Recent measurement based studies reveal that most of the Internet connections are short in terms of the amount of traffic they carry (mice), while a small fraction of the connections are carrying a large portion of the traffic (elephants). A careful study of the TCP protocol shows that without help from an Active Queue Management (AQM) policy, short connections tend to lose to long connections in their competition for bandwidth. This is because short connections do not gain detailed knowledge of the network state, and therefore they are doomed to be less competitive due to the conservative nature of the TCP congestion control algorithm.

Is it possible to reduce the expected response time ofevery request at a web server, simply by changing the order in which we schedule the requests? That is the question we ask in this paper. This paper proposes a method for improving the performance of web servers servicing static HTTP requests. The idea is to give preference to those requests which are short, or have small remaining processing requirements, in accordance with the SRPT (Shortest Remaining Processing Time) scheduling policy. The implementation is at the kernel level and in-volves controlling the order in which socket buffers are drained into the network.Experiments are executed both in a LAN and a WAN environment. We use the Linux operating system and the Apache and Flash web servers. Results indicate that SRPT-based scheduling of connections yields significant reductions in delay at the web server. These result in a substantial reduction inmean response time, mean slowdown, and variance in response time for both the LAN and WAN environments. Significantly, and counter to intuition, the large requests are only negligibly penalized or not at all penalized as a result of SRPT-based scheduling.

It is common to classify scheduling policies based on their mean response times. Another important, but sometimes opposing, performance metric is a scheduling policy’s fairness. For example, a policy that biases towards short jobs so as to minimize mean response time, may end up being unfair to long jobs. In this paper we define three types of unfairness and demonstrate large classes of scheduling policies that fall into each type. We end with a discussion on which jobs are the ones being treated unfairly. 1

Most well-managed web servers perform well most of the time. Occasionally, however, every popular web server experiences transient overload. An overloaded web server typically displays signs of its affliction within a few seconds. Work enters the web server at a greater rate than the web server can complete it, causing the number of connections at the server to build up. This implies large delays for clients accessing the server. This paper provides a systematic performance study of exactly what happens when a web server is run under transient overload, both from the perspective of the server and from the perspective of the client. Second, this paper proposes and evaluates a particular kernel-level solution for improving the performance of web servers under overload. The solution is based on SRPT connection scheduling. We show that SRPT-based scheduling improves overload performance across a variety of client and server-oriented metrics.

Recent studies of Internet traffic have shown that flow size distributions often exhibit a high variability property in the sense that most of the flows are short and more than half of the total load is constituted by a small percentage of the largest flows. In the light of this observation, it is interesting to revisit scheduling policies that are known to favor small jobs in order to quantify the benefit for small and the penalty for large jobs. Among all scheduling policies that do not require knowledge of job size, the least attained service (LAS) scheduling policy is known to favor small jobs the most.

This note briey summarizes some results from two papers: [4] and [23]. These papers pose the following question: Is it possible to reduce the expected response time of every request at a web server, simply by changing the order in which we schedule the requests? In [4] we approach this question analytically via an M/G/1 queue. In [23] we approach the same question via implementation involving an Apache web server running on Linux.

We explore the performance of an M/GI/1 queue under various scheduling policies from the perspective of a new metric: the slowdown experienced by largest jobs. We consider scheduling policies that bias against large jobs, towards large jobs, and those that are fair, e.g., Processor-Sharing. We prove that as job size increases to infinity, all work conserving policies converge almost surely with respect to this metric to no more than 1=(1 \Gamma ae), where ae denotes load. We also find that the expected slowdown under any work conserving policy can be made arbitrarily close to that under Processor-Sharing, for all job sizes that are sufficiently large.

We investigate the problem of online scheduling of jobs to minimize flow time and stretch on m identical machines. We consider the case where the algorithm is given either (1 + #)m machines or m machines of speed (1 + #), for arbitrarily small #>0. We show that simple randomized and deterministic load balancing algorithms, coupled with simple single machine scheduling strategies such as SRPT (shortest remaining processing time) and SJF (shortest job first), are O(poly(1/#))-competitive for both flow time and stretch. These are the first results which prove constant factor competitive ratios for flow time or stretch with arbitrarily small resource augmentation. Both the randomized and the deterministic load balancing algorithms are nonmigratory and do immediate dispatch of jobs.

We present MobiDesk, a mobile virtual desktop computing hosting infrastructure that leverages continued improvements in network speed, cost, and ubiquity to address the complexity, cost, and mobility limitations of today’s personal computing infrastructure. MobiDesk transparently virtualizes a user’s computing session by abstracting underlying system resources in three key areas: display, operating system and network. MobiDesk provides a thin virtualization layer that decouples a user’s computing session from any particular end user device and moves all application logic from end user devices to hosting providers. MobiDesk virtualization decouples a user’s computing session from the underlying operating system and server instance, enabling high availability service by transparently migrating sessions from one server to another during server maintenance or upgrades. We have implemented a MobiDesk prototype in Linux that works with existing unmodified applications and operating system kernels. Our experimental results demonstrate that MobiDesk has very low virtualization overhead, can provide a full-featured desktop experience including full-motion video support, and is able to migrate users’ sessions efficiently and reliably for high availability, while maintaining existing network connections.

We define the class of SMART scheduling policies. These are policies that bias towards jobs with small remaining service times, jobs with small original sizes, or both, with the motivation of minimizing mean response time and/or mean slowdown. Examples of SMART policies include PSJF, SRPT, and hybrid policies such as RS (which biases according to the product of the remaining size and the original size of a job). For many policies in the SMART class, the mean response time and mean slowdown are not known or have complex representations involving multiple nested integrals, making evaluation difficult. In this work, we prove three main results. First, for all policies in the SMART class, we prove simple upper and lower bounds on mean response time. Second, we show that all policies in the SMART class, surprisingly, have very similar mean response times. Third, we show that the response times of SMART policies are largely insensitive to the variability of the job size distribution. In particular, we focus on the SRPT and PSJF policies and prove insensitive bounds in these cases.