KR20120068279A - The automated test bed system using virtual network - Google Patents

Abstract

PURPOSE: An automatic test bed establishing system using virtual network is provided to minimize times required for establishing test beds and to enable reusing of the test beds. CONSTITUTION: An web interface(110) receives a request for establishing a test bed. The web interface is interlinked with various test beds. A resource allocating block(120) allocates resources extracted from the request for establishing the test bed. An automating block(130) automatically establishes an actual test bed. A test bed database(140) manages virtual image and information of the test bed and reuses the test bed.

Description

The automated test bed system using virtual network resource

The present invention relates to a network system.

The present invention is to design a system structure for the test bed construction using the resources provided in the virtual network environment, more specifically to the user to provide a variety of virtual networks separated by introducing the concept of virtualization in the existing network and end users Is to build a testbed through an application between

The structure of the Internet has been further advanced by the development of various terminals such as smart phones and the development of network devices that provide high bandwidth at the transport layer. However, it was not possible to provide an optimized network to satisfy various demands of users by upgrading the network alone. For this reason, a variety of Qos technologies have been developed for a long time. And the concept of virtualization has brought a new paradigm to these existing networks. The separation and operation of the existing network through virtualization enabled the operation of various separate services on top of the existing Internet structure, and many researches are being conducted under the name of the future Internet to provide such a separate network. have. In particular, empirical experiments on various services are becoming important through the construction of test beds using these virtual network resources, and many research teams are working on KOREN in Korea. But now that network virtualization has matured and the number of virtualized network resources has grown, there is a need to manage these resources collectively and make these testbeds easier for users.

Therefore, this article aims to maximize the utilization of virtual network resources by maximizing the utilization of virtual network resources by minimizing the inefficiency and optimizing the utilization of virtual network resources. We present the structure of system and resource allocation method for test bed construction.

Therefore, the technical problem to be achieved by the present invention proposes a virtual network resource allocation technique for building the automated testbed based on the virtual network as described above, and utilizes this to build a highly intelligent virtual network testbed.

In order to achieve the above object, the present invention provides a web service interface for accepting a testbed request; Optimized resource allocation includes a resource allocation block for; A server for managing a virtual machine image for reuse of a test bed; After all the test bed resource allocation is made, it is characterized by having an automation block for actually setting this content.

As described above, the automatic test bed construction system using the virtual network resources proposed in the present invention maximizes the utilization of the overall network resources through the virtualized network resources using the virtualization technology and the test bed configured using the same, The build of the testbed is automated, making it easier for users to build the testbed. Through this automated test bed construction, the user does not need to rely on the resource state of the virtualized network at all, and the system provides a test bed optimized for the user's needs. Through the database and open testbed of the testbed, it is possible to share information among people who need the same testbed, and to facilitate the reuse of the testbed, so that the user is wasted for building the same testbed. Minimize your time.

1 is a screen showing the structure of an automatic test bed building system utilizing a virtual network resource according to the present invention.
2 is a flowchart of actually constructing a test bed when a test bed request according to the present invention arrives.

In order to achieve the above object, the method of the present invention includes a first step of receiving a request for setting a test bed; Extracting optimized nodes and paths in consideration of the request of the received test bed; A third step of determining whether there is a resource that satisfies the request of the test bed; Determining whether it is possible to reuse an existing test bed when the test bed exists; And a fifth step of setting the test bed through the automation system.

Hereinafter, with reference to the accompanying drawings will be described in detail the specific flow of the present invention.

1 illustrates a structure of an automatic test bed construction system utilizing a virtual network resource according to the present invention.

As shown in FIG. 1, the automatic test bed construction system utilizing the virtual network resource according to the present invention accepts a request for test bed construction and includes a web interface 110 for interworking with various test beds and corresponding test bed requests. A resource allocation block 120 for allocating the extracted resources, an automation block 130 for automatically configuring the actual testbed after the resource is allocated, and a virtual image of the testbed thus formed, and for managing and reusing information. The test bed database 140 is provided. As described above, the system of the present invention separates the virtual network resources required in the test bed through the modified Constraint Shortest Path First (CSPF), and the Dijkstra algorithm considering the weight of each virtual network resource in the graph left through this separation. The path is obtained through. After the final graph of the virtual network resources is drawn, it is determined whether the existing testbed can be reused according to the service type of the testbed, and when the result is determined, the virtual machine image of the existing testbed is copied. The nodes of the testbed are configured in a distributed manner, and after the nodes are configured, a path is established through a signaling protocol. If the same test bed does not exist, the standardized image is copied and distributed to each node. Since only a minimal system is configured for such a standard image, the user needs to install the necessary items according to the type of service. The automated testbed building system utilizing the virtual network resources enables users to build the testbed more conveniently and quickly than the existing testbed building system, and reused through the database of testbeds. , Can reduce a lot of waste. At the same time, network providers can provide users with a much larger number of testbeds in a unified network system through optimized distribution of network resources.

FIG. 2 is a flowchart illustrating an optimal test bed construction and procedure under limited virtual network resources when a request for test bed construction is received in an automatic test bed construction system utilizing virtual network resources according to the present invention.

As shown in FIG. 2, when a test bed request is received through the web interface of the automatic test bed building system utilizing the virtual network resources of the present invention (S100), the system determines the virtual network resource requirements of the test bed. Will be extracted. In this system, the available resources are limited to three resources: storage size, CPU usage, and bandwidth. The received requirements are removed in the order of storage size (S101), required CPU usage (S102), required bandwidth (S103), nodes and links whose current resource state does not satisfy the requirements in the overall network graph. . The weight of the link is determined according to the degree of resource state of each node and link in the derived graph (S104).

A node pair having a minimum weight value is determined between all node pairs through a dijkstra algorithm through a graph having a weight value determined by the above (S105), and links and nodes having the least hop count among the pairs of these nodes are determined (S106). . Then, it is determined whether there is a test bed of links and nodes satisfying the requirements (S107).

As a result of the determination, if there is no test bed that satisfies the requirements, the test bed construction fails (S108), and if present, the same test bed search through the automation system and the test bed setting process through the automation system Will be skipped.

As a result of the determination, if there is a test bed that satisfies the requirement, it is searched whether a test bed of the same service type exists in advance in order to reuse the test bed (S109).

As a result of the search, if there is the same testbed, the image of the same virtual machine is copied from the testbed database and distributed to the nodes of the corresponding testbed (S110).

As a result of the search, if the same testbed does not exist, the standardized virtual machine image is copied from the testbed database and distributed to the nodes of the testbed. However, there is a hassle that the user has to newly set many elements to fit the service (S111).

As described above, after the virtual machine image and the path are determined, the actual test bed is set through the automation block (S112).

As illustrated in FIG. 1, a web interface 110 for accepting a request for testbed construction and interworking with various testbeds, a resource allocation block 120 for allocating resources extracted from a corresponding testbed request, and resource allocation After the automatic block 130 is automatically configured to configure the actual testbed and a virtual image of the testbed thus formed, and a testbed database 140 for managing and reusing information.

110: Web interface for accepting testbed build requests and interworking with various testbeds.
120: resource allocation block for allocating resources extracted from the corresponding testbed request
130: automation block for automatically configuring the actual testbed
140: Testbed virtual image and testbed database for managing and reusing information

Claims (10)

A web interface for accepting test bed build requests and interworking with various test beds;
A resource allocation block for allocating resources extracted from the test bed request;
Automation blocks for automatically configuring the actual test bed;
A virtual image of the testbed and a testbed database for managing and reusing information;
Automatic test bed building system utilizing the virtual network resources, characterized in that it comprises a. The method of claim 1,
The web interface is the entry point for the user to access the automated testbed building system utilizing the virtual network resources. The web interface can communicate with various user interfaces, and the API call allows the user to communicate with the system. . 3. The web interface of claim 2 not only accepts a request from a user, but also provides communication with a variety of different testbed systems. This allows connections between different tests. The method of claim 1,
The resource allocation block is a block that satisfies various requirements of the user and performs calculations to maximize resource utilization of the overall network. This creates a logical graph for building the testbed. The method of claim 4, wherein
The resource allocation block uses existing CSPF algorithms to extract nodes and links with free resources based on various requirements such as storage size, CPU usage, and bandwidth. The method of claim 4, wherein
The resource allocation block extracts a weight value between each node and link through a graph of nodes and links extracted from claim 5. At this time, the weight value is determined by the average of the ratio of surplus storage size, CPU usage, and bandwidth. And through this, the graph of the total weight value is extracted. The method of claim 4, wherein
In the resource allocation block, the final path between the nodes is determined through the Dijkstra algorithm according to the pair of nodes in the graph extracted from claim 6, and among the pair of nodes, the node having the smallest sum of the hop number and the weight of the path The path of the pair of fields is optional. The method of claim 1,
The automation block is a block for activation of nodes and connections between them after the nodes and paths of the actual testbed are determined. Since these nodes are virtual machines based on virtualization, they can be controlled through API calls of virtualization software, and in the case of links, links are formed through various network signaling protocols. The method of claim 8,
Although configured through an automation block, the link between the configured link and the node is provided for the accuracy of the test bed configuration through an automated verification process. The method of claim 1,
The testbed database is a database in which virtual images of testbeds that have been previously performed for reuse of testbeds are stored and the testbeds are managed.In this case, testbeds of the same service type can be copied and distributed through virtual images. Testbed copy is possible.

Images