Dynamic Parallel Schedules (DPS) is a flow graph based framework for developing parallel applications on clusters of workstations. The DPS flow graph execution model enables automatic pipelined parallel execution of applications. DPS supports graceful degradation of parallel applications in case of node failures. The fault-tolerance mechanism relies on a set of backup threads stored in the vola...

The microgrids are composed of small scale fueled generation capacities, renewable energy sources, storage energy systems, controllable loads, and autonomously can connect or disconnect from the mains supply. The microgrids can operate connected to the upstream main grid, or in an islanded operation mode following a large perturbation in the upstream grid. The microgrid analyzed in this paper i...

Abstmct Our success in deploying the illusion of infinite stomge to applications rests in the use of database tebhnology. Tht3 paper presents the support for transactions in the ADSTAR Dhtributed Stomge Manager, (A DSiU) system. For a user, ADSM offers a backup and archive service in a heterogeneous cliint-server environment. It also operates as a file migratton mposito~ in some Unix environmen...

Classical cloud storage based on external data providers has been recognized to suffer from a number of drawbacks. This is due to its inherent centralized architecture which makes it vulnerable to external attacks, malware, technical failures, as well to the large premium charged for business purposes. In this paper, we propose an alternative distributed peer-to-peer cloud storage model which i...

In “Using free web storage for data backup” by Traeger et al., they developed crawlable HTML pages that contained encoded (and optionally encrypted) files that could be cached by a search engine and retrieved later on. As the figure below shows, their program could convert a file into an image or plain text. They found that HTML was the best medium to use since images and other binary documents...

We introduce a method for computing the difference between a pattern string (data block or file) and a much larger, iteratively updated set of strings (an entire file system or cache thereof), while only examining a small fraction of this set. We show how our method generalizes existing algorithms used in a unique block storage system (venti), and a low bandwidth network file system (LBFS), and...

iSCSI is one of the most popular protocols among the storage area network (SAN) working on an IP network. iSCSI is a standard to encapsulate a SCSI command into a TCP/IP packet, thus we can make an access to a storage system with commoditized IP devices only. In this paper, iSCSI sequential write access for a remote backup system is focused. It is well known that iSCSI experiences performance d...

Data deduplication is a technology that eliminates redundant data to save storage space. Most previous studies on data deduplication target backup storage, where the deduplication ratio and throughput are important. However, data deduplication on primary storage has recently been receiving attention; in this case, I/O latency should be considered equally with the deduplication ratio. Unfortunat...

This thesis presents the design and implementation of the All-or-Nothing package transform, presented by Ronald Rivest in 1997 [24], and its subsequently integration in an existing peer-to-peer backup system. This encryption mode provides a third alternate way of ensuring confidentiality, integrity and availability of backups. Resilia, the existing prototype where this extension is built on, co...