Optimizing Performance in Migrating Data between Non-cloud Infrastructure and Cloud Using Parallel Computing
Mimouna Alkhonaini, Hoda El-Sayed · 2018
Clouds have offered users a very rich and flexible internet computing experience. Therefore, many individuals, companies, organizations and institutions migrated from traditional local computing services to the cloud environment. Despite the numerous benefits of cloud computing, there are still major concerns when it comes to security. Among the major security vulnerabilities is the migration process from the physical local servers to the cloud servers. During this migration, data can be lost due to corruption, incomplete transmission and interception. To address these threats, we have designed and implemented a system that will secure the data throughout the migration process. The method being implemented combines different security techniques to ensure that data is fully transferred from the source non-cloud computing environment to the cloud servers without being corrupted or intercepted by a malicious third party. The combination of data segmentation, error control/correction, encryption, decryption and hashing of data, which are applied in our system will ensure that the highest level of data security during the migration process is attained. However, due to the associated heavy computation processes, the whole system can slow down, which can result in poor and inefficient performance. Most chips today are multicore and parallelization, therefore, can be leveraged to enhance the overall system performance and efficiency. In this research, we detail the security improvements of our system and demonstrate the implementations and performance improvements of the parallel implementation.