We consider architecture of data center networks (DCN), which adds wireless network card to each router [1 -- 3]. Existing redundancy elimination (RE) mechanisms reduce link loads and increase network capacity in several environments by removing strings that have appeared in earlier packets through encoding and decode them several hops downstream. This article is the first to explore RE mechanisms in large-scale DCNs, and the first to exploit cooperative RE among servers. We describe a centralized REDaCeN (RE in DCNs) algorithm, which greedily and repeatedly selects to cache data unit with the largest caching utility among routers with remaining capacity. We show that approximation ratio of REDaCeN is high through our simulation. We then discuss distributed implementation of REDaCen based on the observation that servers and routers are physically close and within transmission range of each other. Servers and routers update their status in a single wireless transmission, using an efficient prioritized schedule. We also address cache replacement and consistency of information in servers and routers. Our framework is tested on different parameters and shows superior performance in comparison to other mechanisms (imported directly to this setting).
展开▼
