Performance evaluation and asymptotics for Content Delivery Networks

@article{Shah2014PerformanceEA,
  title={Performance evaluation and asymptotics for Content Delivery Networks},
  author={Virag Shah and Gustavo de Veciana},
  journal={IEEE INFOCOM 2014 - IEEE Conference on Computer Communications},
  year={2014},
  pages={2607-2615}
}
  • Virag Shah, G. Veciana
  • Published 2014
  • Computer Science
  • IEEE INFOCOM 2014 - IEEE Conference on Computer Communications
Large scale Content Delivery Networks (CDNs) are one of the key components of today's information infrastructure. This paper proposes and analyzes a simple stochastic model for a file-server system wherein servers can work together, as a pooled resource, to meet individual user requests. In such systems basic questions include: How and where to replicate files? What is the impact of dynamic service allocation across request types, and whether it can provide substantial gains over simpler load… 

Figures from this paper

Impact of Fairness and Heterogeneity on Delays in Large-scale Content Delivery Networks
TLDR
This work considers multi-class queueing systems where the per class service rates depend on the network state, fairness criterion, and is constrained to be in a symmetric polymatroid capacity region, and studies performance robustness to heterogeneity in per class loads and fairness criteria.
Resource Pooling in Large-Scale Content Delivery Systems
TLDR
It is shown that the benefits of resource pooling depend on the popularity profile of the contents offered by the VoD service, and if the popularity does not vary drastically across contents, then resource Pooling leads to an order wise reduction in central server transmission rate as the system size grows.
Resource pooling in large-scale content delivery systems
TLDR
It is shown that the benefits of resource pooling depend on the popularity profile of the contents offered by the VoD service, and if the popularity does not vary drastically across contents, resource Pooling can lead to significant improvements in the system performance.
Fair throughput allocation in Information-Centric Networks
Impact of fairness and heterogeneity on delays in large-scale centralized content delivery systems
TLDR
It is shown that large randomly configured systems with a logarithmic number of file copies are robust to substantial load and server heterogeneities for a class of fairness criteria.
Distributed Server Allocation for Content Delivery Networks
TLDR
A dynamic formulation of file-sharing networks in terms of an average cost Markov decision process with constraints and an index policy in the spirit of Whittle is proposed and compared by simulations with other natural heuristics.
Caching Partial Files for Content Delivery
TLDR
This work treats incoming requests as requests for segments of files and proposes a Markovian request model which captures the time-correlation in requests, characterized as the fundamental limit on the performance of caching policies which only cache full files.
Statistical Multiplexing and Traffic Shaping Games for Network Slicing
TLDR
This work provides a solution to robust share dimensioning for SCPF-based network slicing, which can be modeled as a “traffic shaping” game with an achievable Nash equilibrium under high loads.
Statistical multiplexing and traffic shaping games for network slicing
TLDR
A solution to robust share dimensioning for SCPF-based network slicing and how to unilaterally manage their users' performance via admission control can be modeled as a "traffic shaping" game with an achievable Nash equilibrium.
Asymptotic independence of servers’ activity in queueing systems with limited resource pooling
TLDR
For large-scale cloud systems where heterogeneous pools of servers collaborate in serving jobs of diverse classes, a concentration in server activity indicates that the overall power and network capacity that need to be provisioned may be substantially lower than the worst case, thus reducing costs.
...
1
2
3
...

References

SHOWING 1-10 OF 23 REFERENCES
Bipartite graph structures for efficient balancing of heterogeneous loads
TLDR
This paper describes the asymptotic system efficiency under an optimal matching policy, and shows that -in contrast to greedy matching- optimal matching incurs an inefficiency that is exponentially small in the server storage size, even at critical system loads.
Insensitive Bandwidth Sharing in Data Networks
TLDR
This work uses some key properties of Whittle queueing networks to characterize the class of allocations which are insensitive in the sense that the stationary distribution of this stochastic process does not depend on any traffic characteristics except the traffic intensity on each route.
Calculating the flow level performance of balanced fairness in tree networks
An infinite server system with customer-to-server packing constraints
  • A. Stolyar
  • Mathematics
    2012 50th Annual Allerton Conference on Communication, Control, and Computing (Allerton)
  • 2012
TLDR
This paper shows that some versions of a greedy strategy are asymptotically optimal as the system scale (the total customer arrival rate) goes to infinity.
Power-aware speed scaling in processor sharing systems: Optimality and robustness
Randomized load balancing with general service time distributions
TLDR
A modularized program based on an ansatz which asserts that any finite set of queues in a randomized load balancing scheme becomes independent as n to ∞ is described, which allows one to derive queue size distributions and other performance measures of interest.
A queueing analysis of max-min fairness, proportional fairness and balanced fairness
TLDR
This paper compares the performance of three usual allocations, namely max-min fairness, proportional fairness and balanced fairness, in a communication network whose resources are shared by a random number of data flows and shows this model is representative of a rich class of wired and wireless networks.
Path Selection and Multipath Congestion Control
TLDR
This paper investigates the potential benefits of coordinated congestion control for multipath data transfers, and shows that with RTT bias (as in TCP Reno), uncoordinated control can lead to inefficient equilibria, and introduces the notion of a Nash equilibrium.
Dynamic right-sizing for power-proportional data centers
TLDR
It is proved that the optimal offline algorithm for dynamic right-sizing has a simple structure when viewed in reverse time, and this structure is exploited to develop a new ‘lazy’ online algorithm, which is proven to be 3-competitive.
Multi-Path TCP: A Joint Congestion Control and Routing Scheme to Exploit Path Diversity in the Internet
TLDR
This paper presents a class of algorithms that can be implemented at the sources to stably and optimally split the flow between each source-destination pair and shows that the connection-level throughput region of such multi-path routing/congestion control algorithms can be larger than that of a single-path congestion control scheme.
...
1
2
3
...