Research On Offline Update System For Virtual Machine Images In Cloud Environment

With the rapid development of cloud computing technology, large numbers of applications and services are deployed in virtual machines within the cloud environment,which leads to a large number of virtual machine images stay in the cloud platform. If the overdue software inside the virtual machines cannot be updated timely, it will affect the safety of the virtual machines. The timely software update is a fundamental method for the virtual machines to deal with security threats. In addition, many virtual machines are often offline, so the offline update for images is very important to ensure the security after the normal operation of images. The traditional software update method requires an online status of the machines, and these methods are not suitable for offline virtual machine images apparently. If a virtual machine image is manually booted for online status, the update is performed and then the image is shut down to offline state, it wastes a lot of time and this method is not suitable for cloud platforms where a large number of different types of virtual machine images need to be updated. Therefore, to achieve efficient and reliable offline update for virtual machine images is important.The virtual machine image offline update system firstly collects all software information from all images through efficient collection subsystem. In order to reflect the needs of different users, software update priority policy is designed and implemented and the system can update all software based on the update priority. Secondly, the design of a phased update method based on the separation of the script completes the update tasks for all software inside image efficiently. Finally, the reliability verification for updated image ensures that the software update is correct.The test results show that the virtual machine image offline update system has a high software collection efficiency, and it completes the information collection tasks for large numbers of images as well as for the latest software information. Also it can search all target images to be updated quickly which ensures the update tasks can be performed timely,and complete the update tasks for the software inside images efficiently, while the software support is 100%. At last, it can complete the reliability verification tasks of a large numberreduces the time for verification greatly.of updated images efficiently, which ensures the correctness of the software update and...