This paper considers the potentially negative impacts of an increasing deployment of non-congestion-controlled best-effort traffic on the Internet.’ These negative impacts range from extreme unfairness against competing TCP traffic to the potential for congestion collapse. To promote the inclusion of end-to-end congestion control in the design of future protocols using best-effort traffic, we argue that router mechanisms are needed to identify and restrict the bandwidth of selected high-bandwidth best-effort flows in times of congestion. The paper discusses several general approaches for identifying those flows suitable for bandwidth regulation. These approaches are to identify a high-bandwidth flow in times of congestion as unresponsive, “not TCP-friendly,” or simply using disproportionate bandwidth. A flow that is not “TCP-friendly ” is one whose long-term arrival rate exceeds that of any conformant TCP in the same circumstances. An unresponsive flow is one failing to reduce its offered load at a router in response to an increased packet drop rate, and a disproportionate-bandwidth flow is one that uses considerably more bandwidth than other flows in a time of congestion.

Abstract | Memory hierarchies have long been studied by many means: system building, trace-driven simulation, and mathematical analysis. Yet little help is available for the sys-tem designer wishing to quickly size the di erent levels in a memory hierarchy to a rst-order approximation. In this pa-per, we present a simple analysis for providing this practical help and some unexpected results and intuition that come out of the analysis. By applying a speci c, parameterized model of workload locality, we are able to derive a closed-form solution for the optimal size of each hierarchy level. We verify the accu-racy of this solution against exhaustive simulation with two case studies: a three-level I/O storage hierarchy and a three-level processor-cache hierarchy. In all but one case, the con gura-tion recommended by the model performs within 5 % of optimal. One result of our analysis is that the rst place tospend money is the cheapest (rather than the fastest) cache level, particularly with small system budgets. Another is that money spent on an n-level hierarchy is spent in a xed proportion until another level is added.

ION LAYER DEVICE LAYER USER SPACE KERNEL SPACE SOURCE DEVICE NETWORK ADAPTER CLIENT APPLICATION LAYER STANDARD I/O LAYER ABSTRACTION LAYER DEVICE LAYER USER SPACE KERNEL SPACE NETWORK ADAPTER SINK DEVICE ROUTER ROUTER NETWORK Figure 2 An End-to-End Data Path from a Source Device on One Workstation to a Sink Device on Another Workstation device is a network adapter. In this paper, we use the term data transfer problemto refer to the problem of reducing these overheads to achieve high throughput between a source device and a sink device, either of which can be a network adapter within a single workstation. Although the data transfer problem may also exist in intermediate routers, it does so to a much lesser degree than with end-user workstations (assuming modern router software and hardware technology). This is because of a router's simplified execution environment and its reduced needs for transfers across multiple protected domains. However, there is nothing that precludes the appli...