by Felix C. Gartner, Henning Pagnia
http://www.informatik.tu-darmstadt.de/~felix/publications/wss1999b.ps.gz
Add To MetaCart
Abstract:
Usually, load distribution schemes for replicated servers are based on a many-to-one mapping between client and server, meaning that while a server may serve many clients, a client has a single specific server which it queries at any point in time. In some cases, however, it is desirable that the number of accesses of a client may be distributed over multiple servers, thus yielding a many-to-many mapping between clients and servers. In this paper, we present a simple method to efficiently realize such a many-to-many mapping between clients and servers. For the sake of transparency, we add a component called "distribution module " to the communication interface of client and server. This module is responsible for distributing server accesses over multiple target machines in a well defined way. We present algorithms for the client and server component and show that they are self-stabilizing, meaning that they converge to a stable state once the access pattern becomes regular. Due to this property, the components can tolerate any internal transient fault in a non-masking way. Additionally, our approach is highly modular since servers may run an off-theshelf load distribution algorithm and replica consistency is not affected. 1.
Citations
|
449
|
Self-Stabilizing Systems in Spite of Distributed Control
– Dijkstra
- 1974
|
|
240
|
The case for geographical push caching
– Gwertzman, Seltzer
- 1995
|
|
230
|
Introduction to Distributed Algorithms
– Tel
- 1994
|
|
126
|
Dynamic Server Selection Using Bandwidth Probing in Wide-Area Networks
– Carter, Crovella
- 1996
|
|
111
|
Load distribution for locally distributed systems
– Shivaratri, Krueger, et al.
- 1992
|
|
101
|
Weak Consistency group communications and memberships
– Golding
- 1992
|
|
79
|
Closure and convergence: A foundation of fault-tolerant computing
– Arora, Gouda
- 1993
|
|
62
|
Dynamic server selection in the Internet
– Crovella, Carter
- 1995
|
|
36
|
Introducing Application-level Replication and Naming into Today’s Web
– Baentsch, Molter, et al.
- 1996
|
|
22
|
Constraint satisfaction as a basis for designing nonmasking faulttolerance
– Arora, Gouda, et al.
- 1996
|
|
20
|
Measured performance of the Network Time Protocol in the Internet system
– Mills
- 1990
|
|
19
|
Demand-based document dissemination for the World-Wide Web
– Bestavros
- 1995
|
|
16
|
Dynamic Load Distribution with the WINNER System
– Arndt, Freisleben, et al.
- 1998
|
|
6
|
Network time protocol (version 3). Internet Request for Comments RFC 1305
– Mills
- 1992
|
|
4
|
Load balancing: An exercise in constrained convergence
– Arora, Gouda
- 1995
|
|
3
|
DNS support for load balancing. Internet Request for Comments 1794
– Brisco
- 1995
|
|
2
|
On the correctness criteria of load balancing programs. Internet: ftp://ftp.cis.ohiostate. edu/pub/anish/papers/load-balancing.ps.gz
– Arora, Gouda
- 1997
|
|
2
|
Stepwise development of a distributed load balancing algorithms
– Gr��nning, Nielsen, et al.
- 1990
|
|
2
|
Enhancing the World-Wide Web through a transparent integration of document replication
– Theel, Pagnia, et al.
- 1998
|
