The multicast policy and its relationship to replicated data placement
ACM Transactions on Database Systems (TODS)
Optimal Placement of Replicas in Trees with Read, Write, and Storage Costs
IEEE Transactions on Parallel and Distributed Systems
Optimal allocation of electronic content
Computer Networks: The International Journal of Computer and Telecommunications Networking
Choosing Replica Placement Heuristics for Wide-Area Systems
ICDCS '04 Proceedings of the 24th International Conference on Distributed Computing Systems (ICDCS'04)
Optimal Placement of Replicas in Data Grid Environments with Locality Assurance
ICPADS '06 Proceedings of the 12th International Conference on Parallel and Distributed Systems - Volume 1
Replicated Server Placement with QoS Constraints
IEEE Transactions on Parallel and Distributed Systems
Optimal replica placement in hierarchical Data Grids with locality assurance
Journal of Parallel and Distributed Computing
Replica Placement and Access Policies in Tree Networks
IEEE Transactions on Parallel and Distributed Systems
| Hi-index | 0.00 |
In this paper, we discuss and compare several policies to place replicas in tree networks subject to server capacity. The server capacities and client requests are known beforehand, while the number and location of the servers are to be determined. The standard approach in the literature is to enforce that all requests of a client be served by a single server in the tree (Single ). One major contribution of this paper is to assess the impact of a new policy in which requests of a given client can be processed by multiple servers (Multiple ), thus distributing the processing of requests over the platform. We characterize problem instances for which Multiple cannot be more than two times better than the optimal Single solution, if this latter exists. For such instances, we provide a procedure which builds a Single solution with a guarantee on its cost. This is a very interesting result for applications which do not accept multiple servers for a given client, since it might be more difficult to implement such a complex strategy.