Abstract
Grid computing emerges in part from the need to integrate a
collection of distributed computing resources to offer performance
unattainable by any single machine. Grid technology facilitates data sharing
across many organizations in different geographical locations. Data
replication is an excellent technique to move and cache data close to users.
Replication reduces access latency and bandwidth consumption. It also
facilitates load balancing and improves reliability by creating multiple data
copies. One of the challenges in data replication is to select the candidate
sites where replicas should be placed, which is known as the allocation
problem. One performance metric to determine the best place to host replicas
is select for optimum average (or aggregated ) response time. We use the
p-median model for the replica placement problem. The p-median model has been
exploited in urban planning to find locations where new facilities should be
built. In our problem, the p-median model finds the locations of p candidate
sites to place a replica that optimize the aggregated response time. A Grid
environment is highly dynamic so user requests and network latency vary
constantly. Therefore, the candidate sites currently holding replicas may not
be the best sites to fetch replica on subsequent requests. We propose a
dynamic replica maintenance algorithm that re-allocates to new candidate
sties if a performance metric degrades significantly over last K time
periods. Simulation results demonstrate that the dynamic maintenance
algorithm with static placement decisions performs best in dynamic
environments like Data Grids.
Notes
We are currently acquiring citations for the work deposited into this collection. We recognize the distribution rights of this item may have been assigned to another entity, other than the author(s) of the work.If you can provide the citation for this work or you think you own the distribution rights to this work please contact the Institutional Repository Administrator at digitize@ucalgary.ca