LSMAC and LSNAIT: Two Approaches for Cluster-based Scalable Web Servers

Server responsiveness and scalability are more important than ever in today’s client/server dominated network environments. Recently, researchers have begun to consider cluster-based computers using commodity hardware as an alternative to expensive specialized hardware for building scalable Web servers. In this paper, we present performance results comparing two cluster-based Web servers based on different server infrastructures: MAC-based dispatching (LSMAC) and IP-based dispatching (LSNAT). Both cluster-based server systems were implemented as application-space programs running on commodity hardware. We point out the advantages and disadvantages of both systems. We also identify when servers should be clustered and when clustering will not improve performance.

Scalable Web Server Clustering Technologies

The exponential growth of the Internet, coupled with the increasing popularity of dynamically generated content on the World Wide Web, has created the need for more and faster Web servers capable of serving the over 100 million Internet users. Server clustering has emerged as a promising technique to build scalable Web servers. In this article we examine the seminal work, early products, and a sample of contemporary commercial offerings in the field of transparent Web server clustering. We broadly classify transparent server clustering into three categories.

Load Balancing a Cluster of Web Servers using Distributed Packet Rewriting

In this paper, we propose and evaluate an implementation of a prototype scalable web server. The prototype consists of a load-balanced cluster of hosts that collectively accept and service TCP connections. The host IP addresses are advertised using the Round Robin DNS technique, allowing any host to receive requests from any client. Once a client attempts to establish a TCP connection with one of the hosts, a decision is made as to whether or not the connection should be redirected to a different host---namely, the host with the lowest number of established connections. We use the low-overhead Distributed Packet Rewriting (DPR) technique to redirect TCP connections. In our prototype, each host keeps information about connections in hash tables and linked lists. Every time a packet arrives, it is examined to see if it has to be redirected or not. Load information is maintained using periodic broadcasts amongst the cluster hosts.

Distributed Packet Rewriting and its Application to Scalable Server Architectures

To construct high performance Web servers, system builders are increasingly turning to distributed designs. An important challenge that arises in distributed Web servers is the need to direct incoming connections to individual hosts. Previous methods for connection routing have employed a centralized node which handles all incoming requests. In contrast, we propose a distributed approach, called Distributed Packet Rewriting (DPR), in which all hosts of the distributed system participate in connection routing. We argue that this approach promises better scalability and fault-tolerance than the centralized approach. We describe our implementation of four variants of DPR and compare their performance. We show that DPR provides performance comparable to centralized alternatives, measured in terms of throughput and delay under the SPECweb96 benchmark. Finally, we argue that DPR is particularly attractive both for small scale systems and for systems following the emerging trend toward increasingly intelligent I/O subsystems.

Connection Scheduling in Web Servers

Under high loads, a Web server may be servicing many hundreds of connections concurrently. In traditional Web servers, the question of the order in which concurrent connections are serviced has been left to the operating system. In this paper we ask whether servers might provide better service by using non-traditional service ordering. In particular, for the case when a Web server is serving static files, we examine the costs and benefits of a policy that gives preferential service to short connections. We start by assessing the scheduling behavior of a commonly used server (Apache running on Linux) with respect to connection size and show that it does not appear to provide preferential service to short connections. We then examine the potential performance improvements of a policy that does favor short connections (shortest-connection-first). We show that mean response time can be improved by factors of four or five under shortest-connection-first, as compared to an (Apache-like) size-independent policy. Finally we assess the costs of shortest-connection-first scheduling in terms of unfairness (i.e., the degree to which long connections suffer). We show that under shortest-connection-first scheduling, long connections pay very little penalty. This surprising result can be understood as a consequence of heavy-tailed Web server workloads, in which most connections are small, but most server load is due to the few large connections. We support this explanation using analysis.

A Generic Reference Software Architecture for Load Balancing over Mirrored Web Servers: NaSr Case Study

With the rapid expansion of the internet and the increasing demand on Web servers, many techniques were developed to overcome the servers' hardware performance limitation. Mirrored Web Servers is one of the techniques used where a number of servers carrying the same "mirrored" set of services are deployed. Client access requests are then distributed over the set of mirrored servers to even up the load. In this paper we present a generic reference software architecture for load balancing over mirrored web servers. The architecture was designed adopting the latest NaSr architectural style [1] and described using the ADLARS [2] architecture description language. With minimal effort, different tailored product architectures can be generated from the reference architecture to serve different network protocols and server operating systems. An example product system is described and a sample Java implementation is presented.

DiffServer: Application Level Differentiated Services for Web Servers

Web content hosting, in which a Web server stores and provides Web access to documents for different customers, is becoming increasingly common. For example, a web server can host webpages for several different companies and individuals. Traditionally, Web Service Providers (WSPs) provide all customers with the same level of performance (best-effort service). Most service differentiation has been in the pricing structure (individual vs. business rates) or the connectivity type (dial-up access vs. leased line, etc.). This report presents DiffServer, a program that implements two simple, server-side, application-level mechanisms (server-centric and client-centric) to provide different levels of web service. The results of the experiments show that there is not much overhead due to the addition of this additional layer of abstraction between the client and the Apache web server under light load conditions. Also, the average waiting time for high priority requests decreases significantly after they are assigned priorities as compared to a FIFO approach.

On-the-fly web content integrity check boosts users' confidence

Web servers are accessible by anyone who can access the Internet. Although this universal accessibility is attractive for all kinds of Web-based applications, Web servers are exposed to attackers who may want to alter their contents. Alterations range from humorous additions or changes, which are typically easy to spot, to more sinister tampering, such as providing false or damaging information.

Improving web server performance by network aware data buffering and caching

In this paper we propose a new method of data handling for web servers. We call this method Network Aware Buffering and Caching (NABC for short). NABC facilitates reduction of data copies in web server's data sending path, by doing three things: (1) Layout the data in main memory in a way that protocol processing can be done without data copies (2) Keep a unified cache of data in kernel and ensure safe access to it by various processes and kernel and (3) Pass only the necessary meta data between processes so that bulk data handling time spent during IPC can be reduced. We realize NABC by implementing a set of system calls and an user library. The end product of the implementation is a set of APIs specifically designed for use by the web servers. We port an in house web server called SWEET, to NABC APIs and evaluate performance using a range of workloads both simulated and real. The results show a very impressive gain of 12% to 21% in throughput for static file serving and 1.6 to 4 times gain in throughput for lightweight dynamic content serving for a server using NABC APIs over the one using UNIX APIs.

On Reducing Dynamic Web Page Construction Times

Many web sites incorporate dynamic web pages to deliver customized contents to their users. However, dynamic pages result in increased user response times due to their construction overheads. In this paper, we consider mechanisms for reducing these overheads by utilizing the excess capacity with which web servers are typically provisioned. Specifically, we present a caching technique that integrates fragment caching with anticipatory page pre-generation in order to deliver dynamic pages faster during normal operating situations. A feedback mechanism is used to tune the page pre-generation process to match the current system load. The experimental results from a detailed simulation study of our technique indicate that, given a fixed cache budget, page construction speedups of more than fifty percent can be consistently achieved as compared to a pure fragment caching approach.

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.

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.

The Open Web of Things as a means to unlock the potential of the IoT

Abstract: There is a lot of hype around the Internet of Things along with talk about 100 billion devices within 10 years time. The promise of innovative new services and efficiency savings is fueling interest in a wide range of potential applications across many sectors including smart homes, healthcare, smart grids, smart cities, retail, and smart industry. However, the current reality is one of fragmentation and data silos. W3C is seeking to fix that by exposing IoT platforms through the Web with shared semantics and data formats as the basis for interoperability. This talk will address the abstractions needed to move from a Web of pages to a Web of things, and introduce the work that is being done on standards and on open source projects for a new breed of Web servers on microcontrollers to cloud based server farms. Speaker Biography -Dave Raggett : Dave has been involved at the heart of web standards since 1992, and part of the W3C Team since 1995. As well as working on standards, he likes to dabble with software, and more recently with IoT hardware. He has participated in a wide range of European research projects on behalf of W3C/ERCIM. He currently focuses on Web payments, and realising the potential for the Web of Things as an evolution from the Web of pages. Dave has a doctorate from the University of Oxford. He is a visiting professor at the University of the West of England, and lives in the UK in a small town near to Bath.

A locality aware cache diffusion system

Web caching is a widely deployed technique to reduce the load to web servers and to reduce the latency for web browsers. Peer-to-Peer (P2P) web caching has been a hot research topic in recent years as it can create scalable and robust designs for decentralized internet-scale applications. However, many P2P web caching systems suffer expensive overheads such as lookup and publish messages, and lack locality awareness. In this paper, we present the development of a locality aware cache diffusion system that makes use of routing table locality, aggregation, and soft state to overcome these limitations. The analysis and experiments show that our cache diffusion system reduces the amount of information processed by nodes, reduces the number of index messages sent by nodes, and improves the locality of cache pointers.

Detection and defense of application-layer DDoS attacks in backbone web traffic

Web servers are usually located in a well-organized data center where these servers connect with the outside Internet directly through backbones. Meanwhile, the application-layer distributed denials of service (AL-DDoS) attacks are critical threats to the Internet, particularly to those business web servers. Currently, there are some methods designed to handle the AL-DDoS attacks, but most of them cannot be used in heavy backbones. In this paper, we propose a new method to detect AL-DDoS attacks. Our work distinguishes itself from previous methods by considering AL-DDoS attack detection in heavy backbone traffic. Besides, the detection of AL-DDoS attacks is easily misled by flash crowd traffic. In order to overcome this problem, our proposed method constructs a Real-time Frequency Vector (RFV) and real-timely characterizes the traffic as a set of models. By examining the entropy of AL-DDoS attacks and flash crowds, these models can be used to recognize the real AL-DDoS attacks. We integrate the above detection principles into a modularized defense architecture, which consists of a head-end sensor, a detection module and a traffic filter. With a swift AL-DDoS detection speed, the filter is capable of letting the legitimate requests through but the attack traffic is stopped. In the experiment, we adopt certain episodes of real traffic from Sina and Taobao to evaluate our AL-DDoS detection method and architecture. Compared with previous methods, the results show that our approach is very effective in defending AL-DDoS attacks at backbones. © 2013 Elsevier B.V. All rights reserved.