Abstract-This paper presents a simple method for computing throughputs of links in a CSMA network. We call our method back-of-the-envelop (BoE) computation, because for many network configurations, very accurate results can be obtained by simple hand computation. BoE beats prior methods in terms of both speed and accuracy. To explain BoE, we construct a theory based on the model of an "ideal CSMA network" (ICN). We find that link throughputs are insensitive to the distributions of the backoff countdown time and transmission time in ICN given the ratio of their mean c. The BoE computation method emerges from ICN in the limit 0 → c . The insensitivity result explains why BoE works so well for IEEE 802.11 networks, in which the backoff countdown process is one that has memory and the transmission time can be arbitrarily distributed. Furthermore, c does not have to be very small for BoE to be highly accurate. BoE allows us to make shortcuts in performance evaluation, bypassing complicated stochastic analysis. An immediate application of BoE is for quick identification of starved links in the network so that remedies can be devised to solve the problem.

Large-scale Internet applications, such as content distribution networks, are deployed across multiple datacenters and consume massive amounts of electricity. To provide uniformly low access latencies, these datacenters are geographically distributed and the deployment size at each location reflects the regional demand for the application. Consequently, an application's environmental impact can vary significantly depending on the geographical distribution of end-users, as electricity cost and carbon footprint per watt is location specific. In this paper, we describe FORTE: Flow Optimization based framework for request-Routing and Traffic Engineering. FORTE dynamically controls the fraction of user traffic directed to each datacenter in response to changes in both request workload and carbon footprint. It allows an operator to navigate the threeway tradeoff between access latency, carbon footprint, and electricity costs and to determine an optimal datacenter upgrade plan in response to increases in traffic load. We use FORTE to show that carbon taxes or credits are impractical in incentivizing carbon output reduction by providers of large-scale Internet applications. However, they can reduce carbon emissions by 10% without increasing the mean latency nor the electricity bill.

In this paper, we propose an unorthodox topology for datacenters that eliminates all hierarchical switches in favor of connecting nodes at random according to a small-worldinspired distribution. Specifically, we examine topologies where the underlying nodes are connected at the small scale in a regular pattern, such as a ring, torus or cube, such that every node can route efficiently to nodes in its immediate vicinity, and amended by the addition of random links to nodes throughout the datacenter, such that a greedy algorithm can route packets to far away locations efficiently. Coupled with geographical address assignment, the resulting network can provide content routing in addition to traditional routing, and thus efficiently implement key-value stores. The irregular but self-similar nature of the network facilitates constructing large networks easily using prewired, commodity racks. We show that Small-World Datacenters can achieve higher bandwidth and fault tolerance compared to both conventional hierarchical datacenters as well as the recently proposed CamCube topology. Coupled with hardware acceleration for packet switching, small-world datacenters can achieve an order of magnitude higher bandwidth than a conventional datacenter, depending on the network traffic.

No abstract

The cloud was originally designed to provide general purpose computing using commodity hardware and its focus was on increasing resource consolidation as a means to lower cost. Hence, it was not particularly adapted to the requirements of multimedia applications that are highly latency sensitive and require specialized hardware, such as graphical processing units.Existing cloud infrastructure is dimensioned to serve general purpose workloads and to meet end-user requirements by providing high throughput. In this paper, we investigate the effectiveness of using this general-purpose infrastructure for serving latency-sensitive multimedia applications. In particular, we examine on-demand gaming, also known as cloud gaming, which has the potential to change the video game industry. We demonstrate through a large-scale measurement study that the existing cloud infrastructure is unable to meet the strict latency requirements that is necessary for acceptable on-demand game play. Furthermore, we investigate the effectiveness of incorporating edge servers, which are servers located near end-users (e.g. CDN servers), to improve end-user coverage. Specifically, we examine an edge-server only infrastructure and a hybrid infrastructure that consists of using edge servers in addition to the cloud. We find that a hybrid infrastructure significantly improves the number of end-users served. However, the number of satisfied end-users in a hybrid deployment largely depends on the various deployment parameters. Therefore, we evaluate various strategies that determine two such parameters, namely, the location of on-demand gaming servers and the games that are placed on these servers. We find that, through both a careful selection of on-demand gaming servers and the games to place on these servers, we significantly increase the number of end-users served over the basic random selection and placement strategies.