Evaluating Update Dissemination Strategies in Distributed Systems: A Focus on Anti-Entropy Protocol (December 2023)
Clarian Makungu
Abstract
In the realm of distributed systems, efficient update dissemination stands as a critical challenge. The Anti-Entropy Protocol, a prominent strategy within epidemic protocols, strives to maintain system-wide consistency by swiftly updating or removing susceptible servers. This study investigates three primary update exchange strategies—pushing updates, pulling updates, and bidirectional pushing-pulling—emphasizing the limitations of pushing updates in contrast to the other methods.
Keywords
distributed systems
update dissemination
the Anti-Entropy Protocol
epidemic
protocols
pushing updates
pulling updates
bidirectional communication
system-wide
consistency
efficiency
goss
References
[1] Abu-Libdeh, H., P. Costa, A. Rowstron, G. O’Shea, & A. Donnelly (2010, August). Symbiotic
routing in future data centers. SIGCOMM Comput. Commun. Rev. 41, 51–62.
[2] Agrawal, D., A. El Abbadi, & R. C. Steinke (1997). Epidemic algorithms in replicated
databases (extended abstract). In Proc. of the sixteenth ACM SIGACT-SIGMOD-SIGART
symposium on Principles of database systems, PODS ’97, New York, NY, USA, pp. 161– 172.
ACM.
[3] Benson, T., A. Akella, & D. A. Maltz (2010). Network traffic characteristics of data centers in
the wild. In IMC 2010, pp. 267–280.
routing in future data centers. SIGCOMM Comput. Commun. Rev. 41, 51–62.
[2] Agrawal, D., A. El Abbadi, & R. C. Steinke (1997). Epidemic algorithms in replicated
databases (extended abstract). In Proc. of the sixteenth ACM SIGACT-SIGMOD-SIGART
symposium on Principles of database systems, PODS ’97, New York, NY, USA, pp. 161– 172.
ACM.
[3] Benson, T., A. Akella, & D. A. Maltz (2010). Network traffic characteristics of data centers in
the wild. In IMC 2010, pp. 267–280.
