Total recall: system support for automated availability management
NSDI'04 Proceedings of the 1st conference on Symposium on Networked Systems Design and Implementation - Volume 1
A Practical Study of Regenerating Codes for Peer-to-Peer Backup Systems
ICDCS '09 Proceedings of the 2009 29th IEEE International Conference on Distributed Computing Systems
High availability in DHTs: erasure coding vs. replication
IPTPS'05 Proceedings of the 4th international conference on Peer-to-Peer Systems
Measuring bandwidth between planetlab nodes
PAM'05 Proceedings of the 6th international conference on Passive and Active Network Measurement
Hi-index | 0.00 |
Distributed storage systems provide large-scale reliable data storage by storing a certain degree of redundancy in a decentralized fashion on a group of storage nodes. To recover from data losses due to the instability of these nodes, whenever a node leaves the system, additional redundancy should be regenerated to compensate such losses. In this context, the general objective is to minimize the volume of actual network traffic caused by such regenerations. A class of codes, called regenerating codes, has been proposed to achieve an optimal trade-off curve between the amount of storage space required for storing redundancy and the network traffic during the regeneration. In this paper, we jointly consider the choices of regenerating codes and network topologies. We propose a new design, referred to as RCTREE, that combines the advantage of regenerating codes with a tree-structured regeneration topology. Our focus is the efficient utilization of network links, in addition to the reduction of the regeneration traffic. With the extensive analysis and quantitative evaluations, we show that RCTREE is able to achieve a both fast and stable regeneration, even with departures of storage nodes during the regeneration.