We introduce a new method and system to curb junk e-mail by employing extended e-mail addresses. It enables a party to use her (core) e-mail address with different extensions and consequently classify incoming e-mail messages according to the extension they were sent to. Our contributions are threefold: First, we identify the components of a system that realizes the concept of extended e-mail addresses and investigate the functionality of these components in a manner which is backwards compatible to current e-mail tools. Secondly, we specify an adversarial model, and give the necessary properties of extended e-mail addresses and of the procedure to obtain them in the presence of the adversary. Finally, we design cryptographic functions that enable realizing extended e-mail addresses which satisfy these properties.

An increasing number of web-sites require users to establish an account before they can access the information stored on that site ("personalized web browsing"). Typically, the user is required to provide at least a unique username, a secret password and an e-mail address. Establishing accounts at multiple web-sites is a tedious task. A security-and privacy-aware user may have to invent a distinct username and a secure password, both unrelated to his/her identity, for each web-site. The user may also desire mechanisms for anonymous e-mail. Besides the information that the user supplies voluntarily to the web-site, additional information about the user may flow (involuntarily) from the user's site to the web-site, due to the nature of the HTTP protocol and the cookie mechanism. This paper describes the Janus Personalized Web Anonymizer, which makes personalized web browsing simple, secure and anonymous by providing convenient solutions to each of the above problems. Janus serves as an intermediary entity between a user and a web-site. Given a user and a web-site, Janus automatically generates an alias - typically a username, a password and an e-mail address - that can be used to establish an anonymous account at the web-site. Different aliases are generated for each user, web-site pair; however the same alias is presented whenever a particular user visits a particular web-site. Janus frees the user from the burden of inventing and memorizing distinct usernames and secure passwords for each web-site, and guarantees that an alias (including an e-mail address) does not reveal the true identity of the user. Janus also provides mechanisms to complete an anonymous e-mail exchange from a web-site to a user, and filters the information-flow of the HTTP protocol to preserve user privacy. Thus Janus provides simultaneous user identification and user privacy, as required for anonymous personalized web browsing.

Emerging applications in electronic commerce of ten involve very low cost transactions which ex ecute in the context of ongoing extended client server relationships For example consider a web site server which o ers repeated authenticated personalized stock quotes to each of its subscribers clients The value of a single transaction e g de livery of a web page with a customized set of quotes does not warrant the cost of executing a handshake and key distribution protocol Also a client might not always use the same machine during such an extended relationship e g a PC at home a laptop on a trip Typical transport session layer security mechanisms such as SSL and S HTTP either require handshake key distribution for each transaction or do not support client mobility We propose a new security framework for extended relationships between clients and servers based on persistent shared keys We argue that this is a pre ferred model for inexpensive transactions executing within extended relationships Our main contri bution is the design and implementation of a set of lightweight application layer primitives for generating and maintaining persistent shared keys without requiring a client to store any informa tion between transactions and securing a wide range of web transactions e g subscription au thenticated and or private delivery of information receipts with adequate computational cost Our protocols require public key infrastructure only for servers vendors and its usage only once per client upon rst interaction

Extensive research on system support for enabling I/O-intensive applications to achieve performance close to the limits imposed by the hardware suggests two main approaches: Low overhead I/O protocols and the flexibility to customize I/O policies to the needs of applications. One way to achieve both is by supporting user-level access to I/O devices, enabling user-level implementations of I/O protocols. User-level networking is an example of this approach, specific to network interface controllers (NICs). In this paper, we argue that the real key to high-performance in I/O-intensive applications is user-level file caching and user-level network buffering, both of which can be achieved without user-level access to NICs. Avoiding the need to support user-level networking carries two important benefits for overall system design: First, a NIC exporting a privileged kernel interface is simpler to design and implement than one exporting a user-level interface. Second, the kernel is re-instated as a global system resource controller and arbitrator. We develop an analytical model of network storage applications and use it to show that their performance is not affected by the use of a kernel-based API to NICs.

The Direct Access File System (DAFS) is a distributed file system built on top of direct-access transports (DAT). Direct-access transports are characterized by using remote direct memory access (RDMA) for data transfer and user-level networking. The motivation behind the DAT-enabled distributed file system architecture is the reduction of the CPU overhead on the I/O data path.We have created an implementation of DAFS for the FreeBSD platform. In this paper we describe the performance evaluation study of DAFS that we have performed using this software. The goal of this study is to determine whether the architecture of DAFS brings any fundamental performance benefits to applications compared to traditional distributed file systems, such as NFS. We perform comparison of DAFS to a version of NFS optimized to reduce the I/O overhead. In order to thoroughly understand the impact of DAFS on application performance, we consider a diverse range of applications workloads.We conclude that DAFS can accomplish superior performance for latency-sensitive applications, outperforming NFS by up to a factor of 2. Bandwidth-sensitive applications do equally well on both systems, unless they are CPU-intensive, in which case they perform better on DAFS. We also found that RDMA is a less restrictive mechanism to achieve copy avoidance than that used by the optimized NFS.

In this paper we present StarFish, a highly-available geographically-dispersed block storage system built from commodity servers running FreeBSD, which are connected by standard high-speed IP networking gear. StarFish achieves high availability by transparently replicating data over multiple storage sites. StarFish is accessed via a host-site appliance that masquerades as a host-attached storage device, hence it requires no special hardware or software in the host computer. We show that a StarFish system with 3 replicas and a write quorum size of 2 is a good choice, based on a formal analysis of data availability and reliability: 3 replicas with individual availability of 99%, a write quorum of 2, and read-only consistency gives better than 99.9999% data availability. Although StarFish increases the per-request latency relative to a direct-attached RAID, we show how to design a highly-available StarFish configuration that provides most of the performance of a direct-attached RAID on an I/O-intensive benchmark, even during the recovery of a failed replica. Moreover, the third replica may be connected by a link with long delays and limited bandwidth, which alleviates the necessity of dedicated communication links to all replicas.

The viability of storage outsourcing is critically dependent on the access performance of remote storage. We study this issue by measuring the behavior of a broad variety of I/O-intensive benchmarks as they access remote storage over an IP network. We measure the effect of network latencies that correspond to distances ranging from a local neighborhood to halfway across a continent. We then measure the effect of latency-hiding mechanisms. Our results indicate that, in many cases, the adverse effects of network delay can be rendered inconsequential by clever file system and operating system techniques.

Today, caching web proxies use general-purpose file systems to store web objects. Proxies, e.g., Squid or Apache, when running on a UNIX system, typically use the standard UNIX file system (UFS) for this purpose. UFS was designed for research and engineering environments, which have different characteristics from that of a caching web proxy. Some of the differences are high temporal locality, relaxed persistence requirements, and a different read/write ratio. In this paper, we characterize the web proxy workload, describe the design of Hummingbird, a light-weight file system for web proxies, and present performance measurements of Hummingbird. Hummingbird has two distinguishing features: it separates object naming and storage locality through direct application-provided hints, and its clients are compiled with a linked library interface for memory sharing. When we simulated the Squid proxy, Hummingbird achieves document request throughput 2.3-9.4 times larger than with several different versions of UFS. Our experimental results are verified within the Polygraph proxy benchmarking environment.

Malloc(3) is considered to be a robust building block. However, we found that many malloc implementations suffer from excessive heap fragmentation when used with Hummingbird, a long-running application which stores a large number of fixed-sized and variable-sized objects in dynamic memory. This paper characterizes the dynamic memory activity pattern of Hummingbird and GNU Emacs. It compares the behavior of nine different mallocs when used with Hummingbird and GNU Emacs dynamic memory activity traces. In the Hummingbird case, the best malloc caused 30.5% fragmentation (increased heap size above the amount of live memory), while the worst malloc caused a heap overflow. In the GNU Emacs case, the best malloc caused 2.69% fragmentation, and the worst one caused 101.5% fragmentation.

This work proposes a hybrid solution, called the smart box architecture (SBoX), that provides quality of service (QoS) in internet protocol (IP)-based networks. SBoX architecture consists of SBoX servers, which are located at the network boundary, and SBoX routers, which are add-on label switching routers (LSR), which are located at interior network nodes. This approach combines the advantages of three existing technologies: integrated services (Intserv), differentiated services (Diffserv), and multi-protocol label switching (MPLS). SBoX aggregates traffic in three levels: commodity-flows, macro-flows and micro-flows. Commodity-flows aggregate flow between every pair of edge points. The packets of the same commodity-flow are marked by an MPLS label. Commodity-flows are composed of macro-flows, which aggregate the traffic of a particular enterprise. Macro-flows are associated with an explicit service level agreement (SLA), which is offered to users. Macro-flows are composed of micro-flows, which are the traffic associated with a particular individual in an enterprise or a particular application. SBoX servers provide Diffserv like SLA to users, and use class-based queuing (CBQ) with a hierarchy of flow aggregation. SBoX servers manage macro-flows and commodity-flows only, and leave the management of micro-flows to the enterprise/users which signed the SLA for the macro-flow. SBoX routers perform MPLS routing of commodity-flows in interior network nodes. SBoX routers can be combined with existing best-effort routers to provide QoS in a network that lacks end-to-end deployment of LSRs. This paper describes the SBoX architecture and its operation in detail. It also reports experimental results obtained on a prototype network. The results indicate that the SBoX architecture can indeed provide guaranteed performance in a congested network, and that SBoX routers can be combined with commodity best-effort routers to enable QoS in a heterogeneous network.