A Virtual Deadline Scheduler for Window-Constrained Service Guarantees

This paper presents a new approach to window-constrained scheduling, suitable for multimedia and weakly-hard real-time systems. We originally developed an algorithm, called Dynamic Window-Constrained Scheduling (DWCS), that attempts to guarantee no more than x out of y deadlines are missed for real-time jobs such as periodic CPU tasks, or delay-constrained packet streams. While DWCS is capable of generating a feasible window-constrained schedule that utilizes 100% of resources, it requires all jobs to have the same request periods (or intervals between successive service requests). We describe a new algorithm called Virtual Deadline Scheduling (VDS), that provides window-constrained service guarantees to jobs with potentially different request periods, while still maximizing resource utilization. VDS attempts to service m out of k job instances by their virtual deadlines, that may be some finite time after the corresponding real-time deadlines. Notwithstanding, VDS is capable of outperforming DWCS and similar algorithms, when servicing jobs with potentially different request periods. Additionally, VDS is able to limit the extent to which a fraction of all job instances are serviced late. Results from simulations show that VDS can provide better window-constrained service guarantees than other related algorithms, while still having as good or better delay bounds for all scheduled jobs. Finally, an implementation of VDS in the Linux kernel compares favorably against DWCS for a range of scheduling loads.

Application-Level Document Caching in the Internet

With the increasing demand for document transfer services such as the World Wide Web comes a need for better resource management to reduce the latency of documents in these systems. To address this need, we analyze the potential for document caching at the application level in document transfer services. We have collected traces of actual executions of Mosaic, reflecting over half a million user requests for WWW documents. Using those traces, we study the tradeoffs between caching at three levels in the system, and the potential for use of application-level information in the caching system. Our traces show that while a high hit rate in terms of URLs is achievable, a much lower hit rate is possible in terms of bytes, because most profitably-cached documents are small. We consider the performance of caching when applied at the level of individual user sessions, at the level of individual hosts, and at the level of a collection of hosts on a single LAN. We show that the performance gain achievable by caching at the session level (which is straightforward to implement) is nearly all of that achievable at the LAN level (where caching is more difficult to implement). However, when resource requirements are considered, LAN level caching becomes much more desirable, since it can achieve a given level of caching performance using a much smaller amount of cache space. Finally, we consider the use of organizational boundary information as an example of the potential for use of application-level information in caching. Our results suggest that distinguishing between documents produced locally and those produced remotely can provide useful leverage in designing caching policies, because of differences in the potential for sharing these two document types among multiple users.

Characteristics of WWW Client-based Traces

The explosion of WWW traffic necessitates an accurate picture of WWW use, and in particular requires a good understanding of client requests for WWW documents. To address this need, we have collected traces of actual executions of NCSA Mosaic, reflecting over half a million user requests for WWW documents. In this paper we describe the methods we used to collect our traces, and the formats of the collected data. Next, we present a descriptive statistical summary of the traces we collected, which identifies a number of trends and reference patterns in WWW use. In particular, we show that many characteristics of WWW use can be modelled using power-law distributions, including the distribution of document sizes, the popularity of documents as a function of size, the distribution of user requests for documents, and the number of references to documents as a function of their overall rank in popularity (Zipf's law). Finally, we show how the power-law distributions derived from our traces can be used to guide system designers interested in caching WWW documents.

Explaining World Wide Web Traffic Self-Similarity

Recently the notion of self-similarity has been shown to apply to wide-area and local-area network traffic. In this paper we examine the mechanisms that give rise to self-similar network traffic. We present an explanation for traffic self-similarity by using a particular subset of wide area traffic: traffic due to the World Wide Web (WWW). Using an extensive set of traces of actual user executions of NCSA Mosaic, reflecting over half a million requests for WWW documents, we show evidence that WWW traffic is self-similar. Then we show that the self-similarity in such traffic can be explained based on the underlying distributions of WWW document sizes, the effects of caching and user preference in file transfer, the effect of user "think time", and the superimposition of many such transfers in a local area network. To do this we rely on empirically measured distributions both from our traces and from data independently collected at over thirty WWW sites.

Characterizing Reference Locality in the WWW

As the World Wide Web (Web) is increasingly adopted as the infrastructure for large-scale distributed information systems, issues of performance modeling become ever more critical. In particular, locality of reference is an important property in the performance modeling of distributed information systems. In the case of the Web, understanding the nature of reference locality will help improve the design of middleware, such as caching, prefetching, and document dissemination systems. For example, good measurements of reference locality would allow us to generate synthetic reference streams with accurate performance characteristics, would allow us to compare empirically measured streams to explain differences, and would allow us to predict expected performance for system design and capacity planning. In this paper we propose models for both temporal and spatial locality of reference in streams of requests arriving at Web servers. We show that simple models based only on document popularity (likelihood of reference) are insufficient for capturing either temporal or spatial locality. Instead, we rely on an equivalent, but numerical, representation of a reference stream: a stack distance trace. We show that temporal locality can be characterized by the marginal distribution of the stack distance trace, and we propose models for typical distributions and compare their cache performance to our traces. We also show that spatial locality in a reference stream can be characterized using the notion of self-similarity. Self-similarity describes long-range correlations in the dataset, which is a property that previous researchers have found hard to incorporate into synthetic reference strings. We show that stack distance strings appear to be strongly self-similar, and we provide measurements of the degree of self-similarity in our traces. Finally, we discuss methods for generating synthetic Web traces that exhibit the properties of temporal and spatial locality that we measured in our data.

Measuring the Behavior of a World-Wide Web Server

Server performance has become a crucial issue for improving the overall performance of the World-Wide Web. This paper describes Webmonitor, a tool for evaluating and understanding server performance, and presents new results for a realistic workload. Webmonitor measures activity and resource consumption, both within the kernel and in HTTP processes running in user space. Webmonitor is implemented using an efficient combination of sampling and event-driven techniques that exhibit low overhead. Our initial implementation is for the Apache World-Wide Web server running on the Linux operating system. We demonstrate the utility of Webmonitor by measuring and understanding the performance of a Pentium-based PC acting as a dedicated WWW server. Our workload uses a file size distribution with a heavy tail. This captures the fact that Web servers must concurrently handle some requests for large audio and video files, and a large number of requests for small documents, containing text or images. Our results show that in a Web server saturated by client requests, over 90% of the time spent handling HTTP requests is spent in the kernel. Furthermore, keeping TCP connections open, as required by TCP, causes a factor of 2-9 increase in the elapsed time required to service an HTTP request. Data gathered from Webmonitor provide insight into the causes of this performance penalty. Specifically, we observe a significant increase in resource consumption along three dimensions: the number of HTTP processes running at the same time, CPU utilization, and memory utilization. These results emphasize the important role of operating system and network protocol implementation in determining Web server performance.

The Network Effects of Prefetching

Prefetching has been shown to be an effective technique for reducing user perceived latency in distributed systems. In this paper we show that even when prefetching adds no extra traffic to the network, it can have serious negative performance effects. Straightforward approaches to prefetching increase the burstiness of individual sources, leading to increased average queue sizes in network switches. However, we also show that applications can avoid the undesirable queueing effects of prefetching. In fact, we show that applications employing prefetching can significantly improve network performance, to a level much better than that obtained without any prefetching at all. This is because prefetching offers increased opportunities for traffic shaping that are not available in the absence of prefetching. Using a simple transport rate control mechanism, a prefetching application can modify its behavior from a distinctly ON/OFF entity to one whose data transfer rate changes less abruptly, while still delivering all data in advance of the user's actual requests.

Load Profiling for Efficient Route Selection in Multi-Class Networks

High-speed networks, such as ATM networks, are expected to support diverse Quality of Service (QoS) constraints, including real-time QoS guarantees. Real-time QoS is required by many applications such as those that involve voice and video communication. To support such services, routing algorithms that allow applications to reserve the needed bandwidth over a Virtual Circuit (VC) have been proposed. Commonly, these bandwidth-reservation algorithms assign VCs to routes using the least-loaded concept, and thus result in balancing the load over the set of all candidate routes. In this paper, we show that for such reservation-based protocols|which allow for the exclusive use of a preset fraction of a resource's bandwidth for an extended period of time-load balancing is not desirable as it results in resource fragmentation, which adversely affects the likelihood of accepting new reservations. In particular, we show that load-balancing VC routing algorithms are not appropriate when the main objective of the routing protocol is to increase the probability of finding routes that satisfy incoming VC requests, as opposed to equalizing the bandwidth utilization along the various routes. We present an on-line VC routing scheme that is based on the concept of "load profiling", which allows a distribution of "available" bandwidth across a set of candidate routes to match the characteristics of incoming VC QoS requests. We show the effectiveness of our load-profiling approach when compared to traditional load-balancing and load-packing VC routing schemes.

Evaluation of a Load Profiling Approach to Routing Guaranteed Bandwidth Flows

To support the diverse Quality of Service (QoS) requirements of real-time (e.g. audio/video) applications in integrated services networks, several routing algorithms that allow for the reservation of the needed bandwidth over a Virtual Circuit (VC) established on one of several candidate routes have been proposed. Traditionally, such routing is done using the least-loaded concept, and thus results in balancing the load across the set of candidate routes. In a recent study, we have established the inadequacy of this load balancing practice and proposed the use of load profiling as an alternative. Load profiling techniques allow the distribution of "available" bandwidth across a set of candidate routes to match the characteristics of incoming VC QoS requests. In this paper we thoroughly characterize the performance of VC routing using load profiling and contrast it to routing using load balancing and load packing. We do so both analytically and via extensive simulations of multi-class traffic routing in Virtual Path (VP) based networks. Our findings confirm that for routing guaranteed bandwidth flows in VP networks, load balancing is not desirable as it results in VP bandwidth fragmentation, which adversely affects the likelihood of accepting new VC requests. This fragmentation is more pronounced when the granularity of VC requests is large. Typically, this occurs when a common VC is established to carry the aggregate traffic flow of many high-bandwidth real-time sources. For VP-based networks, our simulation results show that our load-profiling VC routing scheme performs better or as well as the traditional load-balancing VC routing in terms of revenue under both skewed and uniform workloads. Furthermore, load-profiling routing improves routing fairness by proactively increasing the chances of admitting high-bandwidth connections.

Admission Control and Scheduling for High-Performance WWW Servers

In this paper we examine a number of admission control and scheduling protocols for high-performance web servers based on a 2-phase policy for serving HTTP requests. The first "registration" phase involves establishing the TCP connection for the HTTP request and parsing/interpreting its arguments, whereas the second "service" phase involves the service/transmission of data in response to the HTTP request. By introducing a delay between these two phases, we show that the performance of a web server could be potentially improved through the adoption of a number of scheduling policies that optimize the utilization of various system components (e.g. memory cache and I/O). In addition, to its premise for improving the performance of a single web server, the delineation between the registration and service phases of an HTTP request may be useful for load balancing purposes on clusters of web servers. We are investigating the use of such a mechanism as part of the Commonwealth testbed being developed at Boston University.

Measuring Bottleneck Bandwidth of Targeted Path Segments

Accurate measurement of network bandwidth is crucial for flexible Internet applications and protocols which actively manage and dynamically adapt to changing utilization of network resources. These applications must do so to perform tasks such as distributing and delivering high-bandwidth media, scheduling service requests and performing admission control. Extensive work has focused on two approaches to measuring bandwidth: measuring it hop-by-hop, and measuring it end-to-end along a path. Unfortunately, best-practice techniques for the former are inefficient and techniques for the latter are only able to observe bottlenecks visible at end-to-end scope. In this paper, we develop and simulate end-to-end probing methods which can measure bottleneck bandwidth along arbitrary, targeted subpaths of a path in the network, including subpaths shared by a set of flows. As another important contribution, we describe a number of practical applications which we foresee as standing to benefit from solutions to this problem, especially in emerging, flexible network architectures such as overlay networks, ad-hoc networks, peer-to-peer architectures and massively accessed content servers.

Scalability of Multicast Delivery for Non-sequential Streaming Access

To serve asynchronous requests using multicast, two categories of techniques, stream merging and periodic broadcasting have been proposed. For sequential streaming access where requests are uninterrupted from the beginning to the end of an object, these techniques are highly scalable: the required server bandwidth for stream merging grows logarithmically as request arrival rate, and the required server bandwidth for periodic broadcasting varies logarithmically as the inverse of start-up delay. However, sequential access is inappropriate to model partial requests and client interactivity observed in various streaming access workloads. This paper analytically and experimentally studies the scalability of multicast delivery under a non-sequential access model where requests start at random points in the object. We show that the required server bandwidth for any protocols providing immediate service grows at least as the square root of request arrival rate, and the required server bandwidth for any protocols providing delayed service grows linearly with the inverse of start-up delay. We also investigate the impact of limited client receiving bandwidth on scalability. We optimize practical protocols which provide immediate service to non-sequential requests. The protocols utilize limited client receiving bandwidth, and they are near-optimal in that the required server bandwidth is very close to its lower bound.

PeriScope: An Active Internet Probing and Measurement API

Growing interest in inference and prediction of network characteristics is justified by its importance for a variety of network-aware applications. One widely adopted strategy to characterize network conditions relies on active, end-to-end probing of the network. Active end-to-end probing techniques differ in (1) the structural composition of the probes they use (e.g., number and size of packets, the destination of various packets, the protocols used, etc.), (2) the entity making the measurements (e.g. sender vs. receiver), and (3) the techniques used to combine measurements in order to infer specific metrics of interest. In this paper, we present Periscope: a Linux API that enables the definition of new probing structures and inference techniques from user space through a flexible interface. PeriScope requires no support from clients beyond the ability to respond to ICMP ECHO REQUESTs and is designed to minimize user/kernel crossings and to ensure various constraints (e.g., back-to-back packet transmissions, fine-grained timing measurements) We show how to use Periscope for two different probing purposes, namely the measurement of shared packet losses between pairs of endpoints and for the measurement of subpath bandwidth. Results from Internet experiments for both of these goals are also presented.

Differentiated Control of Web Traffic: A Numerical Analysis

Internet measurements show that the size distribution of Web-based transactions is usually very skewed; a few large requests constitute most of the total traffic. Motivated by the advantages of scheduling algorithms which favor short jobs, we propose to perform differentiated control over Web-based transactions to give preferential service to short web requests. The control is realized through service semantics provided by Internet Traffic Managers, a Diffserv-like architecture. To evaluate the performance of such a control system, it is necessary to have a fast but accurate analytical method. To this end, we model the Internet as a time-shared system and propose a numerical approach which utilizes Kleinrock's conservation law to solve the model. The numerical results are shown to match well those obtained by packet-level simulation, which runs orders of magnitude slower than our numerical method.

A Hierarchical Characterization of a Live Streaming Media Workload

We present what we believe to be the first thorough characterization of live streaming media content delivered over the Internet. Our characterization of over five million requests spanning a 28-day period is done at three increasingly granular levels, corresponding to clients, sessions, and transfers. Our findings support two important conclusions. First, we show that the nature of interactions between users and objects is fundamentally different for live versus stored objects. Access to stored objects is user driven, whereas access to live objects is object driven. This reversal of active/passive roles of users and objects leads to interesting dualities. For instance, our analysis underscores a Zipf-like profile for user interest in a given object, which is to be contrasted to the classic Zipf-like popularity of objects for a given user. Also, our analysis reveals that transfer lengths are highly variable and that this variability is due to the stickiness of clients to a particular live object, as opposed to structural (size) properties of objects. Second, based on observations we make, we conjecture that the particular characteristics of live media access workloads are likely to be highly dependent on the nature of the live content being accessed. In our study, this dependence is clear from the strong temporal correlations we observed in the traces, which we attribute to the synchronizing impact of live content on access characteristics. Based on our analyses, we present a model for live media workload generation that incorporates many of our findings, and which we implement in GISMO [19].