PROJECT TITLE :

iConSnap: An Incremental Continuous Snapshots System for Virtual Machines

ABSTRACT:

In cloud environments, one of the most important concerns is the dependability of the data and services that are hosted on a virtual machine (VM). Continuous Snapshots are the method of choice for protecting long-running systems because they minimize the amount of data that is lost in the event of a system failure. The currently available methods, on the other hand, are plagued by significant performance loss, significant VM downtime, and long snapshot intervals. This article introduces iConSnap, a system that was developed to take fine-grained continuous snapshots of virtual machines without affecting the performance of the VMs. First, iConSnap uses a copy-on-write (COW) mechanism to save the memory pages on-demand, which brings the VM downtime to about 200 milliseconds and reduces the amount of time it takes to backup the virtual machine. Second, we extend the concept of COW and propose a lazily incremental method to save the delta data between two successive snapshots only once. This results in a significant reduction in both the duration of the snapshot and the amount of data stored in the snapshot. Third, we suggest a scheduling mechanism as a potential solution to the problem of the VM's performance penalty. In conclusion, we present a method that reduces storage costs through the combination of compression and a time-aware multi-granularity reclamation strategy. This method maintains performance and availability while cutting costs. We evaluate iConSnap by implementing it on QEMU/KVM and then conducting a series of experiments with it. The results of the experiments show that iConSnap is superior to other methods in terms of the amount of VM downtime, the duration of snapshots, the costs of storage, and the performance of VMs.