CN103984589A - Virtual machine-based lab practice environment establishment method for realizing computer experiment teaching - Google Patents

Abstract

The invention discloses a computer environment establishment method for realizing computer experiment teaching based on a virtual machine. This method is as follows: 1) Students complete authentication and login on the system web service home page, and then select the current course; wherein, the main image file of the course has been deployed on each physical node; 2) Check and manage the status of the virtual machine through the interactive interface; 3 ) Obtain the desktop of the virtual machine through the system Web service page; then use the virtual machine in the browser. Compared with the prior art, the present invention can help to shield the underlying physical hardware differences through virtualization technology, and provide diversified execution environments for upper-layer users, and the virtual machines are isolated from each other, so users can use private virtual machines as the computer platform , Freely install personalized software, user data is stored in the virtual machine image, which can ensure data security.

Description

A method for establishing computer environment based on virtual machine to realize computer experiment teaching

technical field

The invention relates to a method for establishing a computer environment for realizing computer experiment teaching based on a virtual machine, and belongs to the technical field of virtual machines.

Background technique

The computer experiment teaching center is an important place to implement experiment teaching. It should provide students with good computer services, facilitate teachers to participate in teaching practice, and facilitate administrators to participate in monitoring and management. With the development of the education model, the shortcomings of the traditional experimental teaching center have become increasingly prominent. First of all, when carrying out experimental teaching activities in traditional experimental teaching centers, we often face security problems: in the public experimental environment, users share storage, and private file data can be deleted and altered at will, resulting in loss; in order to ensure system security, students usually do not have permission to install software, which not only sacrifices students User experience also increases the management burden for administrators; hardware failures can also easily cause serious consequences. Secondly, the resource utilization rate of the traditional experimental teaching environment is not high: the hardware and software resources of the experimental teaching center are usually not fully utilized by users. Finally, the management activities of traditional experimental teaching centers are mainly participated by administrators, lacking the effective participation of teachers, and administrators lack a convenient and efficient experimental teaching center management mechanism.

Contents of the invention

Based on the above requirements, this paper proposes to use virtualization technology to solve the above problems. Virtualization technology can help shield the underlying physical hardware differences, provide a variety of execution environments for upper-level users, and isolate virtual machines from each other, so users can use private virtual machines as a platform to freely install personalized software and user data Stored in the virtual machine image, which can ensure data security; when a hardware failure occurs, migrate the virtual machine or transfer the image file, and the original operating system and application software can be quickly restored; at the same time, with the popularization of multi-core technology, a physical machine Running multiple virtual machines will not affect the user experience, so the use of virtualization technology can help improve resource utilization and alleviate the current situation of hardware shortage.

This paper proposes the concept of a personalized computer environment based on virtual machines to support computer experiment teaching, aiming to provide a complete set of teaching management software modules for computer experiment teaching, which can be convenient for computer room administrators, teachers and students Provide efficient experimental teaching services.

1. Design goals

1.1 Demand Design

First of all, there are three types of service objects provided by the personalized computer environment of computer experiment teaching: student users, teacher users, and computer room administrator users. Through the analysis of the needs of these three kinds of users, we put forward corresponding functional design concepts respectively.

For student users, it mainly provides students with free-to-operate private virtual machines, supports students to put forward more flexible demands on personalized operating systems, application software, computing resource configuration, etc., and provides convenient interactive interfaces for students to control and Use its personalized virtual machine (such as accessing its own virtual machine through a client or browser). Specifically, there are the following functional requirements, as shown in Figure 1.

It enables students to conveniently carry the operating system and application software used in the experimental teaching center, and restore the use of the virtual machine environment on other platforms. Support students to remotely access their personalized virtual machines outside the experimental teaching center, and provide a variety of access methods. Remote users share computing resources with internal users. Students can close, save, backup, and restore personalized virtual machines to achieve Courses related to the administration of personalized virtual machines.

For teacher users, provide a convenient management platform for course teaching experiments for teacher users. The virtualization experiment teaching center also needs to be oriented to courses, and provide a variety of computer environments for course experiments (for example, the introduction to computing course requires windows, Microsoft office, vc++6.0, etc., while the compilation practice course requires GCC, Java, etc.). The system provides a complete set of management interfaces to allow teachers and users to manage course-related information, and can easily create, deploy and manage virtual machines for course experiments. The virtualization technology realizes the mechanism that the underlying hardware resources are transparent to the upper-level virtual machine users, so a special experimental platform can be built in the virtualization experiment center to expand the teaching experiment mode. Take the two courses of Operating System Internship and Introduction to Computer Networks of the Computer Department of Peking University as an example: In order to protect the software and hardware from being damaged, the operating system internship did not allow students to spy on the real operating system (such as debugging the Linux kernel); Due to resource constraints, the experimental course of network theory cannot set up an independent private local area network for each student to assist network experiment exploration. Using virtualization technology, students can debug VM kernel code, build a virtual local area network, etc., and provide a unified mechanism to manage user information and personalize virtual machines. The functional requirements from teacher users are shown in Figure 2.

For administrator users, it aims to provide a comprehensive system platform for deployment, monitoring, management and maintenance. The system can generate relevant usage data of various users and an overview of computing resources for administrator users, and provide interactive functions for monitoring and management, such as managing students to participate in courses and using personalized virtual machines; managing teachers to deploy course experiment platforms to monitor the entire teaching Real-time status of virtual machines in the lab center. The functional requirements of the administrator user are shown in Figure 3.

1.2 Functional Design

In view of the deficiencies in the traditional experimental environment, we put forward the following design concepts and goals in this system:

●Personalization: Provide students with freely operable personalized virtual machines in the experimental teaching center. Virtualization technology can support students to more flexibly configure the experimental environment they need, including operating system, application software, computing resources, etc. (for example, students use Windows7 operating system, 4GB memory). Provide a convenient interactive interface for students to control and use their virtual machines (such as accessing their own virtual machines through clients and browsers).

●Curriculum-oriented: To meet the needs of diversified course teaching, it provides a variety of computer environments. Using virtualization technology, teachers can break through the limitations of the existing experimental environment and set up unique experimental courses. Provide a complete set of operation interface for teachers to manage course information, which is convenient for creation, deployment and management. It can be used to expand the experimental teaching mode: for example, to provide students with an operating system that can be freely modified when carrying out system security course experiments; to provide students with several virtual machines to form a local area network when carrying out network course experiments.

●Portable: Students use USB and other devices to carry their personalized experiment environment used in the computer experiment teaching center, and resume operation on other platforms.

●Remote access: Provides a variety of virtual machine access methods. Such as supporting students to remotely access their personalized virtual machines. Remote users can share computing resources with local users, and provide managers with a unified mechanism to manage all user information and personalized virtual machines.

●Monitoring management: In order to ensure the good operation of the experimental teaching center, an overview map of various computing resources should be generated for the administrator, and an interactive interface for monitoring management should be provided: such as managing students to participate in courses and using personalized virtual machines; managing teachers to deploy course experiment platforms , to monitor the real-time status of the virtual machines in the entire teaching experiment center.

1.3 System Architecture Design

To achieve the above goals, we propose the architecture design shown in Figure 4. When student users participate in course experiments, they can choose any node that deploys the Xen/KVM virtualization platform to use their personalized virtual machines. The virtual machine is installed with the operating system and application software that meet the needs of the course experiment. Starting a personalized virtual machine requires two parts, the main image and the sub-image. The main image file contains the system and software required by the course experiment, and the sub-image file stores the student's personalized data (including personalized software and user data, etc.). Each node stores the main image of each course, and the sub-image is stored on the storage server in the form of a mirror resource pool. The main image and sub-image are stored separately, so that students can flexibly choose computing nodes when using virtual machine services. Since the sub-mirror file is usually small, the user can also copy it to a USB flash drive for backup and recovery to use its personalized environment. In order to improve the resource utilization rate of the experimental teaching center, the personalized virtual machines of remote users are also run in the nodes. We provide interfaces to support users to access services from the outside, and provide interfaces for teachers and administrators to manage courses, students, computing resources, etc. . The system is deployed on a server cluster and includes a mirror image management module, a virtual machine management module and a metadata management module.

The mirror management module manages the organization of the primary mirror of the local mirror file repository on each node and the deployment of the sub-mirror file repository in the storage server and the sub-mirror pool. It also manages the file system of the virtual machine image.

The virtual machine management module manages the virtual machines running on each node. Such as managing virtual machine operation, creating virtual machine snapshots, migrating virtual machines, and providing a unified virtual machine management interface for local and remote users.

The metadata management module manages the object information stored in the database server (such as students, teachers, nodes, virtual machines) and the relationship between objects (such as course selection, the relationship between the student's personalized image and the main image of the course, etc.).

The management modules cooperate with each other to provide services for users to use the management interface: the image management module provides the main image and sub-image for the virtual machine management module; the metadata management module maintains the mapping between the course main image and personalized sub-image managed by the image management module relationship; the metadata management module also maintains the relationship between the student's personalized virtual machine and its related image files managed by the virtual machine management module.

The system provides three types of user access interfaces according to the different needs of students, teachers and administrators. Students mainly focus on the use of virtual machines, teachers mainly focus on the configuration and deployment of the main image of the course, and administrators need to manage the operation status of the entire experimental environment.

Compared with prior art, positive effect of the present invention is:

The present invention can help to shield the underlying physical hardware differences through virtualization technology, and provide diversified execution environments for upper-level users, and the virtual machines are isolated from each other, so users can use private virtual machines as the upper-machine platform to freely install personalized software , user data is stored in the virtual machine image, which can ensure data security.

Description of drawings

Figure 1 is a schematic diagram of students' functional requirements;

Figure 2 is a schematic diagram of the teacher's functional requirements;

Figure 3 is a schematic diagram of administrator functional requirements;

Figure 4 is a schematic diagram of the system architecture;

Fig. 5 is a mirror image template tree structure diagram;

Fig. 6 is a schematic diagram of students using virtual machines;

Figure 7 is an overview of the class diagram;

Fig. 8 is a flow chart of data processing;

Figure 9 is a flow chart of course selection.

Detailed ways

According to the design requirements, the modules of the system are as follows: course image management module, student personalized computer module, administrator monitoring management module.

Course Mirroring Management Module

Administrator users are mainly involved in the creation and management of courses and master images. We call the master image installed with the operating system and application software required by the course the course image template. The basic usage process of the administrator user is: authentication login, management of basic course information, and management of the main image template of the course. The most critical issue is to prepare the master image template of the course.

In order to efficiently prepare the main image template of the course, we propose the concept of image template tree. Figure 5 shows the image template tree of the system, wherein the root node is a blank image file; the node with a depth of 1 is an image template with only various common distribution operating systems installed; when the node depth k>=1, its The child node with a depth of k+1 is based on the image template derived from installing additional software on it, and the software set contained in the parent node is a proper subset of the child node software set. We build and organize the template repository for the course master image based on the concept of an image template tree.

The process for an administrator user to prepare a mirror image template will be carried out around the template tree. The system pre-prepared some mirror templates by investigating the course offerings in previous years. First, the teacher user submits the course template configuration to specify the operating system and software set required by the experiment, and then the administrator queries the mirror template tree according to the teacher user's needs. The query process can usually find mirror templates that meet the course requirements in the template library pre-deployed by the system. If the match fails, return the maximum proper subset template of the desired mirror template. At the same time, the system also supports teacher users to prepare mirror image templates by themselves. Teacher users can choose the following two ways to prepare templates:

The first way is to configure mirror templates online for teachers. According to the above query results, the teacher can use the returned template to start the virtual machine, and only need to install a small amount of software to complete the preparation of the image template, thereby improving the efficiency of template preparation.

The second way is that the teacher prepares the image template independently, then uploads it to the image library, and adds it to the template library after the image management module converts the format. The image formats supported for uploading include RAW, IMG, VMDK, QCOW2, etc.

When the course image template is prepared, the system will distribute it to the local image file library of each node, store it as /localTemps/teacherID/courseID/templateName.img according to the teacher and course it belongs to, and maintain the course by the metadata management module Mirror the relationship between templates and teachers, courses, etc.

Personalized hands-on modules for students

The system provides students with the function of using and managing virtual machines. Running a virtual machine requires a course master image template and a personalized sub-image. After determining the list of students participating in the course, the system will create a QCOW2-type sub-image for each student based on the main image of the course. As shown in Figure 6, students have two ways to use personalized virtual machines: client and browser.

The basic process for students to use the personalized virtual machine through the client is: 1) authenticate and log in, 2) select the current course, 3) view and manage the status of the virtual machine through the interactive interface, 4) connect to the virtual machine through the Spice client or VNCViewer, etc. Desktop for course experiments.

Students can also use their personalized virtual machines through a browser: A) complete the authentication login on the system web service home page, B) select the current course, C) view and manage the personalized virtual machine, D) obtain the virtual machine on the system web service page machine desktop, E) use the virtual machine in the browser.

The process for students to use a personalized virtual machine mainly involves three key issues: managing the virtual machine QCOW2 image, controlling the virtual machine, and connecting to the virtual machine desktop.

To manage QCOW2 images, on the one hand, it is necessary to manage the QCOW2 image file system. The image management module of the system is based on libguestfs API (libguestfs is a set of APIs for accessing and modifying virtual machine image files) and tools such as qemu-img to manage QCOW2 files, and supports viewing and editing Files in the mirror, monitoring the idle status of the mirror file system, etc. On the other hand, the relationship between the student QCOW2 image and the course image template needs to be maintained. According to the student and course information, the personalized sub-mirror is stored as /gluster/studentID/courseID/QCOW2Name.QCOW2, and the relationship between mirror, student and course is maintained by the metadata management module.

Virtual machine control includes virtual machine power management, virtual machine image snapshot management, virtual machine migration, etc. The pVCE virtual machine management module realizes the control of the virtual machine through the libvirt API, and can support various virtualization platforms such as Xen/KVM.

Connect to the virtual machine desktop. When using the client, students can connect to the virtual machine desktop through the pVCE client installed on each node such as Spicec and VNCViewer, or use the virtual machine in ssh mode, or connect through other remote desktop access tools (such as Windows Remote Desktop) Virtual machine desktop. For web access, we have integrated noVNC, spice-html5 and other tools to support students to use the virtual machine directly through a browser without installing any software.

Administrator Monitoring Management Module

The administrator mainly participates in the following three aspects of work in the system:

(1) Maintain the mirror template library. By investigating the requirements of the operating system and software in the experimental environment of previous courses, the administrator will pre-prepare a part of the image template to initialize the system's image template library.

(2) Manage all objects in the system and the relationship between objects, and control the life cycle of objects. The class diagram in Figure 7 describes the main classes of the system and shows the relationships between the classes. The administrator uses the metadata management module to manage the system through the interactive interface.

(3) Monitoring system. The virtual machine management module and the image management module are responsible for recording system information (such as node load status, virtual machine running status and capacity status of the mirror template library, etc.) to help system developers and administrators optimize system operation. For example, the system can graphically display the load status of each node, which is convenient for the administrator to migrate some virtual machines to balance the system load.

Design and implementation of metadata management

overview

The objects in the system include: course information, student user information, teacher information, course selection information, course master image file information, student personalized image information, physical node information, virtual machine running instance information, etc., which are stored in the database table are collectively referred to as metadata information in the system. The metabase management module not only manages the various objects mentioned above, but also needs to manage the relationship between objects, so as to support the correct operation of the virtual machine management module and the image file management module. It can provide a relatively complete system monitoring and management mechanism for the front-end administrator user interface, such as system resource status monitoring, storage server capacity monitoring, and network traffic monitoring.

Database Design

The tables in the system are summarized in Table 1:

English table name Chinese table name Function Description pVCE_student Student User Form Store student user information pVCE_teacher Teacher User Table Store teacher user information pVCE_host physical node table Store physical node information

pVCE_course Class Schedule Store course information pVCE_flavor configuration table Store virtual common configuration list pVCE_selective Course selection relationship table Store course selection relationship information

Table 2 is the student user table

Table 3 is the teacher user table

Table 4 is the physical node table

Table 5 is the curriculum schedule

Table 6 is the configuration table

Table 7 is the course selection relationship table

Data processing flow

students, teachers, courses

As shown in Figure 8, the metadata about students, teachers, and courses are initially completed by administrators, and later in the use process of students and teachers, they can modify their own information, such as email addresses, login passwords, etc. The image file entity corresponding to the main image name of the course indicated in the metadata database is organized by the teacher or administrator to a specific file path, and the naming rules are organized according to the teacher ID/course ID/image name.

Course selection management

As shown in Figure 9, the maintenance of the course selection relationship starts after the student, teacher user, and course records are created. That is, when creating a course, it is necessary to ensure that the course master image file has been deployed on each physical node, and Make sure that there is a teacher user in the system as the main teacher who opens the course. In order to facilitate the management of administrators and teachers, the system has developed the function of importing student course selection list. Submitting the student course selection list for a certain course can generate course selection relationship with one click. The most important step in generating course selection relationship is to participate in this course. Each student has a personalized incremental image file based on the basic image file of the virtual machine of the course, and organizes the incremental image file according to certain rules. The naming rules we adopt in the system It is: student ID/course ID/student incremental image file name.

Image Metadata Management

The management of image metadata is related to the process that students use the records in the system database to obtain the two image files involved in their course virtual machines after login authentication.

The basic image is related to the course. Although the basic image file name is specified when creating the course, we need to confirm whether the image is actually on each physical node.

At the same time, the process of generating incremental mirroring is also based on the records in the course selection relational database table. It is also necessary to ensure that the mirroring file can match the main mirroring of the course and can support the startup of the virtual machine.

Virtual Machine Metadata Management

The management of nodes in the system is mainly divided into two categories: one is physical nodes. When the administrator initializes the system, he will configure the ip of all physical nodes in the system, and then input them into the data table of the physical nodes; The second category is the management of virtual machine nodes. When a virtual machine is running, it depends on the physical node. Therefore, it is possible to discover which virtual machines of which users are running on a certain physical node according to the virtual machine management interface such as libvirt. In order to facilitate the management of runtime resources in the system, the system will track and record the running status of the virtual machine.

Claims (6)

1. based on virtual machine, realize a upper machine environment method for building up for Computer Experiment Teaching, the steps include:

1) student completes to authenticate and logs on system Web service homepage, then selects current course; Wherein, the main image file of course has been deployed on each physical node;

2) by interactive interface, check and managing virtual machines state;

3) by the system Web service page, obtain virtual machine desktop; Then in browser, use virtual machine.

2. the method for claim 1, is characterized in that by Spice client or VNCViewer, being connected to virtual machine desktop carries out curriculum experiment.

3. the method for claim 1, is characterized in that described system Web service homepage is provided with monobasic database table; In described metadatabase table, store curriculum information, User information, teacher's information, curricula-variable information, course primary mirror as fileinfo, student individuality Mirror Info, physical node information, virtual machine running example information.

4. method as claimed in claim 3, is characterized in that each course primary mirror picture name indicated in described metadatabase is corresponding with an image file entity respectively.

5. the method for claim 1, is characterized in that on described system Web service homepage more previously prepared mirror image templates and sets up a mirror image template tree; First by teacher user, submit to course template configuration to indicate operating system and the software collection information of requirement of experiment, keeper goes to inquire about according to teacher's user's request the template that coupling was set and returned to mirror image template subsequently, as it fails to match, return to the maximum proper subclass template of required mirror image template.

6. method as claimed in claim 5, is characterized in that the method that builds described mirror image template tree is: first initialization one root node is blank image file; The degree of depth is that 1 node is for only having installed the mirror image template of all kinds of common distribution version operating systems; When node degree of depth k>=1, the child that its degree of depth is k+1 is, based on it, the derivative mirror image template of additional software is installed, and father node comprises the proper subclass that software collection is child node software collection.