We consider the problem of task assignment in a distributed system (such as a distributed Web server) in which task sizes are drawn from a heavy-tailed distribution. Many task assignment algorithms are based on the heuristic that balancing the load at the server hosts will result in optimal performance. We show this conventional wisdom is less true when the task size distribution is heavy-tailed (as is the case for Web file sizes). We introduce a new task assignment policy, called Size Interval Task Assignment with Variable Load (SITA-V). SITA-V purposely operates the server hosts at different loads, and directs smaller tasks to the lighter-loaded hosts.

In this paper, we consider the problem of modeling machine availability in enterprise-area and wide-area distributed computing settings. Using availability data gathered from three different environments, we detail the suitability of four potential statistical distributions for each data set: exponential, Pareto, Weibull, and hyperexponential. In each case, we use software we have developed to determine the necessary parameters automatically from each data collection.

In our quest to better understand network tra#c dynamics, we examine Internet chat systems. Although chat as an application does not contribute huge amounts of tra#c, chat systems are known to be habit-forming. This implies that catering to such users can be a promising way of attracting them, especially in low bandwidth environments such as wireless networks.

In this paper we develop a general methodology for characterizing the access patterns of Web server requests based on a time-series analysis of finite collections of observed data from real systems. Our approach is used together with the access logs from the IBM Web site for the Olympic Games to demonstrate some of its advantages over previous methods and to construct a particular class of benchmarks for large-scale heavily-accessed Web server environments. We then apply an instance of this class of benchmarks to analyze aspects of large-scale Web server performance, demonstrating some additional problems with commonly used methods to evaluate Web server performance at different request traffic intensities.

We review evidence that Internet traffic is characterized by long-tailed distributions of interarrival times, transfer times, burst sizes and burst lengths. We propose a new statistical technique for identifying long-tailed distributions, and apply it to a variety of datasets collected on the Internet. We find that there is little evidence that interarrival times and transfer times are long-tailed, but that there is some evidence for long-tailed burst sizes. We speculate on the causes of long-tailed bursts. I.

Over the last decade an important new...

Approximating general distributions by phase-type (PH) distributions is a popular technique in queueing analysis, since the Markovian property of PH distributions often allows analytical tractability.

Abstract not found

Abstract—Long-range dependence has been observed in many recent Internet traffic measurements. In addition, some recent studies have shown that under certain network conditions, TCP itself can produce traffic that exhibits dependence over limited timescales, even in the absence of higher-level variability. In this paper, we use a simple Markovian model to argue that when the loss rate is relatively high, TCP’s adaptive congestion control mechanism indeed generates traffic with OFF periods exhibiting power-law shape over several timescales and thus introduces pseudo-long-range dependence into the overall traffic. Moreover, we observe that more variable initial retransmission timeout values for different packets introduces more variable packet inter-arrival times, which increases the burstiness of the overall traffic. We can thus explain why a single TCP connection can produce a time-series that can be misidentified as self-similar using standard tests.

In recent years, several studies have shown that network trac exhibits the property of self-similarity. Traditional (Poissonian) modelling approaches have been shown not to be able to describe this property and generally lead to the underestimation of interesting performance measures. Crovella and Bestavros have shown