CN100372299C

CN100372299C - Network managing method for supporting distributed management information tree

Info

Publication number: CN100372299C
Application number: CNB2004100581843A
Authority: CN
Inventors: 施广宇; 李斌; 邓明昆
Original assignee: Huawei Technologies Co Ltd
Current assignee: Huawei Technologies Co Ltd
Priority date: 2004-08-13
Filing date: 2004-08-13
Publication date: 2008-02-27
Anticipated expiration: 2024-08-13
Also published as: CN1735032A

Abstract

The present invention discloses a network management method for supporting a distributed management information tree (MIT), which comprises: a method for constructing the distributed MIT and a method for managing a network are constructed. The method for constructing the distributed MIT comprises: respective network devices are abstracted into management objects (MOs); based on the containment relationship between the MOs, the MIT is organized. The MIT is divided into one or a plurality of MIT partitions; an MO partition manager is configured for the whole MIT, and MO managers are configured for the respective partitions. The method for managing the network comprises: each MO manager starts each partition thereof and registers the partition to the MO partition manager; external application accesses the respective partitions through the MO partition manager and the MO managers of the respective partitions. When a system is expanded, new network devices are organized into a new MIT partition which is started by a new MO manager and is registered to the MO partition manager. The adoption of the present invention can realize the management for the distributed MIT, and the system in expansion can be expanded in a mode transparent to users.

Description

A kind of network management of supporting the distributed management inforamtion tree

Technical field

The present invention relates to a kind of network management technology, particularly a kind of Network Management method of supporting distributed management inforamtion tree (MIT).

Background technology

At present, along with the progress of the communication technology and the development of communication service, the equipment that telecommunication network management system need be managed is that quantity or device type are all increasing.But existing network management system is owing to be subjected to the restriction of system architecture, and its management capacity is subject to the disposal ability of separate unit server.

Fig. 1 is a prior art networking mode schematic diagram.As shown in Figure 1, network management system 1 is managed existing equipment by connecting 1, when network has had fairly large expansion, need manage newly added equipment, and existing network management system 1 has not possessed enough disposal abilities when managing newly added equipment, usually can adopt following dual mode to solve problem: one, mode by HardwareUpgring enlarges the network management capacity, promptly stop the operation of server in the existing network management system 1, for this server adds new internal memory or central processing unit (CPU), sometimes even have to change the work station of whole network management system, make network management system 1 can manage by connecting 2 pairs of newly added equipments simultaneously; Two, server comes the extended network range of management by installing additional independently, promptly adds an other cover network management system 2, manage newly added equipment by connecting 3, and original network management system 1 is maintained the statusquo, and manages existing equipment by connecting 1.

According to above description to existing networking mode as seen, when network need increase more equipment, and when having to expand the legacy network management capacity, following problem may appear in the scheme of prior art: the upgrading of the network equipment and dilatation may need network management server to stop service operation, even shut down and upgrade; If the capacity expansion and upgrading of network will be accomplished non-stop-machine words; just must set up new independently network management system; the network management system that cover more than causing is same is also deposited; increased operating cost of operator greatly; in addition; whole network equipment is isolated in two covers or the network management system more than two covers, just can't realize the whole network positioning problems.

Therefore, telecom operators are in order to realize minimumization of operation cost, and the ability of raising the whole network positioning problems, will inevitably propose to use one to overlap the demand that network management system is managed all devices, this has just proposed unprecedented requirement to the capacity and the extensibility of network management system.In the face of this situation, just must make existing network management system expand its disposal ability in a kind of mode of user transparent according to the network size that will manage.

Summary of the invention

In view of this, main purpose of the present invention is to provide a kind of Network Management method of supporting distributed MIT, can realize distributed MIT management, make the disposal ability of network management system not be subject to the disposal ability of separate unit server, when the network equipment surpasses the capacity of network management system, system is expanded in the mode of user transparent.

For achieving the above object, technical scheme of the present invention is achieved in that

The invention discloses a kind of Network Management method of supporting distributed MIT, this method comprises: set up distributed MIT method and the management this support the method for the network of distributed MIT;

The described method of setting up distributed MIT comprises: each network equipment is abstracted into management object (MO), organize MIT according to the containment relationship between the MO, and MIT is divided into one or more MIT subregions, for being provided with MO manager (Manager), each subregion is used to visit the MO of subregion inside separately, for whole M IT is provided with the MO manager that the MO zone manager is used to visit each subregion;

Described management this support the method for the network of distributed MIT to comprise: the MO manager by each subregion starts subregion separately, and the log-on message of subregion is registered in the log-on message table of MO zone manager separately; Applications is visited the MO manager of each subregion according to the log-on message in the log-on message table in the MO zone manager, again by the network equipment in its subregion of MO manager accesses; When System Expansion, to increase the network equipment newly and be abstracted into MO, organize new MIT subregion according to the containment relationship between the MO of this newly-increased network equipment correspondence, by for the new MO manager of this new MIT subregion setting starts this new MIT subregion, and log-on message that will this new MIT subregion is registered in the log-on message table of MO zone manager.

Set up in the method for distributed MIT, describedly MIT is divided into one or more MIT partitioned method is:

A1. the MO that selects to be arranged in whole M IT low level is defined as the root node of MIT, and these root nodes are divided into root partition;

A2. all MO beyond the described root partition of step a1 among the MIT are divided into zero, one or more subregion.

Wherein, after described step a2, also further comprise: will divide in each the described subregion obtain through step a2, and be subdivided into root partition with the MO of the MO direct correlation of an above subregion.

Management supports in the method for distributed MIT network, and the MO manager of described each subregion starts separately that partitioned method is:

B1. system supervisor sends the MO manager of starting command to each subregion;

B2. each MO manager loads the interior network equipment of subregion separately, and the log-on message of subregion is sent to the MO zone manager separately;

Described will be separately the log-on message of subregion be registered in the log-on message table of MO zone manager and be: the b3.MO zone manager is kept at the described log-on message of step b2 in the log-on message table of self.

Among the step b1, the mapping relations table of the address information of the MO manager that each subregion that system supervisor has been provided with according to self is corresponding with it obtains the address information of each subregion MO manager, and sends starting command by the address information of this MO manager.

Among the step b2, described MO manager loads the network equipment of each MO correspondence in the subregion of described self correspondence according to the node identifier of all MO in self the corresponding subregion that has been provided with.

Wherein, described management this support the method for the network of distributed MIT further to comprise: each MIT subregion is abstracted into zone object, externally provides MO manager agency (Proxy) class to encapsulate each zone object, and MO manager acting server is set in system;

Described applications is visited the MO manager of each subregion according to the log-on message in the log-on message table in the MO zone manager, and the method by the network equipment in its subregion of MO manager accesses is again:

C1. applications sends access request to MO manager acting server; Described access request comprises the node identifier of MO to be visited at least;

C2.MO manager acting server is transmitted this access request to the MO zone manager;

The c3.MO zone manager obtains the root node identifier of MO to be visited place subregion according to the node identifier of MO to be visited in the described access request of step c2, obtain the address information of the MO manager of MO to be visited place subregion again according to the root node identifier inquiry and registration information table of this MO to be visited place subregion, and return this MO manager address information and give MO manager acting server;

C4.MO manager acting server is sent to this access request by this MO manager address information the MO manager of MO to be visited place subregion again;

The MO manager of MO c5. to be visited place subregion carries out corresponding operating according to the node identifier of MO to be visited in this access request to MO to be visited.

Wherein, described access request is a request to create;

Described MO to be visited is MO to be created; Described MO to be visited place subregion is the direct father node of a MO to be created place subregion;

Described step c5 is: the MO manager of the direct father node of MO to be created place subregion is inquired about the node identifier of all MO in the subregion that self has been provided with according to the node identifier of MO to be created in the request to create, obtain the node identifier of the direct father node of MO to be created, and under the direct father node of this MO to be created, create this MO to be created.

Wherein, when described MO to be created was the partition root node, described request to create further comprised the title of MO to be created place subregion; Described MO to be visited place subregion is a MO to be created place subregion;

Among the step c3, the MO zone manager is with the node identifier of MO to be created in the described request to create of the step c2 root node identifier as MO to be created place subregion, according to MO to be created place partition name inquiry and registration information table in the request to create, obtain the address information of the MO manager of MO to be created place subregion again;

Described step c5 is: the MO manager of MO to be created place subregion is created this MO to be created by the node identifier of MO to be created in the request to create in this subregion.

Wherein, when described MO to be created is the partition root node, after described step c5, further comprise: the MO manager of MO to be created place subregion sends log-on message to the MO zone manager, this MO to be created is registered in the log-on message table of MO zone manager.

Wherein, described access request is a query requests; Described MO to be visited is MO to be checked; The node identifier of described MO to be visited is represented query context;

Described step c5 is: the node identifier of all MO in the MO manager of the MO to be checked place subregion subregion that inquiry has been provided with self according to the query context in this query requests, obtain the node identifier of all MO in the query context, obtain all MO information to be checked in the corresponding separately subregion according to the query contents prescribed information in the query requests again, then by the extraction of the query contents in query requests Query Result separately.

Wherein, after described step c5, further comprise: the Query Result that the MO manager of MO to be checked place subregion will be extracted separately returns to MO manager acting server, and MO manager acting server is organized each Query Result that receives and obtained whole Query Results and should return to applications by whole Query Results.

Wherein, described access request is for revising request; Described MO to be visited is MO to be revised;

Described step c5 is: the MO manager of waiting to revise MO place subregion is according to the node identifier of waiting to revise MO in this modification request, and by the modification content of revising in the request this MO to be revised made amendment.

Wherein, described access request is a removal request; Described MO to be visited is MO to be deleted;

Described step c5 is: the MO manager of MO place subregion correspondence to be deleted is deleted MO according to treating in this removal request node identifier is waited to delete MO with this and is deleted from this waits to delete the subregion at MO place.

Wherein, when MO described to be deleted is the partition root node, after described step c5, further comprise:

Wait that the MO manager of deleting MO place subregion sends log-off message to the MO zone manager, waits to delete MO with this and nullifies from the log-on message table of MO zone manager.

Wherein, described MO manager acting server is based on C++ or common object request broker system (CORBA) interface.

Manage in the method for this network of supporting distributed MIT, the method for described System Expansion is specially:

D1. system supervisor sends the MO manager of starting command to new MIT subregion correspondence;

The d2.MO manager loads the network equipment in the new MIT subregion, and log-on message that will this new MIT subregion is sent to the MO zone manager;

The d3.MO zone manager is kept at the log-on message that steps d 2 described MO managers send in the log-on message table.

Wherein, this supports the Network Management method of distributed MIT further to comprise, the mapping relations table of the address information of the MO manager that each subregion that the system supervisor renewal has been provided with self is corresponding with it;

In the steps d 1, system supervisor obtains the address information of the MO manager of new MIT subregion according to the mapping relations table of the address information of each subregion that has upgraded MO manager corresponding with it, and presses the address information transmission starting command of this MO manager.

In the steps d 2, the node identifier of all MO loads the network equipment of each MO correspondence in this new MIT subregion in the subregion that described MO manager has been provided with according to self.

Wherein, described log-on message comprises at least: partition name, root node identifier, MO manager address information; Described log-on message table comprises the mapping relations between partition name, root node identifier and the MO manager address information.

By such scheme as can be seen, key of the present invention is: each network equipment is abstracted into MO, organizes MIT according to the containment relationship between the MO, and MIT is divided into one or more MIT subregions, for whole M IT is provided with the MO zone manager, for each subregion is provided with MO Manager respectively; Each MIT subregion is abstracted into zone object, externally provide MO Manager Proxy class to encapsulate each zone object, make applications can pass through transparent each subregion of visit of MO Manager Proxy server, inner establishment, inquiry, modification and the deletion that realizes MO of each subregion; When System Expansion, the newly-increased network equipment is abstracted into MO and is organized into new MIT subregion, for this new subregion is provided with new MOManager, be registered in the MO zone manager by this new MO Manager this new subregion of startup and with it, make system accomplish seamless online dilatation in the mode of user transparent.

Therefore, the network management of the distributed MIT of support provided by the present invention, can make the disposal ability of network management system not be subject to the disposal ability of separate unit server, when the network equipment surpasses the capacity of network management system, system is expanded in the mode of user transparent.

Description of drawings

Fig. 1 is a prior art networking mode schematic diagram;

Fig. 2 divides MIT subregion one preferred embodiment schematic diagram in the inventive method;

Fig. 3 is each MIT subregion one preferred embodiment handling process schematic diagram of system start-up in the inventive method;

Fig. 4 is that applications is created MO one preferred embodiment handling process schematic diagram in the inventive method;

Fig. 5 is outside application query MO one a preferred embodiment handling process schematic diagram in the inventive method;

Fig. 6 is that applications is revised MO one preferred embodiment handling process schematic diagram in the inventive method;

Fig. 7 is applications deletion MO one a preferred embodiment handling process schematic diagram in the inventive method;

Fig. 8 is System Expansion one a preferred embodiment handling process schematic diagram in the inventive method.

Embodiment

The present invention is further described in more detail below in conjunction with drawings and the specific embodiments.

The inventive method is supported distributed MIT, and MIT is carried out partition management, and this method comprises: this supports the method for distributed MIT network to set up the method for distributed MIT and management.

The described method of setting up distributed MIT is specially: each network equipment is abstracted into a MO, with each MO as a node among the MIT, and the containment relationship of pressing between each MO is formed whole M IT with each node, for being provided with MO Manager, each subregion is used to visit the MO of subregion inside separately, for whole M IT is provided with the MO Manager that the MO zone manager is used to visit each subregion.At first, the MO that selects to be arranged in the MIT low level according to managerial demand is defined as the root node of MIT, and system is organized into root partition with these root nodes.Owing to also form tree-like relation between these root nodes, therefore the root node of a unique node as all root nodes must be arranged, then the root node of described all root nodes is the root node of this root partition.Then, all nodes under the root partition node among the MIT are carried out subregion, system can carry out subregion according to the quantity of MO, also can carry out subregion according to the attribute of MO.Here, also all nodes among the MIT all might be subdivided into root partition, then not have the node under the root partition node, so system no longer carries out subregion, whole M IT only is divided into a root partition.Wherein, described node is MO, sets in advance the node identifier of interior all MO of subregion separately among the MO Manager of described each subregion; Described MO zone manager has been preserved the log-on message table that comprises each subregion log-on message.

Here, by the quantity subregion, be about to all nodes and be evenly distributed to each subregion by quantity, press MO attribute subregion, the MO that is about to same alike result is subdivided into one or more subregions, this attribute can but specifically which kind of principle to carry out subregion by then be provided with according to managerial demand for the function of equipment, kind etc., and the present invention does not limit.Because, the level of MIT is divided from root node, the root node of whole M IT is the ground floor of MIT, it also is the lowermost layer among the MIT, the child nodes of this root node is the second layer of MIT, and by that analogy, therefore among the described MIT MO of low level just with the more approaching level of the root node of whole M IT in MO.

Fig. 2 divides MIT subregion one preferred embodiment schematic diagram in the inventive method.In the present embodiment, system's employing is N subregion by MO attribute partitioned method with the node division beyond the root partition among the MIT.Wherein, N is a nonnegative integer, if N is 0, then this system has only divided a root partition, if N is 1, then this system divides two subregions, comprise: root partition and subregion 1, if N is 2, then this system has divided three subregions altogether, comprise: root partition, subregion 1 and subregion 2, and by that analogy.As shown in Figure 2, concrete partition method is as follows:

At first, node Net, 1@Net, 2@Net, 1@1@Net, 2@1@Net are defined as root node and are subdivided into root partition, node Net is the root node of this root partition.All nodes beyond the root partition are divided into N zone by its attribute, and promptly subregion 1 is to subregion N.

As shown in Figure 2, in subregion 1, node 1@1@1@Net, 2@1@1@Net, 3@1@1@Net and 4@1@1@Net are the root node of this subregion, wherein, node 2@1@1@Net contains two node: 1@2@1@1@Net and 2@2@1@1@Net down, therefore, node 1@2@1@1@Net and 2@2@1@1@Net are two child nodes of subregion 1, and node 2@1@1@Net is their father node.In subregion 2, node 5@1@1@Net, 1@2@1@Net, 2@2@1@Net and 1@1@2@Net are the root node of this subregion.Subregion is divided to subregion N successively, and in subregion N, node 1@2@Net, 2@2@Net, 3@2@Net and 4@2@Net are the root node of this subregion, and node 2@1@2@Net is the child node of subregion N.

Then, will go in the root partition with the node division of the node direct correlation of a plurality of subregions in each subregion again.As shown in Figure 2, node 1@2@Net among the subregion N and the node 1@1@2@Net direct correlation in the subregion 2, and with subregion N in also direct correlation of node 2@1@2@Net, therefore this node 1@2@Net also is divided to root partition.Here, below above-described all nodes, may also contain a plurality of child nodes, not draw among the figure, but not influence the validity of the inventive method.

Because, the inventive method is in order to realize the transparent expansion of system, MIT is carried out partition management, node in root partition and other each subregion is managed respectively, when needs are operated certain node, obtain root node identifier with its coupling according to the node identifier of this node, thus with this root node place subregion as this node place subregion, operate at this subregion this node of inner inquiry and to this node then.If the node relevant with a plurality of subregions do not put into root partition, then can cause the child node of this node may inquire about less than.Such as: if node 1@2@Net does not put into root partition among Fig. 2, and be divided among the subregion N, then when needs are operated node 1@1@2@Net, because its father node 1@2@Net is in subregion N, the root node that then mates the node 1@1@2@Net place subregion that comes out must be the root node of subregion N, therefore system can be in subregion N query node 1@1@2@Net, but the information that does not have node 1@1@2@Net at subregion N, system just can't inquire node 1@1@2@Net like this, thereby can't operate node 1@1@2@Net.If node 1@2@Net is placed into root partition, then node 1@1@2@Net and node 2@1@2@Net become the root node of subregion 2 and subregion N respectively, when needs are operated node 1@1@2@Net or 2@1@2@Net, each node of subregion 2 or subregion N is directly inquired about by system, and the information that node 1@1@2@Net is arranged in subregion 2, at subregion N the information of node 2@1@2@Net is arranged, like this, system just can inquire all child nodes of node 1@2@Net.Therefore, MIT method for organizing of the present invention will all be divided to root partition with the node of a plurality of subregion direct correlation, and the system that guaranteed can carry out partition management to all nodes.

Based on above-mentioned distributed MIT method for building up, this supports that the method for distributed MIT network is described in detail to management in the inventive method below in conjunction with accompanying drawing, described management this support the method for distributed MIT network to comprise: each MIT subregion of system start-up, the network equipment that applications is visited each MIT subregion is promptly: creates, inquiry, revises and deletion MO, and the processing of System Expansion three parts.

Fig. 3 is each MIT subregion one preferred embodiment handling process schematic diagram of system start-up in the inventive method, and as shown in Figure 3, concrete treatment step is as follows:

Step 301: after the system start-up, system supervisor is set up MIT according to the containment relationship of grid equipment, and with MIT information configuration Composition Region file.Here, the MIT information in the partitioned file comprises that whole system is divided into the information such as title of several subregions and each subregion.

Wherein, set up distributed MIT method as seen by the above, system is each zone configuration MO Manager, sets in advance the node identifier of all MO in self corresponding subregion in the MO of each subregion Manager.As shown in Figure 3, be provided with N+1 subregion in this example altogether, comprise: by the root partition of MO Manager0 management, and comprise subregion 1 by MO Manager1 to MO ManagerN management successively, then stored the node identifier of root partition all MO to the subregion N in MO Manager0 to the MO ManagerN successively respectively to subregion N.

In addition, above-mentioned setting up in the distributed MIT method, system also is provided with the MO zone manager and is responsible for the log-on message of each subregion is stored in the log-on message table of self, when system start-up, also can start the MO zone manager simultaneously.

Step 302: system supervisor sends starting command to MO Manager0.

Here, system is provided with the mapping relations table of the corresponding MOManager address information with it of each subregion in system supervisor before startup.That is to say that system need start the MO Manager address which subregion can inquire this subregion correspondence automatically, system can send starting command to this MO Manager then.

Step 303:MO Manager0 receives this starting command, loads each network equipment of root partition according to the node identifier of all MO in self the corresponding subregion that has been provided with, thereby starts root partition.

Step 304:MO Manager0 sends log-on message to the MO zone manager, root partition is registered in the MO zone manager, in the log-on message table that self is provided with, preserves address information, the title of root partition and the root node identifier of root partition of MO Manager0 by the MO zone manager.Wherein, the root node of root partition has and only has one.Here, described log-on message table comprises the mapping relations of partition name, root node identifier and MO Manager address information.

Step 305: system supervisor sends starting command to MO Manager1.

Step 306:MO Manager1 receives this starting command, and according to the node identifier of all MO in self the corresponding subregion that has been provided with, i.e. the node identifier of all MO in the subregion 1 loads each network equipment of subregion 1, thus boot partition 1.

Step 307:MO Manager1 sends log-on message to the MO zone manager, subregion 1 is registered in the MO zone manager, in described log-on message table, preserves the address information of MOManager1, the title of subregion 1 and the root node identifier of subregion 1 by the MO zone manager.Wherein, the subregion of all except that root partition promptly the subregion 1 here one or more root nodes can be arranged to subregion N.

Then, system supervisor sends starting command to MO Manager2 to MO ManagerN-1 successively, to subregion N-1, finishes subregion 2 again to the registration of subregion N-1 in the MO zone manager by MO Manager2 to MO ManagerN-1 boot partition 2.Because boot partition 2 described here is identical to the operation of the described boot partition 1 of step 307 with step 305 to the operation of subregion N-1, therefore, does not describe among Fig. 3, but does not influence validity of the present invention.

Step 308: system supervisor sends starting command to MO ManagerN.

Step 309:MO ManagerN receives this starting command, and then according to the node identifier of all MO in self the corresponding subregion that has been provided with, i.e. the node identifier of all MO in the subregion N loads the network equipment of subregion N.

Step 310:MO ManagerN sends log-on message to the MO zone manager, subregion N is registered in the MO zone manager, in the log-on message table, preserves the address information of MOManagerN, the title of subregion N and the root node identifier of subregion N by the MO zone manager.

So far, system has finished the startup of root partition and N the subregion of distributed MIT, and in the MO zone manager, preserved the log-on message table of each subregion that has started, this log-on message table comprises at least: the mapping relations of partition name, root node identifier and MO Manager address information, thus system brings into operation and can carry out distributed management to each subregion.

In the inventive method, the method that described applications is visited the network equipment of each MIT subregion is specially: adopt OO method, the network equipment is abstracted into MO, organize MIT and MIT is divided into a plurality of subregions according to the containment relationship between the MO, each MIT subregion is abstracted into zone object, externally provide MOManager Proxy class to encapsulate each zone object, and MO Manager Proxy server is set.Externally during each subregion of application access, at first, applications sends access request to MO Manager Proxy server, this access request of MO Manager Proxy server forwards is to the MO zone manager, the MO zone manager obtains the address information of the corresponding MO Manager of subregion to be visited according to access request, and returns this MOManager address information and give MO Manager Proxy server; MO Manager Proxy server is sent to this access request by this MO Manager address information the MOManager of subregion correspondence to be visited again; At last, the corresponding MO Manager of subregion to be visited carries out corresponding operating according to this access request in corresponding subregion.Described access request can be establishment, inquiry, modification or removal request.Like this, applications can be passed through transparent each subregion of visit of MO Manager Proxy server, again by inner establishment, inquiry, modification and the deletion that realizes MO of each subregion.

Wherein, described applications transparent access, in other words, encapsulation by MO Manager Proxy class, applications can be with the form visit zone object of method call, and applications does not need to know what software is the zone object that will visit be present in, use what hardware platform and where be positioned at network.

In addition, described MO Manager Proxy server can be based on the C++ interface, also can be based on corba interface.If based on the C++ interface, then support applications visit based on C++; Because the CORBA system has defined the map standard of its interface to C, C++, Smalltalk, Ada95, JAVA etc. at present, if therefore based on corba interface, then support applications visit based on the above C, C++, Smalltalk, Ada95, JAVA etc.MOManager Proxy server that is to say that based on corba interface applications can be by each subregion of MO Manager Proxy server access MIT based on corba interface in the following stated embodiment of the invention.

Fig. 4 is that applications is created MO one preferred embodiment handling process schematic diagram in the inventive method, and as shown in Figure 4, its concrete treatment step is as follows:

Step 401: applications sends request to create to MO Manager Proxy server, and a sub-MO node is created in request under certain father node, comprises the node identifier of MO to be created in this request at least.

Here, also can ask at certain partition creating root node, then in this request to create except comprising the node identifier of MO to be created, also comprise the title of MO to be created place subregion at least.

Step 402:MO Manager Proxy server is forwarded to the MO zone manager with this request to create.

The node identifier of MO to be created is discerned the root node identifier of the direct father node of this MO to be created place subregion in the step 403:MO zone manager basis request to create that receives, and the log-on message table of self preserving according to this root node identifier inquiry obtains the MO Manager address information of the direct father node of this MO to be created place subregion correspondence then.

Here, if to be created be the partition root node, then the MO zone manager log-on message table that inquiry self is preserved according to the partition name in the request to create that receives obtains the MOManager address information of corresponding subregion.

Among the MIT of network management system, a rule is all followed in being provided with of each node identifier: the identifier that comprises its all father nodes in the identifier of child node, that is: both comprised the identifier of its direct father node in the identifier of a MO, the identifier that also comprises its direct father node place partition root node, such as: comprise the identifier of its direct father node 2@3@Net in the identifier of child node 1@2@3@4@Net, also comprise the identifier of the root node 3@4@Net of its direct father node place subregion; Use the node identifier 1@2@3@4@Net of MO to be created when the MO zone manager, with all root node identifiers in the log-on message table of himself storage when mating, to select and the root node of the immediate root node of MO to be created as the direct father node place subregion of this MO, specifically: if a plurality of root node identifiers such as 3@4@Net, 4@Net and Net are arranged, all be included in the identifier of this MO, then select the root node as its direct father node place subregion with the immediate node 3@4@Net of this MO; Therefore, the MO zone manager can be discerned the root node identifier of its direct father node place subregion according to the node identifier of MO to be created here.

Step 404:MO zone manager returns to MO Manager Proxy server with the resulting MO Manager of step 403 address information.

Step 405:MO Manager Proxy server is sent to this MO Manager according to the MOManager address information that receives in the step 404 with described request to create.

Step 406:MO Manager is according to the node identifier of MO to be created in the request to create that receives, the node identifier of interior all MO of self corresponding subregion that inquiry has been provided with, obtain the direct father node identifier of this MO to be created, and under this father node, create this MO.

Wherein, described establishment MO comprises: MO Manager loads the network equipment of this MO correspondence, upgrade the node identifier of all MO in self corresponding subregion of being preserved among the MO Manager, in original node identifier information, add the node identifier of this new establishment MO, simultaneously, preserve the MO database of information in the update system, in database, add the information of this new establishment MO.Here, in network management system, can be used to preserve each MO information of place subregion for each MIT subregion is provided with database, described MO information comprises information such as the attribute of node identifier, the network equipment of this MO correspondence and current state.

Wherein, during described more new database, at first self MO identifier, attribute and current state etc. are reported to MO Manager Proxy server by the network equipment of this new establishment MO correspondence, MOManager Proxy server finds the address information of the MO Manager of this new establishment MO place subregion in the MO zone manager according to this MO identifier, and then the MO information that this network equipment reports is forwarded to corresponding MO Manager, MO Manager is sent to database with this MO information and upgrades thus at last.Here, the address of the pre-configured MO ManagerProxy server of meeting in each network equipment, thus each network equipment can report MO information to MO ManagerProxy server; Database address information that also can pre-configured corresponding subregion separately among each MO Manager, thus can upgrade database.Because the described processing of this section does not belong to emphasis of the present invention, therefore be not described further here.

In addition, if to be created be the partition root node, then step 406 is: MO Manager is according to the node identifier of MO in the request to create that receives, and directly at this partition root node M of this partition creating O, execution in step 407 again; Step 407:MO Manager sends log-on message to the MO zone manager, the partition root node M O that this is to be created is registered in the MO zone manager, by the partition name inquiry and registration information table of MO zone manager according to this MO correspondence to be created, obtain the log-on message of this subregion correspondence, and the log-on message of this subregion upgraded, the node identifier of the MO that this is to be created is added in the root node identifier information of this subregion, thereby finishes this establishment.

Fig. 5 is outside application query MO one a preferred embodiment handling process schematic diagram in the inventive method, and as shown in Figure 5, its concrete treatment step is as follows:

Step 501: applications sends query requests to MO Manager Proxy server, and this request comprises at least: querying condition and query contents.

Wherein, described querying condition comprises query context at least, and specifically: when the MO node of one or more appointments was inquired about in request, described query context can be the identifier of these one or more MO nodes; When request qualified MO node of inquiry in certain scope, described query context can be for representing the identifier of this query context, such as, when query context is decided to be all MO nodes of specifying under the father node, this query context is the identifier of father node for this reason, when query context was decided to be all MO nodes in specified partition, this query context information is the root node identifier of subregion for this reason.Described querying condition can also further comprise the query contents prescribed information, such as, for qualification of the relevant information value of MO institute map network equipment to be checked etc.Described query contents is the information of the applications MO that meets querying condition that need obtain, can comprise: the node identifier of these MO, and relevant information such as attribute, the state etc. of institute's map network equipment.The concrete establishing method of the above querying condition and query contents will determine that the present invention does not do concrete qualification according to the realization situation of system.

Step 502:MO Manager Proxy server is forwarded to the MO zone manager with this query requests.

Step 503:MO zone manager obtains query context according to the querying condition in the query requests that receives, can obtain the root node identifier of each MO place subregion to be checked according to query context, and obtain the address information of the MO Manager of each MO place subregion correspondence to be checked according to these root node identifier inquiry and registration information tables.

Step 504:MO zone manager returns to MO Manager Proxy server with the address information of all MO Manager that step 503 obtains.

Step 505:MO Manager Proxy server is the address information of the MO Manager of 504 each the MO place subregion correspondence to be checked that obtain set by step, sends this MO query requests to these MO Manager.Here, present embodiment adopts broadcast mode to send query requests to each MO Manager, can certainly adopt other modes to send query requests to MO Manager, and the present invention does not limit.

In the present embodiment, each MO to be checked is distributed in root partition and subregion N, so MO ManagerProxy server is broadcasted this query requests to MO Manager0 and MO ManagerN.Here, each MO to be checked also may be distributed in all subregions from root partition to subregion N, and then MOManager Proxy server is to all MO Manager broadcast query requests of system; Each MO to be checked also may be distributed in one or more subregion, and then MO Manager Proxy server is to one or more MO Manager broadcast query request of correspondence.

Step 506: the MO Manager of each MO place subregion correspondence to be checked by broadcast reception after query requests, the node identifier of all MO in self the corresponding subregion that has been provided with according to the query context in the querying condition of this query requests inquiry obtains the node identifier of all MO in the query context; Then by the identifier inquiry system of these nodes for this subregion setting be used to preserve the MO database of information, and obtain separately all MO information to be checked in the corresponding subregion of institute according to query contents prescribed information in this querying condition; From then on extract the MO information of the MO correspondence to be checked in the corresponding separately subregion as Query Result in the database by the query contents in this query requests at last.Here, because the database address information of pre-configured subregion separately among each MOManager, therefore described MO Manager can extract MO information from database.

Step 507: the MO Manager of each the MO place subregion to be checked Query Result that extracts of step 506 separately returns to MO Manager Proxy server.

The Query Result that each MO Manager that step 508:MO Manager Proxy server will receive returns is organized, and obtains whole Query Results and returns to applications.

Fig. 6 is that applications is revised MO one preferred embodiment handling process schematic diagram in the inventive method, and as shown in Figure 6, its concrete treatment step is as follows:

Step 601: applications sends to revise asks the server to MO Manager Proxy, and this modification request comprises at least to be waited to revise the node identifier of MO and revise content.

Step 602:MO Manager Proxy server should be revised request and be forwarded to the MO zone manager.

Step 603:MO zone manager waits that according in the modification request that receives the node identifier of revising MO obtains the root section identifier that this waits to revise MO place subregion, and the log-on message table of being preserved according to this root node identifier inquiry obtains the address information that this waits to revise the corresponding MO Manager of MO place subregion then.

Step 604:MO zone manager returns to the Proxy server with the address information of the MO Manager that step 603 obtains.

Step 605:Proxy server is the address information of the 604 MO Manager that obtain set by step, again this MO is revised request and sends MO Manager so far.

After step 606:MO Manager receives the request of modification, according to waiting that the node identifier inquiry system of revising MO is used for preserving the MO database of information in this modification request, obtain the MO information that this waits to revise the MO correspondence, according to revising content this is waited that the MO information of revising the MO correspondence makes amendment again.Here, because the database address information of pre-configured subregion separately among each MO Manager, therefore described MO Manager can make amendment to the MO information in the database.

Fig. 7 is applications deletion MO one a preferred embodiment handling process schematic diagram in the inventive method, and as shown in Figure 7, its concrete treatment step is as follows:

Step 701: applications sends removal request to MO Manager Proxy server, and this request comprises the node identifier of waiting to delete MO at least.

Step 702:MO Manager Proxy server is forwarded to the MO zone manager with this removal request.

Step 703:MO zone manager waits that according in the removal request that receives the node identifier of deleting MO obtains the root node identifier that this waits to delete MO place subregion, and the log-on message table of being preserved by this root node identifier inquiry obtains the address information that this waits to delete the corresponding MO Manager of MO place subregion then.

Step 704:MO zone manager returns to MO Manager Proxy server with the address information of the MO Manager that step 703 obtains.

Step 705:MO Manager Proxy server is the address information of the 704 MO Manager that obtain set by step, again this removal request is sent MO Manager so far.

After step 706:MO Manager receives removal request, according to waiting in this removal request that the node identifier Query Database of deleting MO obtains the MO information that this waits to delete the MO correspondence, and with this MO information deletion, simultaneously also with the node identifier deletion of this MO of preserving among the MO Manager.Here, because the database address information of pre-configured subregion separately among each MO Manager, therefore described MO Manager can delete MO information from database.

Wherein, described deletion not only comprises MO information and the node identifier of deleting this MO correspondence, if the leaf node of the non-MIT of this MO, that is: this MO also contains child nodes in MIT, then described deletion also comprises the MO information and the node identifier of the whole subtree correspondence of deletion under this MO, that is: the MO information and the node identifier of all child nodes correspondences of deleting this MO and being contained.

In addition, if MO to be deleted is the partition root node, then also further comprise after step 706: MO Manager sends log-off message to the MO zone manager, the MO zone manager waits to delete node identifier this MO to be deleted of deletion from the log-on message table of self preserving of MO according in this information this, thus the MO partition root endpoint unregistration that this is to be deleted.

Visit the network equipment of each MIT subregion by above-described applications, promptly create, inquiry, the processing of modification and deletion MO as seen, when applications need be operated the MO of MIT inside, needn't be concerned about that it specifically is in that position of network, run on factors such as which kind of workbench, only need provide the whole network of MO to unify identifier, that is: the node identifier among the MIT, MO ManagerProxy server just can be sent to access request the MOManager of this MO place subregion according to this identifier, handle in subregion inside according to this access request by MO Manager, and these operations all are transparent for other subregions, can not impact other subregion and other MO.

What the inventive method realized is the partition management of the network equipment, and each subregion is encapsulated as zone object, described System Expansion can carry out so that the mode of user transparent is online, that is to say: system realizes dilatation by the new MIT subregion of configuration on original MIT basis, and the access of this subregion is sightless for system's other parts, can the operation of system's other parts not impacted.

Fig. 8 is System Expansion one a preferred embodiment handling process schematic diagram in the inventive method, and as shown in Figure 8, its concrete treatment step is as follows:

Step 801: system supervisor is the newly-increased new partitioned file of network equipments configuration.Here, the information in the partitioned file comprises the information such as title of new subregion.

Wherein, system also has been new zone configuration MO Manager, having set in advance self corresponding subregion among the MO Manager is the node identifier of all MO in the new subregion.

Step 802: system supervisor sends starting command to MO Manager.

Here, before System Expansion, system supervisor upgrades the mapping relations table of the corresponding MOManager address information with it of original each subregion, has added the mapping relations of the corresponding MO Manager with it of new subregion address therein.That is to say that system can obtain the MO Manager address information of this new subregion correspondence in advance in the time of need starting new subregion, can send starting command to this MOManager by this address information then.

Step 803:MO Manager receives this starting command, loads the network equipment of each the MO correspondence in the new subregion according to the node identifier of all MO in self the corresponding subregion that has been provided with, thereby starts new subregion.

Step 804:MO Manager sends log-on message to the MO zone manager, new subregion is registered in the MO zone manager, in the log-on message table that self is provided with, preserves the title of the address information of MO Manager, new subregion and the root node identifier of new subregion by the MO zone manager.

Wherein, system also can be provided with the MO information that new database is used to preserve each MO correspondence of new subregion for new subregion, after new subregion starts, each network equipment in the subregion also will be as stated above MO Manager by MOManager Proxy server and new subregion will MO information separately be sent to database and preserve.

As seen from the above description, when the network management system dilatation, can on original MIT basis, dispose new MIT subregion, and new MO Manager is installed new subregion is managed, but new subregion and all shared original network management system of original MIT subregion are come management equipment, therefore can realize the whole network positioning problems; In addition, in the process of the new subregion of system start-up, do not relate to operation, therefore can the operation of existing network not had any impact, thereby can make system with the online dilatation of the mode of user transparent for other subregion in the existing network.

In the methods of the invention, system adopts distributed architecture to realize the distribution character of MIT, each subregion may operate in different processes or server, can realize transparent communication between different processes or the server, system can adopt the CORBA framework to realize the distribution character of MIT, also can adopt other distributed structure/architecture such as Component Object Model (COM) to wait and realize, perhaps adopt the distributed low layer architecture of a lightweight of oneself realizing to realize.What the above embodiment of the invention adopted is the CORBA framework.

In sum; adopt the inventive method embodiment can make network management system support distributed MIT management; by partition management to MIT; make the processing distribution of system core data is carried out; lower the system loading of concentrating; the overall operation efficiency of raising system; the distribution management of simultaneity factor core data; also improved system reliability; the system that makes carries out local online upgrading in the mode of user transparent becomes possibility; avoided shutting down the loss that dilatation brings to operator, also avoided other increase by to overlap the whole network positioning problems difficulty that network management system is independently brought because of webmaster.

Claims

1. a Network Management method of supporting distributed management inforamtion tree MIT is characterized in that, comprising: set up distributed MIT method and the management this support the method for the network of distributed MIT;

The described method of setting up distributed MIT comprises: each network equipment is abstracted into management object MO, organize MIT according to the containment relationship between the MO, and MIT is divided into one or more MIT subregions, for being provided with the MO manager, each subregion is used to visit the MO of subregion inside separately, for whole M IT is provided with the MO manager that the MO zone manager is used to visit each subregion;

2. method according to claim 1 is characterized in that, sets up in the method for distributed MIT, describedly MIT is divided into one or more MIT partitioned method is:

3. method according to claim 2 is characterized in that, after the step a2, also further comprises: will divide in each the described subregion obtain through step a2, and be subdivided into root partition with the MO of the MO direct correlation of an above subregion.

4. method according to claim 1 is characterized in that, management supports in the method for network of distributed MIT, and the MO manager of described each subregion starts separately that partitioned method is:

5. method according to claim 4, it is characterized in that, among the step b1, the mapping relations table of the address information of the MO manager that each subregion that system supervisor has been provided with according to self is corresponding with it obtains the address information of each subregion MO manager, and sends starting command by the address information of this MO manager.

6. method according to claim 4 is characterized in that, among the step b2, described MO manager loads the network equipment of each MO correspondence in the subregion of described self correspondence according to the node identifier of all MO in self the corresponding subregion that has been provided with.

7. method according to claim 1, it is characterized in that, described management this support the method for the network of distributed MIT further to comprise: each MIT subregion is abstracted into zone object, externally provide MO manager proxy class to encapsulate each zone object, and MO manager acting server is set in system;

8. method according to claim 7 is characterized in that, described access request is a request to create;

9. method according to claim 8 is characterized in that, when described MO to be created was the partition root node, described request to create further comprised the title of MO to be created place subregion; Described MO to be visited place subregion is a MO to be created place subregion;

10. method according to claim 9, it is characterized in that, when described MO to be created is the partition root node, after described step c5, further comprise: the MO manager of MO to be created place subregion sends log-on message to the MO zone manager, this MO to be created is registered in the log-on message table of MO zone manager.

11. method according to claim 7 is characterized in that, described access request is a query requests; Described MO to be visited is MO to be checked; The node identifier of described MO to be visited is represented query context;

12. method according to claim 11, it is characterized in that, after described step c5, further comprise: the Query Result that the MO manager of MO to be checked place subregion will be extracted separately returns to MO manager acting server, and MO manager acting server is organized each Query Result that receives and obtained whole Query Results and should return to applications by whole Query Results.

13. method according to claim 7 is characterized in that, described access request is for revising request; Described MO to be visited is MO to be revised;

14. method according to claim 7 is characterized in that, described access request is a removal request; Described MO to be visited is MO to be deleted;

15. method according to claim 14 is characterized in that, when MO described to be deleted is the partition root node, after described step c5, further comprises:

16. method according to claim 7 is characterized in that, described MO manager acting server is based on C++ or common object request broker system corba interface.

17. method according to claim 1 is characterized in that, manages in the method for this network of supporting distributed MIT, the method for described System Expansion is specially:

18. method according to claim 17 is characterized in that, this supports the Network Management method of distributed MIT further to comprise, the mapping relations table of the address information of the MO manager that each subregion that the system supervisor renewal has been provided with self is corresponding with it;

19. method according to claim 17 is characterized in that, in the steps d 2, the node identifier of all MO loads the network equipment of each MO correspondence in this new MIT subregion in the subregion that described MO manager has been provided with according to self.

20. according to claim 1,4,7,9,10,15 or 17 described methods, it is characterized in that,

Described log-on message comprises at least: partition name, root node identifier, MO manager address information;

Described log-on message table comprises the mapping relations between partition name, root node identifier and the MO manager address information.