Considerable research has been directed toward the study of consistent hashing and distributed hash tables. While many useful research results have emerged from this work, existing solutions can be improved in the areas of time efficiency, system growth and change to input distributions. For resolution, systems such as Chord [69][70], Memcached [23] and others use a binary search over the set of intervals to determine the node. Also, relying on a pseudo-random designation of partitions on the continuum can result in poor worst-case time performance due to load imbalance. The work proposes F^2DR, a system that maps an arithmetic distribution of intervals on a continuum to a fluid set of nodes. Any point on the continuum can be resolved to a node in O(1) time, and O(n) space. The system also contains flexible mechanisms for adapting to load patterns through dynamic restructuring. In all, F2DR provides a fresh formulation of consistent hashing that offers several advantages over previous work. / School of Computer Science, University of Guelph
| Identifer | oai:union.ndltd.org:LACETR/oai:collectionscanada.gc.ca:OGU.10214/5327 |
| Date | 17 January 2013 |
| Creators | Rea, Dana |
| Contributors | Calvert, David A. |
| Source Sets | Library and Archives Canada ETDs Repository / Centre d'archives des thèses électroniques de Bibliothèque et Archives Canada |
| Language | English |
| Detected Language | English |
| Type | Thesis |
| Rights | http://creativecommons.org/licenses/by-nc/2.5/ca/ |
Page generated in 0.0016 seconds