Abstract

The problem associated with current file archiving systems is a slow processing time owing to unnecessary data copying. To address this problem, a novel archiving system with zero-copy merging and splitting operations, referred to as AvaTar, is presented herein. For the operations, instead of copying the data, the block allocation information of the files is manipulated at the kernel level. We implemented kernel-level archiving primitives in a Linux kernel, called AvaTar-FS, and a user-level archiving tool, called AvaTar agent. Our evaluation results indicated that AvaTar required only 0.48 s to extract 1,024 files from a 4 GB archive file, which is 132-times faster when compared with traditional GNU Tar archiving. AvaTar affords practical benefits in uploading files to a real-world cloud storage system, and successfully completes the transfer of 1,024 files to Amazon Web Service cloud storage within 60.55% of the processing time required through a traditional approach.

Highlights

  • Sharing files over cloud storage is easy

  • We present the performance of the AvaTar file system, which takes advantage of a zero-copy merge and the split files

  • Because the mount utility is based on file system in user space (FUSE) [2], it can be executed without an installation or modification of the system

Read more

Summary

Introduction

Sharing files over cloud storage is easy. Once the setup is complete, numerous files placed in resources such as Dropbox or Amazon S3 directories can be automatically synchronised and immediately accessed by any computer [16]. Small and numerous file transfers are prevalent on the cloud [6], [17], and they hinder the best-utilising network bandwidth. I-nodes with block numbers are used in numerous UNIX-like traditional file systems such as Unix file system (UFS) [19], Ext2 [7], and Ext3 [26]. In these file systems, data blocks consist of direct and indirect blocks. Direct blocks are used for quick and easy access, whereas indirect blocks are used for storing a number of data blocks for large-size files

Methods
Results
Discussion
Conclusion
Full Text
Paper version not known

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call

Disclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.