Myriad: cost-effective disaster tolerance
Fay W. Chang, Minwen Ji, Shun-Tak A. Leung, John MacCormick, Sharon E. Perl, Li Zhang · 2002
This paper proposes a new approach for achieving dis-aster tolerance in large, geographically-distributed stor-age systems. The system, called Myriad, can achieve the same level of disaster tolerance as a typical single mir-rored solution, but uses considerably fewer physical re-sources, by employing cross-site checksums (via erasure codes) instead of direct replication. The key technical contribution of the paper is a proto-col permitting cross-site checksums to be updated in such a way that data recovery is always possible. Another im-portant contribution is the specification of a protocol for recovering from disasters, explicitly verifying the claim of disaster tolerance. Further, it is shown by direct cal-culation and analytical modeling that Myriad compares favorably with mirroring in terms of both total cost of ownership and reliability. 1