Cloud Storage Monitoring System analyzing through File Access Pattern

Abstract: Cloud computing is an important technology on current demanding business requirements and it has been emerged as unavoidable technology. The usage of IaaS Service storage for Cloud Computing is being expanding exponential every year. The cloud storages are used by the cloud user due to its economy compared with other storage methods. The replications of files helps user for easy access with high availability which reduces the overall access time of the files, but at the same time it occupies more storage space and result in high storage cost. The cloud user holds multiple times of the storage than what he is actually needed. It is a dire need of system to find unwanted files in the cloud and also optimize the storage space by evaluating through file access frequency.

This paper propose Cloud Storage Monitoring (CSM) system, which monitor the IaaS storage usage and analyze the file access patterns by various parameters to identify the frequency of access, size, future access prediction, replication of files in the cloud storage. This allocates a ranking for each file which also predicts future access pattern. This generates a recommendation dashboard to the user who can decide on the operations such as reorganize, delete or archive the files and eliminate duplicate files in the cloud storage which can increase the space for future use. This system is experimented in the CloudSim environment and validate through multiple simulations testing, by using comparison techniques related to file attributes, delta versionhashing, Data de-duplication. The ranking algorithm technique applied on frequency distribution shows that increase in the storage space up to 10.91% higher than the normal system. It also helps to forecast towards future files usability prediction and prevents the duplicate entries.