TY - GEN

T1 - Balanced-replication algorithms for distribution trees

AU - Cohen, Edith

AU - Kaplan, Haim

N1 - Publisher Copyright:
© Springer-Verlag Berlin Heidelberg 2002.

PY - 2002

Y1 - 2002

N2 - In many Internet applications, requests for a certain object are routed bottom-upover a tree where the root of the tree is the node containing the object. When an object becomes popular, the root node of the tree may become a hotspot. Therefore many applications allow intermediate nodes to acquire the ability to serve the requests, for example by caching the object. We call such distinguished nodes primed. We propose and analyse different algorithms where nodes decide when to become primed; these algorithms balance the maximum load on a node and the number of primed nodes. Many applications require both fully distributed decisions and smooth convergence to a stable set of primed nodes. We first present optimal algorithms which require communication across the tree. We then consider the natural previously proposed THRESHOLD algorithm, where a node becomes primed when the incoming flow of requests exceeds a threshold. We show examples where THRESHOLD exhibits undesirable behavior during convergence. Finally, we propose another fully distributed algorithm, GAP, which converges gracefully.

AB - In many Internet applications, requests for a certain object are routed bottom-upover a tree where the root of the tree is the node containing the object. When an object becomes popular, the root node of the tree may become a hotspot. Therefore many applications allow intermediate nodes to acquire the ability to serve the requests, for example by caching the object. We call such distinguished nodes primed. We propose and analyse different algorithms where nodes decide when to become primed; these algorithms balance the maximum load on a node and the number of primed nodes. Many applications require both fully distributed decisions and smooth convergence to a stable set of primed nodes. We first present optimal algorithms which require communication across the tree. We then consider the natural previously proposed THRESHOLD algorithm, where a node becomes primed when the incoming flow of requests exceeds a threshold. We show examples where THRESHOLD exhibits undesirable behavior during convergence. Finally, we propose another fully distributed algorithm, GAP, which converges gracefully.

UR - http://www.scopus.com/inward/record.url?scp=84938080145&partnerID=8YFLogxK

U2 - 10.1007/3-540-45749-6_29

DO - 10.1007/3-540-45749-6_29

M3 - ???researchoutput.researchoutputtypes.contributiontobookanthology.conference???

AN - SCOPUS:84938080145

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 297

EP - 309

BT - Algorithms - ESA 2002 - 10th Annual European Symposium, Proceedings

A2 - Möhring, Rolf

A2 - Raman, Rajeev

PB - Springer Verlag

T2 - 10th Annual European Symposium on Algorithms, ESA 2002

Y2 - 17 September 2002 through 21 September 2002

ER -