharing mode.
6, method according to claim 1 is characterized in that, described network with independent bearing control layer comprises business control function body SCF, resource control function entity RCF, edge/transmission route device entity ER/TR, Access Network AN.
7, method according to claim 6, it is characterized in that, the functional entity that be provided with more than one, has an identical function characteristic for the steps of many ownership functional entitys is: an AN or edge/transmission route device entity ER/TR is set to be connected with a plurality of RCF, and described AN or ER/TR and a plurality of RCF are in same resource management territory.
8, method according to claim 6 is characterized in that, the functional entity that be provided with more than, has an identical function characteristic for the steps of many ownership functional entitys is: a SCF is connected with a plurality of RCF.
9, method according to claim 6 is characterized in that, the functional entity that be provided with more than, has an identical function characteristic for the steps of many ownership functional entitys is: a RCF is connected with a plurality of SCF.
10, method according to claim 6 is characterized in that, the functional entity that be provided with more than, has an identical function characteristic for the steps of many ownership functional entitys is: a RCF is connected with a plurality of RCF in other resource management territory.
Priority Applications (4)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CNA2004100487714A CN1710878A (en) | 2004-06-18 | 2004-06-18 | Method for realizing reliability in network with independent loading control layer |
| PCT/CN2005/000847 WO2005122495A1 (en) | 2004-06-14 | 2005-06-14 | A method for mbms user quantity counting |
| EP05757340.4A EP1758316B1 (en) | 2004-06-14 | 2005-06-14 | Method for ensuring reliability in a network including a separate bearer control layer |
| US11/610,693 US7953015B2 (en) | 2004-06-14 | 2006-12-14 | Method for ensuring reliability in network |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CNA2004100487714A CN1710878A (en) | 2004-06-18 | 2004-06-18 | Method for realizing reliability in network with independent loading control layer |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN1710878A true CN1710878A (en) | 2005-12-21 |
Family
ID=35707061
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CNA2004100487714A Pending CN1710878A (en) | 2004-06-14 | 2004-06-18 | Method for realizing reliability in network with independent loading control layer |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN1710878A (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2008022519A1 (en) * | 2006-08-15 | 2008-02-28 | Huawei Technologies Co., Ltd. | A method and system for realizing service relocation |
| CN101626596B (en) * | 2008-07-09 | 2011-08-31 | 中国移动通信集团公司 | Method, device and system for generating service distributing strategy |
| CN104468408A (en) * | 2013-09-22 | 2015-03-25 | 中国电信股份有限公司 | Method for adjusting dynamically service bandwidth and control center server |
-
2004
- 2004-06-18 CN CNA2004100487714A patent/CN1710878A/en active Pending
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2008022519A1 (en) * | 2006-08-15 | 2008-02-28 | Huawei Technologies Co., Ltd. | A method and system for realizing service relocation |
| CN101128047B (en) * | 2006-08-15 | 2010-04-14 | 华为技术有限公司 | Method for reallocating core network service |
| US8208453B2 (en) | 2006-08-15 | 2012-06-26 | Huawei Technologies Co., Ltd. | Method and system for realizing service reallocation |
| CN101626596B (en) * | 2008-07-09 | 2011-08-31 | 中国移动通信集团公司 | Method, device and system for generating service distributing strategy |
| CN104468408A (en) * | 2013-09-22 | 2015-03-25 | 中国电信股份有限公司 | Method for adjusting dynamically service bandwidth and control center server |
| CN104468408B (en) * | 2013-09-22 | 2018-04-06 | 中国电信股份有限公司 | For dynamically adjusting the method and control centre's server of service bandwidth |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| EP1758298B1 (en) | Method and system for realizing the reliability guarantee of end-to-end quality of service | |
| EP2086197B1 (en) | A method and apparatus for controlling the link aggregation | |
| EP1843537A1 (en) | A process method for dealing with device overload in the communication network | |
| US7953015B2 (en) | Method for ensuring reliability in network | |
| CN1409526A (en) | Intelligent routing for effectively using network signal resource | |
| CN100558105C (en) | A kind of Network sending system and to the control method of issuing service | |
| CN100352215C (en) | Automatic detecting and processing method of label exchange path condition | |
| CN101030917A (en) | Method and apparatus for realizing MPLS TE on VLAN interface | |
| CN100334838C (en) | Method for realizing end-to-end service quality reliability guarantee | |
| CN1710878A (en) | Method for realizing reliability in network with independent loading control layer | |
| CN1330133C (en) | Method for detecting abnormal breaking of user talking | |
| CN1601966A (en) | Route path selection method | |
| EP1758299A1 (en) | A method ensuring the service reliability in the end-to-end service quality framework | |
| CN1595895A (en) | A routing method based on resource restriction | |
| CN1756186A (en) | Resource management realizing method | |
| CN100391154C (en) | Selecting method of path in resource supervisor | |
| CN1601965A (en) | Route path fault repairing method | |
| CN100370736C (en) | Method for managing resources based on planes | |
| CN1881893B (en) | Flow control system and method between QoS signalling points | |
| CN100466613C (en) | Method for processing service | |
| CN100455035C (en) | Routing method of forward constraint backward selecting | |
| CN100486191C (en) | Method for bearing step selecting route in controlling layer | |
| CN1791050A (en) | Method for ensuring service quality when heterogeneous network intercommunication | |
| US8208374B2 (en) | Method, network agent and bandwidth broker for managing the available bandwidth for connections between terminals of a packet-oriented communication network | |
| CN100394736C (en) | Improvement of multi-medium communication reliability by logic media |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C12 | Rejection of a patent application after its publication | ||
| RJ01 | Rejection of invention patent application after publication |
Open date: 20051221 |