KR101674543B1 - System and Method for Improving content Layer in protocol - Google Patents

Abstract

The present invention relates to an apparatus and a method for improving efficiency a content layer of the network configuration protocol (NETCONF) based on dependency of device configuration information with enhanced efficiency. The apparatus comprises: a candidate document checking module checking and sorting and a priority ranking of an XML document including an operation delivered from an NETCONF manager when the XML document comes in a queue for storing XML candidate documents; a VTD-XML module expressing the XML document sorted by the candidate document checking module in a binary array and storing position and content of all elements in the XML document; a data extracting unit extracting required data through an Xpath query with the content of the array by the VTD-XML module; an operation layer operating unit performing an operation of an operation layer of the extracted data; and a device configuration data storage and a data module modeling device configuration data based on dependency of the device configuration data and storing the same while periodically communicating with each other. The apparatus further comprises an NETCONF agent.

Description

[0001] The present invention relates to a device and a method for improving a content layer of a NETCONF protocol,

The present invention relates to a content layer of a NETCONF protocol, and more particularly, to an apparatus and a method for improving a content layer of a NETCONF protocol that increases efficiency based on dependency of equipment configuration information.

Modern networks, which are composed of various heterogeneous equipment, are complicated in configuration and have many types of equipment, so management tasks require a lot of time and manpower.

The NETCONF protocol is a protocol that can be centrally managed. NETCONF, which is the first standard published in 2006, has a four-layer structure.

Among these four layers, the Content layer is a layer related to modeling and creating equipment configuration data and storing it in the device's memory.

However, there is no standard data modeling technique for this content layer. However, it is common to model equipment configuration data in XML using a data modeling language called YANG as the de facto standard.

However, there are the following problems with the method of modeling the equipment configuration data with such XML.

Although the method of processing an XML document can utilize both the subtree filtering and the XPath function in the standard NETCONF, the modeling method for generating the XML document is somewhat inefficient, so that when generating, retrieving, retrieving, and deleting data, The command is executed in the form of locking the entire storage.

That is, when one NETCONF manager accesses one NETCONF agent, another NETCONF manager often receives a failure message if it waits to execute its own request or when it sends a request.

This situation can be a serious problem if the network size increases because the larger the network size, the larger the number of managers for management tasks and the higher the probability of accessing the same equipment.

Therefore, the data modeling method of the Content layer used in the standard NETCONF is very inefficient and the data modeling method is inefficient. As a result, the operation level of the operation layer becomes inefficient.

The network device's configuration data is the default behavior to lock the entire data store to prevent multiple managers from accessing the same data.

However, even on the same interface of the equipment, if you want to logically add only subnets with different port numbers, it is possible for different managers to add their own subnet, but the standard NETCONF will lock the entire data store So that one manager can not access another manager while accessing the memory.

Therefore, it is required to develop a data modeling technique of a new content layer to improve the problem of the data modeling method of the content layer.

Korean Patent Publication No. 10-2006-0087947 Korean Patent Publication No. 10-2007-0057603

The present invention solves the problem of data modeling of a content layer in the NETCONF protocol of the related art, and provides an apparatus and method for improving the content layer of the NETCONF protocol, which improves efficiency based on the dependency of the equipment setting information It has its purpose.

In the present invention, an XML document having equipment configuration information is divided into a plurality of subtrees based on the dependency of the equipment setting information, and the data stored in the memory is divided to allow partial access to the equipment setting data It is an object of the present invention to provide an apparatus and method for improving the content layer of the NETCONF protocol.

The present invention relates to an apparatus and method for improving the content layer of the NETCONF protocol, which allows a plurality of managers to perform setting change operations of the same equipment by making partial access to equipment setting data possible, The purpose is to provide.

The present invention relates to a method of applying VTD-XML (Vitrual Token Descriptor-XML) to a standard NETCONF and designing and implementing a data module for location and data modeling in a NETCONF agent, thereby improving the productivity of NETCONF- And an apparatus and method for improving the content layer of the protocol.

An object of the present invention is to provide an apparatus and method for improving a content layer of a NETCONF protocol using a subtree-based storage method when generating equipment configuration data in XML at a content layer.

It is an object of the present invention to provide an apparatus and method for improving the content layer of the NETCONF protocol, which expresses dependency between equipment setup data and a relative path by XPath and expresses it as a table.

The present invention provides an apparatus and method for improving the content layer of the NETCONF protocol that processes the above technique at high speed and extracts appropriate data from the XML document at high speed by using an extension of VTD-XML, which is open source, It has its purpose.

The objects of the present invention are not limited to the above-mentioned objects, and other objects not mentioned can be clearly understood by those skilled in the art from the following description.

In order to achieve the above object, an apparatus for improving the content layer of the NETCONF protocol according to the present invention comprises: an XML document including an operation to be delivered from a NETCONF manager, The document check module is a VTD-XML module that represents the XML document classified by the candidate document check module as a binary array and stores the location and contents of all the elements of the document of the XML, the XPath query with the contents of the array by the VTD- An operation layer operation unit for performing operation of the operation layer of the extracted data, a device setting data storing unit for modeling and storing the device setting data based on the dependency of the device setting data while periodically communicating with each other, And a data module; and that a NETCONF agent is configured .

Here, when the XML document including the operation transmitted from the NETCONF manager enters the queue storing the XML candidate document, the candidate document check module checks the candidate document, and transmits the XML document transmitted based on the manager's class and the contents of the XML document And classifies them into an urgent XML document or a general XML document.

The data extracting unit extracts necessary data in a state in which required data is randomly accessed through an XPath query and can be circulated without a specific order in the process of extracting necessary data.

And the data module is based on the address of the XML element location in the array generated by the VTD-XML module when modeling the data.

The data module is configured to express a dependency between equipment data by using a relative path that can be represented by XPath, and to generate a table in which data having a dependency is composed of one subtree.

In addition, each subtree can be stored as a single storage unit. In the subtree-based XML storage, the entire XML tree can be stored entirely without division, and can be stored separately for each node.

The NETCONF agent is characterized by classifying the information in an XML document into a subtree of an XML element based on the relative path of the XPath when storing the information in the network device.

The NETCONF agent performs XPath data storage based on the subtree, performs data modeling that integrates data storage through XPath processing and dependency of device setup data, and performs query processing in the data store through XPath processing .

A method for improving the content layer of the NETCONF protocol according to the present invention for achieving another object is to parse an XML document through extended VTD-XML to store the XML document in a subtree based on the NETCONF agent, A step of labeling and assigning each element based on a position address of an element in a document, a step of representing node depth by representing levels of elements, Determining whether the node is a repetition or a selection element if the node is one of a repetition or a selection element, designating the node as a subtree root, assigning it as a path type for a subtree root, And storing the generated record in a table.

Here, the node is characterized by being classified into an XML tag name or an attribute.

In addition, in the step of assigning each element by labeling, each element is labeled, and the inclusion and subordination of the parent element and the subordinate element are expressed and assigned as a labeling range.

Then, in the step of generating a record of the subtree root node and storing it in the table, a record consisting of PathID, begin, end, and value is stored in the table.

If the node is not an iteration or selection element, it is designated as a general node, a path is designated to a designated node up to a designated node, and a PathID is given.

According to another aspect of the present invention, there is provided a method for improving a content layer of a NETCONF protocol, the method comprising the steps of: when a query is input, tokenize a subtree root and a tokenized relative path word Checking whether the token is the last word of the relative path if the matching token exists, checking whether the token is the root of the subtree if the matching is the last word of the relative path, A step of partially locking the memory of the range after checking the labeling range of the root element, performing an operation allocated within the locked range, sending an OK message to the manager, and terminating the query processing do.

Here, if there is no matching token, it is determined that the query is not in the memory or is an incorrect query.

And determining whether the token is the last word of the relative path is to determine at which address the memory lock is to be performed.

If it is not the root of the subtree, it is judged to be one of the child nodes of the subtree, extracts the necessary elements from the subtree, checks the labeling range of the element, And partially locking the memory of the labeling range.

In the process of performing an operation assigned within a locked range, an operation is performed using any one of <get>, <get-config>, <edit-config>, or a combination thereof.

If the matched word is not the last word of the relative path, it is checked whether the token is the root of the subtree. If the matched word is not the last word of the relative path, it is determined whether the token is the root of the subtree. And then moves to the next token after checking the range.

If it is not the root of the subtree, the process moves to the next token without confirming the lower nodes, and returns to the process of confirming whether it is the last word of the relative path and repeats the process until the last word comes out.

The data storage structure for the query is characterized in that a data block is created depending on the data dependency of the equipment and a partial lock operation can be performed when several administrators simultaneously access one equipment at the same time.

In addition, the data storage structure for queries is a structure in which the logical part and the physical part are completely separated from each other, and the physical part is abstracted from the viewpoint of the network manager, .

The apparatus and method for improving the content layer of the NETCONF protocol according to the present invention have the following effects.

First, the efficiency of the content layer of the standard NETCONF protocol can be improved.

Second, we can solve data modeling problem of Content layer in NETCONF protocol by processing data based on dependency of equipment setting information.

Third, based on the dependency of the equipment configuration information, an XML document having equipment configuration information is divided into blocks of several subtrees, and the data stored in the memory is divided to allow partial access to the equipment configuration data.

Fourth, partial access to equipment configuration data is enabled, allowing multiple managers to perform configuration changes on the same machine, and to increase the responsiveness of command execution.

Fifth, the method of applying VTD-XML (Vitrual Token Descriptor-XML) to standard NETCONF and the data module for location and data modeling in NETCONF agent are designed and implemented to increase the productivity of network management work using NETCONF.

Sixth, when the device configuration data is generated in XML, it is stored based on the subtree, and the dependency between the configuration data of the device is made as a relation between relative paths by XPath, and the table is expressed as a control for the network management task Data can be reduced.

Seventh, it is possible to extract the appropriate data from the XML document at high speed while processing the above technique at high speed, using the open-source VTD-XML extension.

1 is a diagram showing an example of an XML document processed in the present invention
FIG. 2 is a block diagram of a byte array of a VTD record.
3 is a block diagram of an apparatus for improving the content layer of the NETCONF protocol according to the present invention
FIG. 4 is a block diagram showing an example in which an actual XML element is stored in a VTD byte array
FIG. 5 is a diagram showing a tree of an example XML document and a sub tree structure of an XML document
FIG. 6 is a diagram showing a structure of an XML document tree labeled for the XML tree illustrated in FIG.
FIG. 7 is an XML path fragmentation schema (XPFS) tree structure diagram for FIG. 6
8 is a flow chart illustrating a process of storing data in a subtree based on an agent
9 is a diagram showing a structure of data storage proposed by the present invention for a NETCONF query
10 is a flowchart showing a process for a query according to the present invention
11 is a diagram showing a state of storage in memory after a query is processed
12 is a diagram showing a detailed structure of a subtree after a query process
FIG. 13 is a schematic diagram of a device screen for improving the content layer of the NETCONF protocol according to the present invention
FIG. 14 is a graph showing an example in which throughputs of queries are compared
FIG. 15 is a graph showing a comparison of processing times for queries

Hereinafter, a preferred embodiment of an apparatus and method for improving the content layer of the NETCONF protocol according to the present invention will be described in detail as follows.

The features and advantages of the apparatus and method for improving the content layer of the NETCONF protocol according to the present invention will be apparent from the following detailed description of each embodiment.

Fig. 1 is a configuration diagram showing an example of an XML document to be processed in the present invention, and Fig. 2 is a configuration diagram of a byte array of a VTD record.

In the present invention, an XML document having equipment configuration information is divided into a plurality of subtrees based on the dependency of the equipment setting information, and the data stored in the memory is divided to allow partial access to the equipment setting data will be.

There is a dependency between equipment configuration information. Dependency refers to data whose configuration information underneath it is treated as a block and can not be accessed while changing specific configuration information.

In the present invention, an XML document having equipment setting information based on the dependency is divided into a plurality of subtrees and the data stored in the memory is divided.

Also, by labeling the range around the root of the subtree and then storing the data in a table form, partial access to the equipment setting data is made possible, so that multiple managers can perform setting change of the same equipment And to increase the responsiveness of command execution.

Data modeling in the Content layer, the top layer of the NETCONF protocol, is the de facto standard for presenting XML documents through YANG.

Based on these standards, the present invention largely carries out data modeling of the content layer using dependency of equipment setting information by applying the following technical structure.

One is the subtype-based storage method used to generate device configuration data in XML at the Content layer. Then, the dependency between the set data of the equipment is made as the relation between the relative path by the XPath, and this is represented as a table and modeled.

The other is to use the extension of VTD-XML (Virtual Token Descriptor for eXtensible Markup Language), which is open source, to process the above technique at high speed and to extract appropriate data from XML document at high speed.

The configuration information of the equipment is finally stored in the memory, but the actual CPU occupation or time consuming is the task of parsing the XML document which is setting data and extracting or storing the necessary data.

FIG. 1 shows an example of receiving a response after performing a < get-config > operation with NETCONF.

Structurally, it can be seen that the <interface> element contains a name, an IP address, and a MAC address in a tree form. It is also possible to check that the <interface> element is repeated twice, though the actual value is different.

In the present invention, a method of expressing and modeling equipment configuration data dependency is derived by suggesting a data storage method based on subtree division and applying a relative path used in XPath.

That is, the contents of FIG. 1 are created and stored in the form shown in Table 1 using the above-mentioned method.

The data extraction using VTD-XML and XPath will be described as follows.

XML parsing based on the VTD-XML model is a 'non-extraction' XML technology called a 'virtual token descriptor'.

'Non-extraction' means that an XML document is originally read into memory in binary form as shown in FIG. 2, and the positions of all elements are analyzed as a byte array, and the information is recorded.

A traversal operation to find appropriate information in XML is performed using information in the byte array of FIG.

VTD-XML parses XML documents and generates VTD records (tokens) in 64-bit binary format for each event. With VTD-XML, you can take advantage of the flexibility of XPath expressions when extracting the necessary data, and XPath can compensate for the slower drawbacks than the subtree method.

Also, because VTD-XML requires less resources while showing better performance compared to the DOM (DOM requires 5 to 10 times the original document size, VTD-XML is 1.3 to 1.5 times the original document Of memory) You can handle the task of extracting data from a document with a size greater than 256GB.

In addition, VTD-XML provides random access capability and enables rapid analysis and traversal.

In one embodiment of the present invention, VTD-XML uses extended VTD-XML combined with a 64-bit JVM to support XPath-based XML processing.

Therefore, XPath queries can be used, and XPath queries are useful when many NETCONF managers frequently create a large number of device configuration commands.

Using the subtree-based XML document storage and data modeling using the dependency of the equipment setting data, which is applied in the present invention, it is possible to bundle related data items using the XPath query and avoid generation of wasteful objects.

As a result, VTD-XML is suitable for generating the table as shown in Table 1 at a high speed based on the tree structure.

VTD-XML consists of three components and has a Java API for VTD-XML that appears at the top level. The three components are VTD generator, VTD navigator, and Autopilot.

The NETCONF agent configuration of the apparatus for improving the content layer of the NETCONF protocol according to the present invention is as follows.

3 is a block diagram of an apparatus for improving the content layer of the NETCONF protocol according to the present invention.

The configuration of the NETCONF agent 100 of the apparatus for improving the content layer of the NETCONF protocol according to the present invention is such that an XML document including an operation transmitted from the NETCONF manager 200 is first placed in a queue 30 storing an XML candidate document A candidate document check module (not shown) for checking the priority candidate of the XML document transmitted from the manager based on the rank of the manager and the contents of the XML document and classifying them into the emergency XML document 32 or the general XML document 33 31), a VTD-XML module (34) which is expressed in a binary array regardless of the type of the XML document and stores the positions and contents of all the elements of the XML document, and a VTD-XML module A data extracting unit 35 for randomly accessing and circulating without any specific order, for extracting necessary data, an operation layer operation unit 36 for performing operations in the operation layer, , And it performs the operation of the operation when the layer comprises a periodic communication with equipment that models the device setting data, based on the dependency of the machine setting data with each other and store setting data storage 37 and the data module (Data Module) (38).

In the present invention, data extraction using VTD-XML and XPath to be applied and data modeling of the content layer are all included in the NETCONF agent 100 side.

The XML document including the operation transmitted from the NETCONF manager 200 enters the queue 30 storing the XML candidate document and is transmitted to the candidate document check module 31 which checks the candidate document. The priority of the XML document transmitted based on the class of the manager and the contents of the XML document is checked and classified into the emergency XML document 32 or the general XML document 33. [

Regardless of the type of the XML document, whether urgent or general, the VTD-XML module 34 expresses the elements in a binary array, and all the elements of the XML document are stored in the position and contents. Then, the necessary data is randomly accessed through the XPath query with the contents of the array, and it can be circulated without any specific order, and the data extracting unit 35 extracts necessary data.

The operation of the overlay layer is performed in the operation layer operating unit 36. The equipment setting data store 37 and the data module 38 periodically communicate with each other to model the equipment setting data based on the dependency of the equipment setting data And store it.

Here, also when modeling the data, it is based on the address of the position of the XML element in the array generated by the VTD-XML module 34.

Especially, it creates a table that expresses the dependency between equipment data by using a relative path expressible by XPath and composes data having dependency into one subtree.

FIG. 4 is a diagram illustrating an example of storing actual XML elements in a VTD byte array, and FIG. 5 is a tree diagram of an exemplary XML document and a subtree of an XML document.

FIG. 6 is a diagram of an XML document tree labeled with respect to the XML tree illustrated in FIG. 5, and FIG. 7 is a diagram illustrating an XML path fragmentation schema (XPFS) tree diagram of FIG.

[Subtree-based XPath data storage method] is as follows.

The first thing to do is to enter an XML document and store it in a subtree form. This is VTD-XML, which is an open-source, memory-efficient parsing technique. Extension of the XML document When the VTD binary array is generated using VTD-XML, it is as shown in FIG.

Next, XML node labeling, which is the latest technique of storing XML data, is performed using the VTD binary array as shown in FIG.

An example of node labeling is shown in (2:25, 2) in FIG. This represents the first interface in Figure 5, where the first two numbers separated by a colon are the coverage of the node indicating the (start: end) and the following number is the level (depth) of the node in the tree do.

Labels enable structured joins as XPath processing proceeds.

n ₁ , n ₂ If the node is in the tree, the ancestor-descendant or parent-child relationship, which is the structural relationship of the nodes, is determined by the label.

That is, if the range of n _{1 includes} a range of n ₂ , then n ₁ is an ancestor of n ₂ .

Also, if there is a level difference between an ancestor-descendant pair relationship, one node is the parent node of another node. The data such as that shown in FIG. 5 is stored in this relationship, and the stored items are composed of the respective nodes. Nodes are tag names of elements or attribute names with a prefix of @.

And the subtree of the XML tree can be regarded as a storage unit of the XML repository.

For example, in the XML tree shown in Fig. 5, a line-bound portion shows that the XML tree is divided into subtrees. Each subtree can be stored as a single storage unit. In the subtree-based XML storage, the entire XML tree can be stored entirely without division, and can be stored completely separately for each node.

We need an XML path schema and an XML fragmentation schema to store XML data in a subtree and process the data extraction in the subtree for XPath queries without parsing.

The path schema can be represented as a tree as shown in FIG. 7 and can be generated from the XML tree of FIG. The notation is simple. The prefix # is added next to each node, which indicates the node's path ID and is assigned as the path type from root to N.

For example, in FIG. 6, the path type / data / interface / ipv4-address is assigned Path ID # 5. The XPFS tree shown in FIG. 7 is created through such a process. That is, Figure 7 shows the path and partition of the XML and contains information about how the original XML tree was divided into subtrees.

There are two kinds of nodes in an XPFS tree, one represented by a single circle and the other a double circle. The double circle represents the root of the subtree to be stored, and the single circle represents the descendant node belonging to the subtree t. Table 1 shows the manner in which the subtree information in the XPFS and the labeling information in FIG. 6 are stored.

In Table 1, seven subtrees are stored. The no column is a simple reference number, so it is not related to the storage contents.

The subtrees (1), (2), and (5) include the symbol "$"

In the present invention, it is assumed that an XML document conforms to a specific DTD.

Therefore, in dividing the XML tree into subtrees, the basic XML segmentation rules are '+, *,?' Is an element that can be repeated or selected that can be the root of another subtree.

For example, interface1 and interface2 in FIG. 6 are repeated elements, and both elements have the same parent node, so the path type is the same as / data / inteface. Under the basic XML partitioning rule, both elements are the root of the subtree.

For the selected element, if the XML node with the tag name 'name' in the path type / data / interface / name is optional, then each instance name ₁ and name _{3 of them} is the root of the subtree.

The advantage of storing in this way as shown in Table 1 is that it reduces data storage space and at the same time allows partial processing of XPath queries without parsing.

That is, accessing the subtree is sufficient to access the root of the subtree, and parsing is required only when accessing the node inside the subtree.

In the case of storing by node, one node is allocated to each XML element and its memory is required. However, in the subtree-based storage method, since only a part of the pointer to the node of the subtree can be stored, the memory can be saved.

In the present invention, the dependency between data and the subtree-based XPath data storage method described above are applied when initially creating and storing equipment setting information.

FIG. 8 shows a process of storing data in the agent used to store the XML document described above in a subtree-based storage manner.

The XML document is parsed through extended VTD-XML (S801) and stored as a VTD byte array (VTD Binary Array or VTD Byte Array) (S802)

In the XML document stored in this array, the position of an element in the document is recognized as an address (S803)

Then, each element is labeled based on this address, and the inclusion and dependency of the upper and lower elements are expressed and assigned as the labeling range (S804).

And the depth of overlap of elements is expressed as level. Subsequently, the node is nodeized and represented in a tree form around the root element (S805)

The node is classified into an XML tag name or attribute (S806), and a basic XML segmentation schema is applied to it (S807) to check whether the node is a repetition or a selection element (S808)

If the node is one of the repetition or selection elements, it is designated as a subtree root (S809), and designated as a path type for the subtree root, and then a PathID is assigned (S810)

Then, a record consisting of PathID, begin, end, and value is generated and stored in the table (S811), and then the data storage process is terminated.

If the node is not an iteration or selection element, it is designated as a general node (S812), the path is designated as one type up to the designated node (S813), and a PathID is assigned (S814)

However, the process of storing the data is terminated without being stored in the table like the root node.

Hereinafter, data modeling that integrates data storage via XPath processing and dependency of device setting data will be described.

As described above, when the subtree storing method of XML data is combined with the dependency between the equipment setting information, the related data can be handled as one block when modeling the data of the setting information of the initial network equipment.

For example, when trying to perform an operation on an interface and a subnet on that interface in NETCONF, the XPath query in a specific command expressed as an XML document would be the same form as shown in Table 2 .

Assuming that the name of the specific interface and the name of the interface of the subnet are changed or searched for the expression, and the data is modeled using the XML storage structure proposed in the present invention, A logical data structure of the same type is created.

For the query in (1) of Table 2, the subtree in the rectangle as shown in (1) in Fig. 9 can be accessed based on the subtree root as one data block.

Likewise, the query in (2) of Table 2 can treat a subtree in the square (2) having the same subtree structure but different storage ranges as one data block.

If another query appears, that is, for queries that return all names of the interface of the instrument, data can be extracted using the dotted rectangle in ⓐ.

Such a data storage structure creates a data block depending on the data dependency of the apparatus, and even when several administrators access one apparatus at the same time, it can perform a partial lock operation by locking each square individually (lock operation).

As shown in FIG. 9, since the logical part and the physical part of the data structure are completely separated from each other and the physical part is abstracted from the viewpoint of the network manager, the configuration of the equipment configuration can be made by viewing only the subtree storage structure of the XML document .

[Query processing in data storage through XPath processing] is as follows.

The process of processing the above query is as follows.

The query terms are set to be the same as in Table 1 (1).

(1) / data / interface [@ xmlns = "urn: ietf: params: xml: ns:

10 shows a processing procedure for a query.

When the query is input as shown in (1), the query format is an XML relative path, so the tokenization is started (S1001)

The subtree root in the storage (memory) is collated with the word of the tokenized relative path (S1002)

Then, it is searched whether there is a match (S1003)

If there are no matching tokens, there are several matching tokens. If there is no matching token, it is a query that is not in memory or a wrong query. Therefore, a Fail message is sent to the Manager and the query processing is terminated (S1009)

If there is a matching token, it is checked whether the token is the last word of the relative path (S1004)

You need to know the end of the relative path before you can decide what address to perform the memory lock on.

If the matching is the last word of the relative path, it is checked whether it is the root of the subtree (S1005). If it is the root of the subtree, the range of the labeling of the root element is confirmed (S1008)

If it is not the root of the subtree, it is one of the child nodes of the subtree. Therefore, the required element is extracted from the subtree (S1006), the labeling range for the element is confirmed (S1007), and the memory corresponding to the labeling range of the element is partially locked. (S1010)

And performs operations such as <get>, <get-config>, <edit-config>, etc. that are allocated within the locked range (S1011)

When the operation is successfully completed, an OK message is sent to the manager and the query processing is terminated (S1012)

If the matched word is not the last word of the relative path, it is checked whether the token is the root of the subtree (S1013). If the matched word is not the last word of the relative path, (S1015) and moves to the next token (S1016)

If it is not the root of the subtree, it moves to the next token because it is not necessary to confirm the child nodes (S1016)

Then, it goes back to the process of confirming whether it is the last word of the relative path and repeats it until the last word comes out.

Figure 11 shows the storage format in memory after query (1) has been processed.

From Fig. 9, which is the result of the modeled data, it is shown that only the part of the square box indicated by (1) is processed in a locked state in response to the query.

① The box inside the box is maintained as another subtree independently. That is, (1) Even if the box is locked, access to (2) is allowed.

12 shows the detailed structure of a subtree rooted at # 2 which is a subtree in a box.

# 3, # 5, and # 6, which are not stored in the table, are all indicated. However, since the location information about these is stored in the pointer # 2, access to # 2, that is, access to the subtree root, It is possible to access them.

The characteristics of the data modeling method applied in the apparatus and method for improving the content layer of the NETCONF protocol according to the present invention will now be described.

To test the performance, configure the following environment.

Because the virtual switch was used, we used the Windows 7 / 64bit operating system that enabled the virtual switch to run smoothly and run the NETCONF agent proposed in this invention.

The schematic configuration of the agent computer consists of the Intel i7 processor and memory 4G, and the remaining elements do not significantly affect the operating environment. The manager computer is also configured with the same settings as the agent. The VTD-XML module uses open source, uses an extended VTD-XML module that supports XPath, and runs on a 64-bit JVM.

We also used the NETCONF GUI program that implements the algorithm developed using C language for data modeling based on equipment configuration data dependency from binary array. The number of network managers is limited to 10, and compared with the improved NETCONF using the subtree storage method and VTD-XML proposed in the present invention, NETCONF using only equipment dependency, and the standard NETCONF.

13 is a part of a screen for connecting and operating with a virtual switch using a GUI program.

In order to compare the throughput of query configuration, it is necessary to compare the actual throughput of each NETCONF manager with respect to the number of commands to modify or read the configuration information for the device.

In other words, it is to check how much efficiency the data modeling technique that constructs the setting information of the equipment based on the subtree shows in other modeling method and partial lock. The experiment measured the number of received OK message messages while increasing the number of commands from 100 to 1000, and suggested the types of commands as <get-config> and <edit-config>.

14, NETCONF improved by applying the technique proposed in the present invention shows a processing success rate of about 95% even when the number of instructions reaches 1000, and shows an average throughput rate of 97%.

It can be confirmed that the NETCONF having added the two techniques of the present invention has a throughput rate which is 4% and 10% higher than that of the remaining NETCONF, respectively.

The comparison of the system processing speed with respect to the query shown in FIG. 15 is an embodiment for comparing the response speed of the NETCONF agent according to the increase in the number of queries, and measures the time from when the manager sends the query, will be.

An important part of this time is the XPath processing time, which is the time that the XPath expression is processed for the configuration data of the device configured in the form of an XML document.

If the method proposed in the present invention is applied, it can be assumed that the configuration data of the equipment is stored in a state as shown in Table 1 in which an XML document is represented by a table-type address.

Therefore, in the embodiment, it is assumed that the data as shown in Table 1 has already been constructed, and only the time at which the response to the command arrives is measured.

When the number of queries is less than 100, the result of multiplying the average value of a single instruction processing by the number of queries is displayed.

That is, as shown in FIG. 10, the NETCONF improved in the present invention and the other two types of NETCONF consumes a little more time than the number of queries multiplied by the number of queries in a single query processing speed of 20 ms up to 200 commands.

In the case of NETCONF improved by the method proposed by the present invention, there is no large difference between the time of processing a single query and the time of processing a single query by about 24 ms even when the number of queries is 1000.

However, as the number of queries increases, the efficiency of the modeled data is increased by storing the device configuration information based on the subtree, and the performance of data extraction of VTD-XML is further enhanced. Specifically, If you check up to the root of the subtree, you can find the dependent data directly, and since the partial lock is also fast based on the subtree, there is almost no drop in performance.

When this method is used, partial simultaneous access is allowed when there are multiple NETCONF managers and there is one network device to change the setting. It is not available in standard NETCONF.

In addition, the configuration data can be blocked and stored based on the dependency of the device setting information, so that the <lock> operation in the operation layer of the NETCONF can be made partially lockable. This means that multiple managers can perform concurrent operations on the same equipment, thereby improving the productivity of network management tasks.

In addition, when performing operations such as creation, search, change, retrieval, and deletion of equipment configuration information, the access speed of data is increased and simultaneous operation can be performed, which can also be connected to increase productivity of network management task.

In addition, since it is unnecessary to transmit unnecessary information additionally, overhead due to network management can be avoided by reducing control data (overhead) for network management service.

The apparatus and method for improving the content layer of the NETCONF protocol according to the present invention can block an XML document having equipment setting information into a plurality of subtrees based on the dependency of equipment setting information and store the XML document in a memory The data is divided into partial access to the equipment setting data.

As described above, it will be understood that the present invention is implemented in a modified form without departing from the essential characteristics of the present invention.

It is therefore to be understood that the specified embodiments are to be considered in an illustrative rather than a restrictive sense and that the scope of the invention is indicated by the appended claims rather than by the foregoing description and that all such differences falling within the scope of equivalents thereof are intended to be embraced therein It should be interpreted.

31. Candidate document check module 32. Emergency XML document
33. Generic XML documentation 34. VTD-XML module
35. Data extraction unit 36. Operation layer operation unit
37. Equipment configuration data store 38. Data module

Claims (22)

When the XML document including the operation transmitted from the NETCONF manager enters the queue storing the XML candidate document, the candidate document check module for checking and classifying the priority order;
A VTD-XML module that expresses classified XML documents in a binary array in a candidate document check module and stores the positions and contents of all the elements of the XML document;
A data extracting unit for extracting necessary data through an XPath query with the contents of the array by the VTD-XML module;
An operation layer operation unit for performing operation of the operation layer of the extracted data;
A device for setting up a content layer of the NETCONF protocol, comprising a device setting data store and a data module for modeling and storing device setting data based on dependency of device setting data while periodically communicating with each other; . 2. The method of claim 1, wherein the candidate document check module is configured to, when an XML document including an operation transmitted from a NETCONF manager enters a queue storing an XML candidate document,
The candidate document is checked and the priority of the XML document transmitted based on the manager's class and the contents of the XML document is checked and classified into an urgent XML document or a general XML document to improve the content layer of the NETCONF protocol Device. The data extracting apparatus according to claim 1, wherein, in the process of extracting necessary data,
Wherein the necessary data is randomly accessed through an XPath query, and necessary data is extracted in a state in which the data can be circulated without a specific order. 2. The method of claim 1,
Wherein the data is based on an address of an XML element location in the array generated by the VTD-XML module when modeling the data. 2. The method of claim 1,
Wherein the dependency between equipment data is expressed using a relative path that can be represented by XPath, and a table in which data having a dependency is composed of one subtree is generated. 6. The method of claim 5, wherein each subtree can be stored as one storage unit. In the subtree-based XML storage, the entire XML tree can be stored as a whole without division, A device for improving the content layer of the NETCONF protocol. The method of claim 1, wherein the NETCONF agent comprises:
A device for improving the content layer of a NETCONF protocol, characterized in that when information of an XML document is stored in a network device, it is classified into a subtree of an XML element based on a relative path of XPath. The method of claim 1, wherein the NETCONF agent comprises:
A data modeling based on subtree-based XPath data storage, data storage through XPath processing and dependency of equipment setting data is performed, and query processing in a data store is performed through XPath processing. A device for improving the content layer of a protocol. In order to store the XML document in the NETCONF agent based on the subtree,
Parsing the XML document via extended VTD-XML and storing the parsed VTD byte array;
Labeling and assigning each element based on a location address of an element in the document;
A step of representing a superimposed depth of an element by a level and nodeized and a tree form centering on a root element;
Applying an XML partitioning scheme to determine whether the node is a repetition or a selection element;
If the node is one of an iteration or a selection element, designating it as a subtree root, designating it as a path type for a subtree root, and assigning a PathID;
Generating a record of a subtree root node and storing the generated record in a table. 10. The method according to claim 9, wherein the node is classified into an XML tag name or an attribute. 10. The method according to claim 9, wherein in the step of assigning each element by labeling,
And labeling each element, and expressing the inclusion and dependency of the parent element and the subordinate element in a labeling range and allocating the labeling range. 10. The method according to claim 9, wherein in the step of generating a record of the subtree root node and storing the record in the table,
A record consisting of PathID, begin, end, and value only for the subtree root node, and then storing the record in the table; and a method for improving the content layer of the NETCONF protocol. The method as claimed in claim 9, wherein if the node is not a repetition or a selection element, the node is designated as a general node, a path is assigned to a designated node, and a PathID is assigned. In order to perform query processing in the NETCONF agent,
If the query is input, tokenize and compare the subtree root with the word of the tokenized relative path;
If the matching token exists, checking whether the token is the last word of the relative path;
Confirming if it is the last word of the relative path if it is the root of the subtree;
Checking the labeling range of the root element if the root of the subtree is present, and partially locking the memory of the range;
Performing an operation assigned in a locked range, sending an OK message to the manager, and terminating the query processing. 15. The method as claimed in claim 14, wherein if there is no matching token, it is determined that the query is not in the memory or is an incorrect query. 15. The method of claim 14, wherein determining whether the token is the last word of the relative path is to determine at which address the memory lock is to be performed. 15. The method of claim 14, further comprising:
If it is not the root of the subtree, it is judged to be one of the child nodes of the subtree and necessary elements are extracted from the subtree,
Identifying a labeling range for the element and partially locking the memory of the labeling range of the element. 15. The method of claim 14, wherein in the process of performing the assigned operation within the locked range,
wherein the operation is performed using one of <get>, <get-config> and <edit-config> or a combination thereof. 15. The method of claim 14, wherein in determining whether the token is the last word of the relative path,
If the matched word is not the last word of the relative path,
Checks if the token is the root of the subtree, locks the memory of that root node if it is root,
And then moves to the next token after confirming the labeling range. 20. The method of claim 19, wherein if the root of the subtree is not the root of the subtree,
And returning to the process of confirming whether the last word of the relative path is again repeated until the last word is reached. 15. The system of claim 14, wherein the data storage structure for the query comprises:
Wherein a data block is created depending on data dependency of the device, and a plurality of administrators simultaneously access one device at the same time, the partial lock operation can be performed by individually locking the data block. 15. The system of claim 14, wherein the data storage structure for the query comprises:
The content layer of the NETCONF protocol is characterized in that the physical part is completely separated from the logical part and the physical part is abstracted from the viewpoint of the network manager, Lt; / RTI &gt;

Images