JP2000151743A - Method for designing network control system - Google Patents

Abstract

PROBLEM TO BE SOLVED: To easily set up component ware by standardizing a display and editing method for data related to component ware by making it possible to access the component ware through a data access interface standardized for plural connected component ware. SOLUTION: Not only syntax appropriateness but also semantic appropriateness is required for connection between components. Restriction information included in each component itself includes not complete but considerably large part of semantic appropriateness. In the case of mutually connecting two components, mutual restricting conditions are checked. When an output side restricting condition is more severe than that of an input side restricting condition, connection processing is executed. When the former condition is not so severe, inter-component connection is rejected. When the procedure is inserted into a tool for connecting between components, a bug generated due to the contradiction of data to be transferred between components can be removed on the stage of Pos design.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to various communication devices (network element: N
The present invention relates to a method for designing a network control system for maintaining, operating, and managing E), and more particularly, to a method for designing a network control system using component orientation.

[0002]

2. Description of the Related Art Software is changing from an era of making from the beginning to an era of combining software (Mikio Aoyama: "Software Big Bang", Information Processing, Vol. 39, No. 4, pp. 3).
38-341, 1998). In the communication network management system,
Prepare plug & play software components (componentware) that can be used as is without compiling,
If you can build a system with that combination,
Both development time and cost can be reduced. Tel
SMAR at eManagement Forum
T TMN (TeleManagement Foru)
m SMARTTMN Work Project, ht
tp: //www.tmforum.org/pages/smart.html) and others are aiming in that direction, but there are still many issues that need to be solved technically before reaching coding-less.

[0003] CORBA (OMG IDL Ver.
2.2 Feb-1998), DCOM (Kazuo Koyama, "D
COM Guidebook ", Ohmsha, 1997) and Ja
vaBeans (JavaBeans ^™ API sp
education 1.01 SUN Micr
osystems, 24-Jul-1997) has become an industry standard, and the ability to acquire specifications such as interface repositories and introspection has been made available in a limited manner, so that components created by different organizations can be combined. Also, the division of labor between tool makers and component makers has become possible.

[0004] The network control system (OpS) maintains, operates, and manages various communication devices constituting a communication network. As a method of designing the OpS, a method using componentware is described in Ikuo Yoda and Koji Yada, "Component Oriented Communication Network Management System" (Proceedings of the IEICE Society Conference 1998, B-14-18,
pp. 565, The Institute of Electronics, Information and Communication Engineers, 1998.
9.7), Taichi Kawahata and Ikuo Yoda, "Design and Implementation of Data Conversion Components for Communication Network Management System" (Proceedings of the 1998 IEICE Society Conference, B-14-19, p.
p. 566, The Institute of Electronics, Information and Communication Engineers, 1998.
7), Hiroyuki Maedo, Tetsuaki Goto, and Masahisa Tago, "Design and Implementation of Protocol Components for Communication Network Management System" (1998)
Proceedings of IEICE Society Conference 2006-B-14
20 pp. 567, The Institute of Electronics, Information and Communication Engineers, 199
8.9.7), Taichi Kawahata and Koji Yada, "Prototype of Component-Oriented 0pS" (Preprints of IEICE Society Conference, 1998, B-14-21, pp. 10-27).
568, published by the Institute of Electronics, Information and Communication Engineers, 1998.9.7).

FIG. 1 shows an outline of a typical configuration of an OpS. FIG. 1A shows a hierarchical system, and FIG. 1B shows a component system.

[0006] In the design of a component type system,
Components called componentware (CW) are used.

FIG. 2 shows a flow of designing a conventional component type OpS. A plurality of types of general-purpose componentware (components) are previously stored on a pallet 1.
2 are prepared. Of these components, a component 2 necessary for actually creating an OpS is selected, the selected component 2 is appropriately customized, and the components are connected to each other.

This method also solves the problem of protocol diversification by componentizing the protocol differences.

FIG. 3 shows how data is set in the above-described OpS design. For setting of various data in the component 2, a setting area called a property sheet 3 is opened, and an expression specific to each data is provided there. This is done by describing the value. Also, data exchanged between the components 2 can be displayed and monitored (4), but this display is also performed in a representation method unique to each data.

In such an OpS, a user 5 operates a network element (NE) 7 through an application 6 as shown in FIG. The application 6 that manages the network element 7 is configured by connecting a number of components 2, and the user operates (GUI operation) using buttons 8 and a graph display 9 that are GUI components.

An event from the network element 7 is exchanged and processed between the components 2 in the application 6, and is transmitted to the user through a graphic display 9, and a user's instruction is transmitted to the network element 7 through a button 8.

[0012]

(A) Unified interface to structured data exchanged between components Traditional management protocols such as CMIP and SNMP, CO
HTML + Apple in addition to the distribution framework such as RBA
The protocols handled by OpS, such as the user interface by t, have been diversified. In addition, many network devices such as routers use HTML to perform detailed settings via a browser, and these are convenient to manage alone, but centrally manage many network devices. If it is desired to do so, it is necessary to regard HTTP as a new management protocol and perform machine processing.

In the case of handling these various management protocols, it has hitherto been necessary to create a program while being aware of the structure of the protocol data unit (PDU) of each management protocol. For example, in CMIP and SNMP, ASN. Data is described using IDL in CORBA, and the coding method and the mapping to the computer language are different from each other. Providing a dedicated component group for each of these protocols not only requires designing a large number of components, but also requires developers to handle many components with different ideas, increasing the burden. .

When implementing a management protocol in a computer language, it is common sense that the PDU structure is mapped to the data structure in the computer language so that the processing can be performed most efficiently. Should respect the data structures used in the existing protocol stack. To share components with different protocols,
If the data passed between the components is easily converted to a common data structure, there is a possibility that information necessary for protocol processing may be lost and communication may become impossible.

In the method of designing a network control system for componentware, debugging, inputting of values, and setting of values in a component property sheet are very troublesome, and it has been recognized that it is necessary to improve these.

(B) Utilization of GUI Interface of Managed Device in Centralized Network Management In the future, the user interface used for network management will not be an application whose development cost is high but will be inexpensive http + applet. Devices to be managed also provide a management method based on http. For this reason, even if a large number of network devices are collectively managed by the conventional method, it is desirable to maximize the management method provided by each device when managing each device.

(C) Data Type and Guidance One of the factors that makes it difficult to reuse software components is a mismatch in the data types to be delivered between components (G
ordon S. Novak Jr. "Softwar
e Reuse by Specialization
of Generic Procedures th
route Views ”, IEEETrans on
Software Engineering, vo
l. 23, no. 7, pp. 401-417, Jul
y, 1997).

On the other hand, strong typing, which is checked by a compiler or the like, is a powerful means for appropriately guiding developers and preventing program errors (NMF040, 4).
1, 42, 43 TMN / C ++ API, 1997). The data structure of a protocol handled in communication network management is often complicated, and such guidance is essential.

However, this data type is generally different for each information definition, and if an attempt is made to design a general-purpose component that can be arbitrarily combined, it becomes necessary to abstract the data type by eliminating the typing. In this case, the developer has to check whether data can be passed at runtime, and the burden on the developer is further increased. Therefore, an induction mechanism instead of typing is required.

(D) Scanning of data exchanged between components In the related art, data exchanged between components is collectively transferred. Therefore, HTML data and Linked Re
ply, ASN. As in the case of one set of data or the like, the data as a whole is one. However, when it is desired to decompose the data into individual elements and process the data individually, it is difficult to cope with the conventional method.

(E) Problems when constructing a 1: n relationship with components on a network One of the advantages of the conventional component technology is that the components are dynamically executed while assembling with tools to confirm the operation. That is possible. However, when this technique is applied to OpS construction, in a 1: n relationship such as a manager agent or a server client that often appears in OpS, an unspecified number of states appear in n. . Therefore, it has been difficult to dynamically simulate such a relationship on a tool.

[0022]

(A) A unified interface to structured data exchanged between components Various data actually exchanged between components (various data structures such as ASN.1 and XML) ), And determined a uniform tree-structured interface for structured data exchanged between components. As a result, the data itself can be accessed with a unified tree-type data access interface, although the data itself is different. For example, it is possible to make various settings of CW used in the management system with a tree structure menu, and when monitoring data exchanged between components, it is easy to see the data in a unified format Now you can.

(B) Use of a GUI Interface of a Managed Device in Centralized Network Management When individual network devices are connected, the
The system that manages the device fetches the HTML file output by the p server as management data. Based on the current connection / setting state of the device, processing such as partial modification or addition of an initial setting value from the beginning is performed on this data. After the processing, the data is converted to HTML again and transferred to the browser for the administrator of the network system.

(C) Data Type and Guidance In order to properly guide the connection between the components, a constraint is introduced for the data that can be exchanged between the components, and based on the constraint, whether or not the connection between the components is possible (induction). ). There are the following two methods for inducing this connection.

(1) Information (constraint information) on possible values of a type is obtained from each component, separately from the type information of the output / input data. Then, only when the output constraint condition of the data output side component is equal to or stricter than the input constraint condition of the data input side component, the component is guided to allow connection.

(2) Constraint information can be obtained not only for components but also for data input and output from components. In this case, the constraints on the data output from the output component are not known until the component actually moves. For this reason, while the components are being assembled with the tool, the data is first flowed to the component that actually outputs the data. According to the constraints of the flowing data, the output constraint information of the component itself is made stricter so that the component can be connected to the input side component. That is, this method is meaningless unless it can be "simulated while assembling".

When the connection between components is determined, the output constraint of the data output component is too strict compared to the input constraint of the data input component. The data may be more stringent than the output component's own output constraints. At this time, it is apparently impossible to connect the components, but the components can actually be connected. Here, in order to actually enable the connection of the component itself, it is necessary to obtain not only the output constraint information from the output component but also the output constraint information from the output data flowing from the output component. .

(D) Scanning of data exchanged between components A component that scans huge data as a plurality of finer data is adopted. At this time, there are prepared a component that reflects the result of processing by the component on each piece of data that is scanned and flowed back to the original huge data, and a component that does not do so.

(E) Problem in constructing a 1: n relationship with components on a network: With a component at a demarcation point of 1: n as a boundary, an instance of a component block on the n side from the demarcation point is defined as a demarcation point. Many components are created and initialized when simulating. The demarcation point component also performs a process of appropriately distributing data on the server or manager side to the client or agent side, and vice versa. In this way, it is possible to appropriately simulate the 1: n relationship of the network even on the tool.

[0030]

BEST MODE FOR CARRYING OUT THE INVENTION

[0031]

Embodiment 1 The problems to be solved by the inventions described in claims 1, 2 and 3 can be summarized as follows. In order to implement software components in an application domain for communication network management, data exchange between components is required. On the other hand, the following is required.

1) Design of an abstract data structure that can also include the PDU structure of a newer management protocol that will appear in the future 2) Creation of a new developer guidance mechanism to replace the typing lost by abstraction 3) Protocol-specific Respect for implementation In the present embodiment, it is assumed that fine-grained components without distributed functions are targeted. That is, each component operates in the same process, and data exchange between components is passed as an argument of a procedure call.

Data exchanged between components connected in the same process can be an object (event object) having a certain function. In order to satisfy the above conditions, the event object holds the data structure unique to each management protocol as it is and implements an access interface independent of the protocol.

FIG. 5 shows an outline of the access interface of the present invention.
Access interface 12 independent of protocol
Is implemented, and the event object 13 is configured so that the components 14 and 15 can be exchanged while maintaining a data structure unique to the protocol.

Hereinafter, an example of the design and implementation of the access interface of the present invention will be described. 1. Access interface 1.1. Arrangement of PDU data structure The information definition method used in communication network management is GDMO.
/ ASN. 1 (CMIP), MIB / ASN. 1 (SN
MP), IDL (CORBA), DTD: Docume
nt type Declaration (HTML,
SGML, XML << Extensible Marku
p Language (XML) 1.0W3C Re
c. 10-Feb-1998. http://www.w3.org/TR/R
EC-xml >>, SQL (RDBMS) and the like. The interface of the event object must be able to manipulate all data structures that can be represented using these techniques. It is also desirable to be as simple as possible.

First, the data structure can be broadly divided into two types: a basic type and a composite type obtained by combining the basic type. The basic type is a type such as an integer type or a character string type that can be represented by an atomic type in a normal computer language. There is not much difference in the basic type in the definition method of each protocol. Since the basic type can be directly represented in a computer language, it is sufficient to prepare a method for changing for each atomic type and a method for acquiring as an access interface.

On the other hand, in the composite type, different expression methods are used in each protocol. In the present invention, Co represents a set of objects in object-oriented design.
lection (Dennis de Champe
aux, Douglas Lea, and Penel
op Faure, "Object-Oriente
d System Development ", Add
(Ison Welsy, 1993).

FIG. 6 shows the types of Collection.
As shown in FIG. 5, a set that allows the same element / a set that does not allow the same element (Bag / Set), an ordered set (Sequence) as a kind of Bag, and k as a kind of Set
Set (Ma) having a set of eye and value as an element
p) and the like.

In the present invention, the composite type in which different types are arranged is referred to as M
ap and sort the same type
e is proposed. What is classified into Map is called a structural type, and what is classified into Sequence is called an array type.

In addition, there is a characteristic data type of the protocol that cannot be applied to the concept of Collection. This is the ASN. 1 in ANY (Ope
nType), any in IDL, and the like. This type needs an interface for acquiring and setting type information at the time of execution, and this type is called an ANY type.

By summarizing the above, there are four types shown in Table 1.

[0042]

[Table 1]

1.2. Constraints If the data structure is abstracted as a set of objects,
Developers cannot be guided at all by the types that exist in the conventional API. Introduce constraints as an alternative. The constraint is acquired from the event object with information representing conditions such as the type of each data, a possible value, the number of elements, and the presence of elements.

The range of application of the constraints is further expanded, and a part of the data structure is represented by the constraints. For example, ASN. 1 CH
The union of OICE and IDL is a complex type in which one of a plurality of type candidates is selected and used.
These types are interpreted as placing restrictions on the structural type such that only one of the elements is present. That is, S
EEQUENCE and CHOICE, structure and unity
For on, the same interface in structure,
The difference in meaning is expressed by the constraint. As described above, the type of interface can be reduced by interpreting a difference that has been conventionally classified as a structure as a difference in constraint.

1.3. Guidance information / metadata The information definition does not include the type name, the name of the structural type element, or the name given to the numerical value, etc. Have been. In order to properly guide the developer using the tool, it is necessary that the tool can extract the information together with the constraint information and the like from the event object itself.

On the other hand, an event object is usually a temporary object generated at the time of execution, and it is conceivable that an instance does not exist when components are combined. Therefore, information about the interface is collected into one object as metadata,
It is necessary to be able to search metadata by the name of the data structure. 2. Interface Definition An example in which the framework of the access interface based on the above concept is defined by Java (JDK1.2) is shown below. Col is the Core API of JDK1.2
Selection has been introduced and is recommended as a common interface regardless of the application. JD
In K1.2, Sequence is called List. An important point of the JDK Collection is that the interface and the implementation are clearly separated, and the interface of the present invention can be used indistinguishably from the general-purpose library prepared in the Core API.

In addition, the same Java Virtual M
As an interface for exchanging data within the “acine”, there is Info Bus (), which is used for setting acquisition of an atomic type. The Info Bus defines a Collection API as one of the APIs of the exchange data, and the interface of the present invention is Info Bu.
s is satisfied.

2.1. Data Interface Table 2 shows the interface used for data access. What is newly defined is Pdu Dat
a, Pdu Any Data only,
The results are shown in Tables 3 and 4.

[0049]

[Table 2]

[0050]

[Table 3]

[0051]

[Table 4]

2.2. Metadata interface Type information is a type of constraint, but since the type is the basic structure related to the creation and deletion of objects, it is separated from the constraint information and incorporated into metadata. Tables 5 and 6 show interfaces for accessing metadata.

[0053]

[Table 5]

[0054]

[Table 6]

2.3. Restriction / Guidance Information The types of restrictions given to data are endless, and there is no general description method. ASN. Sub type of 1 and X
Table 7 summarizes the constraints, with the goal of completely expressing the constraints that can be represented by the ML DTD. In this table,
Constraint information is represented by Map, and key means meaning, value
Represents a value. Table 8 shows information for guidance.

[0056]

[Table 7]

[0057]

[Table 8]

3. Implementation and Evaluation Regarding the access interface of the present invention, ASN.
1 shows an example in which implementation in Java is performed on XML and XML. ASN. 1 and XML have similar expressions,
ASN. 1 and SGML which is the predecessor of XML in some cases. However, since a completely different API is used in a normal implementation, this is an optimal example for testing the effect of the access interface.

ASN. For 1, the access interface is put on the original implementation built on Java. For XML, DOM (Document Object Model) which is being advanced by W3C as an access interface for XML and HTML.
(Document Object Model (DO
M) Level 1 Specification V
er. 1.0 W3C Rec, 1-Oct-199
8. Overlay the access interface on top of the implementation by http://www.w3.org/TR/REC-DOM-Level-1).

Table 9 shows the ASN. Table 10 shows the correspondence between the type 1 and the interface of the present invention, and Table 10 shows the correspondence with XML elements.

[0061]

[Table 9]

[0062]

[Table 10]

ASN. 1, the data structure is strictly defined in advance, whereas XML has a D
Even if there is no knowledge of TD, it is necessary to treat this as a well-formed document. In the DOM, the structure is represented by a tree having document elements (elements, attributes, texts, etc.) as nodes. One node has unique information (eg, tag names of the elements) and subordinate nodes. In addition, any type of node can be set as a subordinate node so as not to depend on the DTD. Therefore, as shown in FIG. 7, one node is connected to three access interfaces (structure type representing node-specific information, subordinate (Array type representing a group of nodes) and a structural type representing the type of a subordinate node.

FIG. 8 shows how data is set according to the present invention. Various settings for each component 2 are made by a menu 22 having a tree structure via a common interface 21 based on the access interface described above.

Further, a data editing component is produced, and ASN. 1. It was confirmed that the same component can be edited for both XML and XML. At this time, it was found that the user can be guided to some extent by appropriately displaying the constraint information. ASN. 1. An example of each XML editing screen is shown in FIGS. 9 and 10, respectively. 4. Utilization of constraints for guidance on a tool The OpS design method of the present invention described in claims 2 and 3
These are shown in FIGS. 11 and 12, respectively.

Conventionally, only the type was used to check the connection between two components placed on the editor as shown in FIG. However, the connection between components should be asked for semantic validity on top of syntactic validity, and the constraint information of the component itself includes a considerable part, if not all, of semantic. I have.

Therefore, in FIG. 11, when two components are connected, mutual constraint conditions are confirmed. If the output-side constraint is stricter than the input-side constraint, a connection process is performed. Otherwise, reject the connection between those components.

Further, in FIG. 12, even if the constraints of the output component itself are loose, if the constraints of the data flowing therethrough are strict, connection is allowed.

As described above, by inserting the procedure described with reference to FIGS. 11 and 12 into a conventional tool for connection between components, a bug caused by inconsistency of data exchanged between components can be removed at the stage of OpS design. .

[0070]

Embodiment 2 The present invention described in claims 4 and 5
This is related to the setting of devices (network elements / NEs) when individual network devices are connected.
Here, a device (NE) to be managed by the network
Is controlled by http, and its html (or XM
L, etc.) The management system treats the output as management protocol data, scans it, modifies necessary parts according to the invention described in claim 5, and sends the result to the user of the management system as html. The invention according to claim 4 for maximizing the operation method provided by each management device will be described.

FIG. 14 shows an example of an embodiment using the method of the present invention.
32 is a network element (NE); 33 is each means described in claim 4, that is, a means for trapping an HTML file output by an http server of a device to be managed in the network management system by a system for managing the device; The trapped HTML file is
HTML processing means for partially modifying or adding an initial setting value from the beginning according to the current connection / setting state of the device, and the processed HTML to a browser for an administrator of the network system. A manager constituting transfer means for transferring, and 34 is a G operated by a user.
UI, for example, a PC with a WWW browser.

The network 31 is configured by connecting a plurality of network elements 32, and a network control system (OpS) manages these devices. Here, each network element 32 to be managed is
Is provided with an http server. These network elements 32 can be collectively managed by a general network management protocol such as SNMP, and for each device, a GUI 34 is utilized by the HTTP protocol to allow the devices to operate. The best form, that is, the GU provided by these devices
It is also possible to manage devices in a way that respects the I-interface.

Here, it is assumed that a user of the network management system wants to set individual network devices using the http protocol. Then, the user can
The p request is sent to each device via the manager 33. Then, in this case, the manager 33
While respecting the content of http sent by each managed device to the maximum, the manager 33 sets the setting items relating to the entire network system from the beginning in the form that the user sets up html or the It is desirable to send similar data. Therefore, once sent html data is processed in the system as management data, and a necessary part of the data is subjected to processing such as data setting by a predetermined method as appropriate. The resulting data is converted back to html data and sent to the user.

In this way, the user is presented with the setting items in a form set from the beginning for the portion relating to the entire network system from among a large number of setting items that will be transmitted from each device. The setting of the network device is simplified, the load of the setting is reduced, and it is possible to prevent the entire network from being damaged due to the setting error. This is the invention described in claim 4.

Next, how the method described in claim 5 is applied as means for realizing this in a component-oriented framework will be described.

According to a fifth aspect of the present invention, as shown in FIG.
When the notification of an event is transmitted by a huge data object 41, a component 43 that performs a process of scanning the notification in units of a plurality of finer data objects 42.
It is.

An object 41 is transmitted from the left side of FIG. 15 to the next component as an event. In response to this event notification, the component 43 that scans the data into a plurality of finer parts outputs the input huge data object 41 as a finer data object 42. By subdividing, the next component can be processed in units of small data objects. The processing result can be reflected back if necessary.

[0078] It should be noted here that it has been assumed that the conventional communication protocol data all fits in one PDU. However, HTML structured type data and CMI
P Linked Reply, ASN. 1 set
It is difficult to apply such a conventional method as it is when it is desired to operate individual elements by components, such as of. Therefore, it is important to consider and prepare a component that can process data object data even in a fragmented form in some way, such as adoption of a component that divides the data object into a plurality of data objects described with reference to FIG.

Here, how the invention of claim 4 is realized by the invention of claim 5 will be specifically described with reference to FIG. First, as a response to the HTTP protocol request from the NE 51, the HTML data is returned to the manager 52 by the HTTP protocol response. At this time, the HT in the manager 52
The TP receiving component 53 converts the HTML data into structural data on a tree that is exchanged between components, like data of other communication protocols.
Send it to the destination component. Here, it is assumed that the HTTP receiving component 53 passes output data to the data scanning component 54 as an input.

The data scanning component 54 individually scans (decomposes) the passed tree-structured data,
As individual data, for example, HTML is checked and passed to a component for setting necessary setting items as appropriate.
The FORM setting processing component 55 sequentially checks each piece of input data that is sent in pieces one after another,
If it is FORM data of the type specified by the customizer, it is changed as necessary. In this way, processing can be performed on each part of the large structured data.

The data processed in this way is sent by the data scanning component 54 to the HTTP (transmission) component 56 for the user who operates the manager 52. The HTTP component 56 converts the received structured type data into HTML again, and executes the GUI 5 of the user.
Send to 7.

By doing so, the user can use the NE 51
It is possible to utilize the GUI provided by the NE 51 almost as it is, except for the data sent from the NE 52, except for the part modified by the manager 52 as necessary.

[0083]

Embodiment 3 Op by the conventional component technology
In the S construction method, at the time of design, such as a server / client or a manager / agent, at the time of operation:
Even in the connection relation of the components that becomes n dynamic relations, 1: 1 static connection was made. However,
In practice, the connection between these managers and agents is not always a 1: 1 static relationship as originally designed, but may be a 1: n relationship in some cases. This includes
This includes the case where an agent is added (a newly installed agent is connected to and added to the same manager) or a part of the agent is removed (an active agent is separated from the manager). A similar dynamic connection / disconnection relationship can occur between a server and a client.

A component-oriented feature is that once a connection is made with a tool, the connection can be executed or simulated immediately on the spot. However, such a 1: n relationship was difficult to simulate on a tool. As a solution, a component that is a 1: n demarcation point generates this n-side instance according to the customizer's setting or sends the instance appropriately to the generation destination. Simulate the relationship.

This mechanism is the same as that of Op.
In the S design method, as shown in FIG. 17, a component is separated into two components that extend between a manager and an agent. On the OpS design tool 60, components 61 to 65 are connected. The component 61 is a client from the beginning of the OpS design, and the NE is 65. The components 62 and 64 are components indicating communication on the network.

When the actual OpS is operated, components (62, 6) are added to the manager (or server) 66.
3, 64, and its components 62,
64 is separated (62 ', 64') and transferred to the client and the agent.

Thus, even if the client side or the agent side changes or a plurality of clients or agents connect to the server or the manager, it is possible to cope with the situation. FIG. 18 shows the connection relationship after NEs are dynamically registered.

In the present invention, the components 62 and 64 increase the instances of the components 61 and 65 virtually at the time of design, and the components 62 and 64 increase the 1: n relationship on the network during operation. Indicates that this situation can be properly simulated by appropriately distributing the data from the component 63 and vice versa.

In a multimedia network handling various main signals, various management protocols are used. The present invention proposes that the communication network management system be made component-oriented, and shows an access interface for handling different types of protocol data by the same component. Also, using the concept of Collection and constraint information, an interface independent of the protocol type is defined, and ASN. It has been shown that general-purpose components can maintain data processing while maintaining their structures for data of different protocols such as 1 and XML. The interface of the present invention is applicable to both of which have completely different APIs specific to the protocol, and both operations can be performed by the same data editing component.

In the embodiment, the interface for exchanging protocol data between components, which is necessary when the communication network management system according to the present invention is manufactured in a component-oriented manner, has been described. Collect
Using the concept of Ion and constraint information, an interface independent of the protocol type is defined, and ASN. 1 and XM
L is actually implemented using Java. Although the APIs specific to both protocols are quite different, the interface of the present invention is applicable to both. A GUI component that edits data through the interface of the present invention can be applied to both, and can operate data without depending on a protocol.

[0091]

As described above, according to the first aspect of the present invention, it is possible to simultaneously access various data structures with a unified data access interface by using a tree-type editor. As a result, the componentware (CW) required for the user to design the OpS
Can be easily set, and data exchanged between CWs can be visually understood.

According to the second and third aspects of the present invention, in order to appropriately guide the connection between the components, a constraint is introduced for data that can be exchanged between the components. The decision of connection (guidance) was made. By doing so, erroneous connections can be eliminated in OpS design, and design errors can be reduced.

According to the invention described in claim 4, when a large number of network devices are connected, one of them is connected.
For a GUI that changes the settings of one device, it is possible to utilize a setting GUI / http server mounted on each device and use a manager to adjust the entire network accompanying the change.

According to the fifth aspect of the present invention, when a component wants to perform individual processing on the contents of the structure of one data object, the component scans and decomposes the data object. And the use thereof, for example, according to claims 2 and 3
Of the invention in a component-oriented framework.

Furthermore, according to the invention of claim 6, with the component at the 1: n demarcation point as a boundary, the instance of the component block on the n-side from the demarcation point is used to simulate the protocol of the demarcation point at the time of simulation. Create and initialize many components to represent. In addition, the demarcation point component appropriately performs a data flow distribution process. This makes it possible to appropriately simulate the 1: n relationship of the network even on the tool.

[Brief description of the drawings]

FIG. 1 is a typical configuration diagram of a network control system.

FIG. 2 is a diagram showing a design flow of a conventional component-type network control system.

FIG. 3 is a diagram showing how data is set in OpS design.

FIG. 4 is a diagram showing an operation state of a network element;

FIG. 5 is a diagram showing an outline of an access interface of the present invention.

FIG. 6: Collec used as a complex data structure
Diagram showing the types of Tions

FIG. 7 is a diagram showing correspondence between a DOM and an access interface;

FIG. 8 is a diagram showing how data is set according to the present invention.

FIG. 9: ASN. Diagram showing an example of one data editing screen

FIG. 10 is a view showing an example of an XML data editing screen;

FIG. 11 is a processing flow of connection between components according to claim 2;

FIG. 12 is a processing flow of connection between components according to claim 3;

FIG. 13 is a processing flow of a conventional connection between components.

FIG. 14 is a diagram showing an example of an embodiment according to claims 4 and 5;

FIG. 15 is a diagram showing a state of processing a huge data object.

FIG. 16 is a diagram showing an example of a process according to claim 5;

FIG. 17 is a diagram showing dynamic registration of an NE according to an example of an embodiment according to claim 6;

FIG. 18 is a diagram showing a connection relationship after registration according to an example of the embodiment corresponding to claim 6;

[Explanation of symbols]

2, 14, 15: component, 11: unique data,
12: access interface, 13: event object, 21: common interface.

 ──────────────────────────────────────────────────続 き Continued on the front page F term (reference) 5B076 DD05 5K034 AA06 AA18 FF01 FF11 FF15 FF18 HH14 HH17 HH18 HH26 JJ23 MM39

Claims (6)

[Claims] 1. A network management system, comprising preparing and selecting componentware as a functional element for controlling a communication device of a network, and customizing the componentware according to the purpose of using the function of the selected componentware. In a network control system design method that performs design by connecting a number of componentwares configured as described above, a unified data access interface is used between the connected componentwares, although the data of various data structures is different. Unifying the display and editing methods for data handled by componentware, data customized by componentware itself, or data monitored for exchange between componentware Design method of a network control system according to claim. 2. When performing an operation of connecting two componentwares, if the output constraint of the componentware on the data output side is the same as or stricter than the constraint on the input side, connection is allowed. 2. The method for designing a network control system according to claim 1, wherein componentware is guided, and otherwise, connection between the componentware is rejected. 3. When connecting between two componentwares at the time of design, data is experimentally exchanged between the componentwares, and constraint information is obtained for data output from the componentware on the data output side. 2. The network control according to claim 1, wherein a tool is provided for making the output constraint information of the componentware itself on the data output side more strict according to the constraint condition of the acquired data, and connecting to the target componentware. How to design the system. 4. A means for trapping an HTML file output from an http server of a device to be managed in a network management system by a system for managing the device, and connecting the trapped HTML file to a current connection / setting of the device. HTML processing means for performing processing such as partially modifying or adding initial setting values from the beginning according to the state, and transfer means for transferring the processed HTML to a browser for an administrator of the network system. The method for designing a network control system according to claim 1, wherein: 5. At the time of design and operation, data to be processed as a single data object is appropriately divided and scanned at the time of design and operation, and individual data is transferred to appropriate componentware. 2. The processing according to claim 1,
5. The method for designing a network control system according to any one of claims 4 to 4. 6. A demarcation point componentware, which is a demarcation point between a componentware group having only one instance during operation and a componentware group having a plurality of components, among componentware groups connected at the time of design. When simulating the time, the demarcation point componentware itself creates and initializes many instances of the demarcation point component block at the boundary of this demarcation point componentware. The method for designing a network control system according to any one of claims 1 to 5, wherein a tool for sorting a data flow and simulating a 1: n relationship in a network is provided.