Search

This paper proposes a backup system based on stackable mirroring filesystem "GMFS." We have developed GMFS to mirror data in realtime on the filesystem layer.GMFS uses NFS which is a typical network file system to communicate with an existing environment, and backup data to one or more NFS servers. Therefore, an existing empty resource can be used. It is necessary to backup data to two or more servers, considering the possibility that the backup data breaks. Moreover, the throughput of read and write has been improved by adopting the method to call system function of NFS from the inside of the filesystem.The result of the experiment from the viewpoint of throughput shows the following. The throughput of the read and write of GMFS is 2.4 times faster than conventional mirroring filesystem, and the decrease of throughput is a little even if the number of backup servers increases.

Nowadays, it is highly important for us to well organize, access, and store digital data in a way which is efficient and secure. Since there are many incidents which may cause the loss of important data, backing up and safe-keeping important data into some reliable and flexibly accessible media is an extremely important practice. In this paper we propose a Backward Web-feed Based Framework (BWBF) for supporting the automatic backup and restoration of important files by using an emerging web technology – the podcast technology. A sample BWBF based backup service designed and implemented to automatically and asynchronously backup files to (and restore files from) BWBF-based remote backup servers was described in this article.