Resumption of virtual machines after adaptive deduplication of virtual machine images in live migration

Naga Malleswari TYJ, Senthil Kumar T., JothiKumar C

Abstract


Cloud Computing is a technology where resources are provided to the users from its pool according to their demand and requirements. During this process, several virtual machines (VMs)are created and destroyed. Due to this, servers are heavily loaded and under-loaded. Thus, balancing the load and energy utilization are the critical problems in cloud computing and are solved by virtual machine migration. The movement of virtual machines from an underloaded machine to the other suitable physical machine that is not yet loaded heavily is known as virtual machine migration. If the migration process carried out without shutting down the power (live) of the VMs, it is known as live virtual machine migration or Live Migration. During this process, the virtual disk image files which are larger size take more significant time to transfer to the target host and hence more migration time and downtime. Some researchers used deduplication methods for decreasing the size of the virtual disk image file in the live VM migration process. In the proposed adaptive deduplication technique, every virtual disk image file undergoes both block-level deduplication processes such as fixed length and variable length, based on the size of the disk file, and the size reduced. The significance of this paper is to deduplicate the disk image files by adaptive deduplication process at source host, transfer the deduplicated pages to the target host—the deduplicated pages reunited at target host into a single virtual disk image file and the VM is resumed. The performance of this process measured by calculating the percentage reduction of the size of the VM image after deduplication, the time is taken to migrate the deduplicated file and the time taken for each VM to resume at the target host after the migration The results show 83 % of overall reduction in image storage, 89.76% reduction in migration time. For a deduplication ratio of 92%, it takes an overall time of 3.52 minutes, which is a 7% reduction in resumption time when compared with the time taken for the total QCOW2 files with original size. For VMDK files the resumption time is reduced by a maximum 17% (7.63 mins) when compared with that of for original files.

Keywords


Computing; Cloud Computing; Pre-copy approach; Load balancing; Deduplication; Streaming Analytics; Akka stream



DOI: http://doi.org/10.11591/ijece.v11i1.pp%25p
Total views : 0 times


Creative Commons License
This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License.

ISSN 2088-8708, e-ISSN 2722-2578