An analysis of Web Server performance with an accelerator

Size: px
Start display at page:

Download "An analysis of Web Server performance with an accelerator"

Transcription

1 An analysis of Web Server performance with an accelerator S.Jha, F. Sabrina, B. Samuels, and M. Hassan School of Computer Science and Engineering The University of New South Wales Sydney, NSW 2052 Australia {sjha, farizas, ABSTRACT One of the most vexing questions facing researchers interest in the World Wide Web is how the performance of Web Server could be improved. In this paper we discuss the results of our investigation on potential of httpd accelerator (phhttpd) to improve web server performance. This paper presents comparative results for running the Web server with and without an accelerator, and measures the effectiveness of the accelerator. As there is little quantitative study of httpd accelerator available, this paper provides a good insight into the potential of the httpd accelerator for improving web server performance by performing empirical study. 1.INTRODUCTION With the explosive growth in the use of World Wide Web, improving the performance of the web has been the subject of much recent research. Server performance has become a critical issue for improving the Quality of Service (QoS) on the World Wide Web. A good understanding of Web Server s behavior and performance issues are necessary to improve the performance of Web server. The performance of Web servers on traditional Unix systems is limited by the underlying operating system s forking process model. This model is very resource intensive, and Web server accelerators or httpd accelerators came as a result to improve performance [2]. On Unix systems when a request is received, it must switch into kernel mode to get access to the hardware. The buffer from the device driver in the kernel needs to be copied to the user process. Efficient TCP/IP stack implementations, buffer copy is done either once through the OS or none at all. Less efficient systems may perform buffer copy through several layers to the user process. On top of this, the process needs some CPU time to execute the request. Recently various httpd accelerators have been used with the web servers to improve performance. The accelerator acts as a front-end for Web servers, initially handling all incoming requests. Requests that cannot be handled by accelerators are forwarded a backing Web server. When a request is received, the accelerator examines its cache for the file, returning the file immediately if found. If there is a cache miss, it can do a system call to fetch the file or pass the request onto a backing Web server to service. A benefit of this model is, that by having some agent examine requests, intelligent decisions can be made on how to best service these requests. The httpd accelerator runs in the kernel as a loadable module. This is very attractive since those requests can be serviced in kernel mode without having to switch to another process. Provos et al [3,5] describe that POSIX Real signal can improve the scalability of server. POSIX Real-time Signals enable network server applications to respond immediately to network requests. An added benefit of RT signals is that they can be queued in the kernel and delivered to an application one at a time, in order, leaving an application free to pick up I/O completion events when convenient. Phhttpd [3,5,7], an httpd accelerator that uses the POSIX RT signal mechanism, thus can improve the scalability of web server. The prior work on performance of phhttpd [3] was only a brief analysis. This paper presents a comprehensive quantitative measurement of performance of phhttpd. Section 2 describes the experimental setup. Measurement and analysis of experiments are discussed in section 3. Conclusion and future work is described in section 4.

2 A Architecture 2. EXPERIMENTAL SETUP The set up of experiments follows the framework outlined in the WAWM project [1]. In Figure 1, the components at the left are located in local network. The Web server (OZ) was the only machine servicing HTTP requests and the two clients were generating HTTP requests. The experiments are controlled from another host, acting as the Webmaster, which also acts as a data store for collection and analysis of data. All hosts in the local network connect via a Mbps switch. Active measurements are taken from the Webmaster, as well as from the remote hosts. Active measurements inject a packet into the network and the response to these packets is measured. Examples include ping and traceroute. Due to resource constraints, there is limited access to remote hosts; so two sites shown on the right in Figure 1 have been used. These sites offer public access to measurement tools running from their site. The Web server performs passive measurements (which gather relatively large amounts of data and record data at a node or endpoint in the network.) of the network using tcpdumpv3.4 and operating system monitoring through vmstat utility. When conducting experiments, simultaneous measurements are performed over different parts of the network. Accurate measurement of end-to-end delay requires synchronized clocks on the client and server. The difficulties of synchronizing clocks in a widely distributed environment led to the development of network time protocol (NTP) daemon, which assures that clocks are synchronized on the order of tens of milliseconds. When comparing measurements all host clocks are set to be synchronized every 30 minutes with a nearby NTP server, ntp.cse.unsw.edu.au by using ntpdate program. Linux also comes with the vmstat utility. This utility monitors operating system activity such as user time, system time, free memory, context switches, interrupts and several more, though the /proc file system. Reading from /proc can incur high overheads as several system calls might be used, and large blocks of data may be copied from the kernel to user space. For our purposes we use vmstat as it is readily available and simple to use. To limit the overheads associated with /proc reads, we only sample at periodic intervals of approximately 10 seconds. We create a shell script on the server to run vmstat at periodic intervals and log the data to a predetermined location for later retrieval by the Webmaster. We also gather process statistics on Apache by referring to the URL, on the Web server. A script is created to log this page at periodic intervals. Figure 1: Logical architecture of experiments B Server Configuration The Web Server runs RedHat Linux v7.0 (Guinness), using Kernel v on a 800 MHz Pentium III CPU and 256 MB of RAM PC with two Mbs Ethernet ports. Light performance tuning is done on the Linux system by removing unnecessary services, and tweaking system parameters. A modified version [11] of Apache v is used as Web server software that supports tunneling, allowing it to talk to an httpd accelerator if needed. Notably we set the pool of initial processes waiting for connections to a moderate size of 20. This value should not be too small in that, many extra httpd processes need to be added to the existing pool to cope with demand, nor should it be too large so that it consumes a lot of memory. A busy httpd process can consume in the order of a few megabytes of memory. For taking passive measurements, we install tcpdump v3.4. A shell script is created to execute tcpdump on both Ethernet ports, logging packet information to disk. C Client Configuration Both client machines have the same hardware specifications with Linux RedHat? v7.0 installed on Pluto and Linux RedHat? v6.2 on Saturn. The PCs were configured with 800 MHz Pentium III CPU and 128 MB of RAM. The clients do not need any other software, besides the workload generator program. For WebStone the binary file webclient, generates load and is distributed at run time. It requires our clients to have rexec and rsh daemons running in order to allow communication with the Webmaster. D Workload generator Workload generators are the tools used to generate HTTP requests on the Web server. There are essentially two types of generators, synthetic workload generators and trace driven generators. Synthetic workload generators are all based on making repeated requests as fast as possible or some at predetermined rate. In our

3 experiments we use the WebBench [6] and WebStone [4] benchmarking tools. E Webmaster configuration The Web clients are controlled by a single Webmaster, which combines the performance results from all clients into a single summary report. The Webmaster (Neptune) computer runs Linux Redhat v6.2. The main WebStone program is installed on this computer and parameters for the experiments can be configured. It was also set up to take ping measurements during experiments and fetch status information from the Web server periodically. These activities are automated by the use of shell scripts, with little user intervention required. For WebBench experiments the Webmaster also has Window? 98 installed. The controller program is installed on this machine. F Phhttpd Phhttpd [7] is an httpd accelerator used in our experiments and can run on the same system where the Web server resides. It features a small I/O core and an aggressive content cache. Currently it is limited to servicing static file requests, passing slower dynamic requests to a backing Web server. Networking is done by using non-blocking system calls, and allows a single execution context to handle as many clients, without having to do a process switch. G Benchmarking procedure The benchmarking process is divided in three phases. Phase 1: When the benchmark begins, the Webmaster distributes the load-generating program to the web clients, along with parameters of the experiment. The web clients will immediately begin generating requests. Scripts that will be executing measurements will be in the sleep mode. Phase 2: The system is well into the benchmark and approaches a steady state. Measurements on the system begin. Phase 3: At the end of the benchmark, the web clients transfer their individual statistics to the Webmaster. The Webmaster provides a summary report. It copies logs from the web server and remote hosts to a central location, for ease of analysis. 3. RESULT In this section we report on the results obtained using the experimental set up described in Section 4. In addition to the information described, we also add one host godzilla.zeta.org.au, located 10 hops away from the Web server, to measure the latency in fetching a file from the server. The duration of these experiments run for four minutes. We allow one minute for the system to start and move into a steady state, and then begin recording measurements. For analysis, the data obtained, firstly needs to be reduced. Logs taken from tcpdump are very large and we use the modified tcptrace [8] to gather information from the packet traces. WebStone already generates its own set of statistics and we use those in our analysis. We summarize the results using a standard spreadsheet for data gathered from active measurements and vmstat. A Performance under varying Loads In our initial set of experiments, we measure the performance of the Web Server under various client loads while retrieving a KB file. Connection statistics, latencies, OS statistics under various server loads are shown in Table 1-3. In Table 1 we observe that increasing client numbers does neither significantly increase the connection rate nor throughput of the system. This may be to do with the fact that at even 5 client processes, WebStone can still generate traffic near the maximum potential of the system. There are four different page mixes that attempt to model real workloads. Type of pages is defined by file size and access frequency. Each page in the mix has a weigh that indicates its probability of being accessed. The total amount of data moved is the product of the total number of pages retrieved and the page sizes. Little s Load Factor (LLF), derived from Little s law [10] reflects the number of clients actually being serviced at a time. The value calculated should be close to the number of clients, otherwise it indicates that the server is overloaded, and some request are not being serviced before they time out. Table 2 shows that increasing loads affect the latency in direct proportion. Connect latency is the time to establish a full TCP connection and response latency is time to transfer file after connection. As there are more clients added to the network there is more contention, hence longer delays. Table 3 shows us that increasing clients causes an increase in the number of processes waiting for execution time. This is because a separate process handles each client and only one process can ever execute at a time on our Linux system. It also shows that for a high load, in this situation, almost 80% of time is spent in the kernel, hence the lower number of context switches for clients. It is evident that on a Unix system, the HTTP servicing process requires a significant time doing system calls, when servicing a high number of concurrent users. Added with the fact that there are also many httpd processes waiting for execution time, we conclude that the forking process model is inefficient under higher loads. In the following section we experiment with a Web server accelerator,

4 phhttpd and look at the performance of using this system, which avoids using the forking model. Table 1:Connection statistics under several loads File Size (KB) Clients Pages Transferre d Connection rate (conn/sec) Server Throughput (Mbits/sec) LLF Total errors Table 2: Latency statistics under several loads File Size (KB) Clients Connect Latency (ms) Response Latency (ms) Local Ping (ms) Remote ping (ms) Remote Trace (ms) Fetch (s) Table 3: OS statistics under several Loads File size (KB) Clien ts Process Waiting (pw/sec) Interrupts (int/sec) Context Switches (cs/sec) %User %Syste m % Idle B Web Server Accelerator Performance In our next set of experiments we look at the performance aspects of phhttpd web accelerators. In figure 2 we see marginal gains in throughput and almost an increase of connections per second using phhttpd. Operating system statistics in Figure 3 shows that there is an 80% decrease in the number of context switches performed per second when Apache is used in conjunction with phhttpd, because httpd accelerator serves requests from the kernel. This increase in performance, allows more requests to come through, hence the increase in interrupts, from the network cards. Looking at how the CPU time is divided [Figure 4], phhttpd breezes through the benchmark, leaving the CPU idle 57% of the time. User time is very minimal at 8% while system time is 34%. The Web accelerator looks very promising, providing increased performance along with efficient utilization of resources. Throughput (Mbit) Connection rate (per sec) Figure 2 Throughput and connection rate Apache Standalone Apache+phhttpd

5 Figure 5: Connections per sec Figure 3 Context switches and interrupts Interrupts/s Context switches/s Apache standalone Apache+phhttpd We use another file set and measure the increase in performance along various number of clients. We run the experiments for four minutes, each time varying the number of clients. Results from the experiments for Apache and Apache+phhttpd are shown in Figures 5 and 6. We notice substantial gains in the connection rate and server throughput when accelerator is used. There is 54% increase in the number of connections and bandwidth usage at clients. This shows a significant improvement in server performance. We also have a look at the operating system statistics under this experiment. From Table 3, we see that Apache with phhttpd only has 1 process waiting on average, that results in a very large drop in the number of context switches done per second. This allows the Web server system to spend more time in the kernel (79.38%) and concentrate on the task of responding to network interrupts and sending out files, which all require system calls. Figure 4: CPU Figure 6. Interrupts /sec On the other hand, the Apache standalone model has 115 processes waiting and a 1600% increase in the number of context switches executed. The fact that there is zero idle time and processes waiting, indicating there is contention for CPU time between processes and system calls. Kernel routines have a higher priority and therefore less useful work from processes can be done, resulting in a degrade in performance. We see that a Web server accelerator like phhttpd, is very attractive and has shown significant performance gains and shows how an alternative model can be an advantage under a Unix based system. Table 4: Apache standalone vs Apache+phhttpd Processes Server Clients waiting (pw/sec) Interrupts (int/sec) Context Switches (cs/sec) % User % System %Idle Apache Apache+ phhttpd

6 4. CONCLUSION In this paper the performance gain from use of httpd accelerator with Web server has been studied. We have shown that under Unix, the forking model is very resource intensive. Experimental results in this paper show that under moderate to high loads, the CPU is being used to its full capacity, creating contention for CPU time amongst processes and system calls. This is caused by the fact that the httpd processes always need to switch in and out. A context switch to kernel mode is required to read from a network device driver. After parsing the request in user space, a system call is yet again required to send the request file onto the network, and yet another context switch is required to service other waiting httpd processes. Experiments were performed with an alternative model using a Web server accelerator, phhttpd. Results show substantial performance gain higher throughput and service rates and low resource compared to Apache. Accelerators run in the kernel and save the overheads of switching processes, although there is a trade off between execution speed and functionality. POSIX Real signals are very efficient mechanism in terms of the overhead and it also provides good throughput, but there are some limitations too. These limitations arise from the fact that the RT signal queue is a limited resource. Since each event results in a signal being appended to the signal queue, a few active connections could dominate the signal queue usage or even trigger an overflow. [12] has proposed a scheme called signal-per-fd which is an enhancement to the default RT signal implementation in the Linux kernel, and it could significantly reduce the complexity of a server implementation, increasing the robustness under high load, and potentially increasing its throughput The phhttpd front-end is very simple, and is limited to static requests for the time being. As is evident from our experiments, it is still worth using with Apache Web servers to serve static files such as images. On a general note, there is a potential for Web server accelerators to extend their functionality as the cost would seem minimal with its major performance gains over a mainstream Web server such as Apache. Smith et al [9] have done work to evaluate techniques for caching dynamic content, but still lots of work need to be done. Future work involves study of the dynamic contents issues and the work is under progress. REFERENCES [1] P. Barford, M. Crovella. Measuring Web Performance in the Wide Area, in Performance Evaluation Review, August, [2] Eric Levy-Abegnoli, Arun Iyengar, Junehwa Song, and Daniel Dias. Design and performance of a Web server accelerator. In Proceedings of IEEE INFOCOM 99, March 1999 [3] Neil Provos, chuck lever, stphen Tweedie, "Analyzing the overload behavior of a simple Web server." CITI Technical report 00-7, August 2000 [4] WebStone. [5] Neil provos, Chuck lever, "Scalable Network I/O in Linux", Proceedings of the Usenix technical conference, FREENIX track, June 2000 [6] WebBench 3.0/ [7] Z.Brown, phhttpd (documentation on phhttpd) [8] Tcptrace. [9] Ben Smith, Anurag Acharya, Tao Yang, Exploting Result Equivalence in Caching Dynamic web content. Department of Computer Science, University of California, Santa Barbara, CA [10] D.A. Menasce, V A F Almeida, capacity planning and performance modeling: From Mainframes to client server systems. Upper Saddle River, NJ: Prentice Hall, 1994 [11] Azer Bestravros, Bob Carter, Mark Crovella, Carlos Cunha, Abdelsalam Heddaya, and suliman Mirdad. Application-Level document caching in the Internet. In Proceedings of SDNE 95: The Second International Workshop on services in Distributed and Network Environment, June [12] Abhishek chandra, David Mosberger, Internet and mobile systems laboratory, HP laboratories Palo Alto, HPL ,"Scalability of linux Event-dispatch Mechanisms". December 14 th, Acknowledgment: Experimental work described in this paper was performed by Bryan Samuels for his minor thesis.

A Scalable Event Dispatching Library for Linux Network Servers

A Scalable Event Dispatching Library for Linux Network Servers A Scalable Event Dispatching Library for Linux Network Servers Hao-Ran Liu and Tien-Fu Chen Dept. of CSIE National Chung Cheng University Traditional server: Multiple Process (MP) server A dedicated process

More information

SEDA: An Architecture for Well-Conditioned, Scalable Internet Services

SEDA: An Architecture for Well-Conditioned, Scalable Internet Services SEDA: An Architecture for Well-Conditioned, Scalable Internet Services Matt Welsh, David Culler, and Eric Brewer Computer Science Division University of California, Berkeley Operating Systems Principles

More information

Introduction. Application Performance in the QLinux Multimedia Operating System. Solution: QLinux. Introduction. Outline. QLinux Design Principles

Introduction. Application Performance in the QLinux Multimedia Operating System. Solution: QLinux. Introduction. Outline. QLinux Design Principles Application Performance in the QLinux Multimedia Operating System Sundaram, A. Chandra, P. Goyal, P. Shenoy, J. Sahni and H. Vin Umass Amherst, U of Texas Austin ACM Multimedia, 2000 Introduction General

More information

Flash: an efficient and portable web server

Flash: an efficient and portable web server Flash: an efficient and portable web server High Level Ideas Server performance has several dimensions Lots of different choices on how to express and effect concurrency in a program Paper argues that

More information

On the Interaction Between an Operating System and Web Server

On the Interaction Between an Operating System and Web Server On the Interaction Between an Operating System and Web Server David J. Yates (djy@cs.bu.edu) Computer Science Department Boston University Boston, MA 2215 USA Virgílio Almeida (virgilio@dcc.ufmg.br) Jussara

More information

Performance Modeling and Evaluation of Web Systems with Proxy Caching

Performance Modeling and Evaluation of Web Systems with Proxy Caching Performance Modeling and Evaluation of Web Systems with Proxy Caching Yasuyuki FUJITA, Masayuki MURATA and Hideo MIYAHARA a a Department of Infomatics and Mathematical Science Graduate School of Engineering

More information

IX: A Protected Dataplane Operating System for High Throughput and Low Latency

IX: A Protected Dataplane Operating System for High Throughput and Low Latency IX: A Protected Dataplane Operating System for High Throughput and Low Latency Belay, A. et al. Proc. of the 11th USENIX Symp. on OSDI, pp. 49-65, 2014. Reviewed by Chun-Yu and Xinghao Li Summary In this

More information

Ch. 13: Measuring Performance

Ch. 13: Measuring Performance Ch. 13: Measuring Performance Kenneth Mitchell School of Computing & Engineering, University of Missouri-Kansas City, Kansas City, MO 64110 Kenneth Mitchell, CS & EE dept., SCE, UMKC p. 1/3 Introduction

More information

Memory Management Outline. Operating Systems. Motivation. Paging Implementation. Accessing Invalid Pages. Performance of Demand Paging

Memory Management Outline. Operating Systems. Motivation. Paging Implementation. Accessing Invalid Pages. Performance of Demand Paging Memory Management Outline Operating Systems Processes (done) Memory Management Basic (done) Paging (done) Virtual memory Virtual Memory (Chapter.) Motivation Logical address space larger than physical

More information

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

LSMAC and LSNAT: Two Approaches for Cluster-based Scalable Web Servers 1 LSMAC and LSNAT: Two Approaches for Cluster-based Scalable Web Servers Xuehong Gan a,1, Trevor Schroeder b, Steve Goddard b, and Byrav Ramamurthy b a Microsoft Corporation b Department of Computer Science

More information

Ch. 7: Benchmarks and Performance Tests

Ch. 7: Benchmarks and Performance Tests Ch. 7: Benchmarks and Performance Tests Kenneth Mitchell School of Computing & Engineering, University of Missouri-Kansas City, Kansas City, MO 64110 Kenneth Mitchell, CS & EE dept., SCE, UMKC p. 1/3 Introduction

More information

Profiling Grid Data Transfer Protocols and Servers. George Kola, Tevfik Kosar and Miron Livny University of Wisconsin-Madison USA

Profiling Grid Data Transfer Protocols and Servers. George Kola, Tevfik Kosar and Miron Livny University of Wisconsin-Madison USA Profiling Grid Data Transfer Protocols and Servers George Kola, Tevfik Kosar and Miron Livny University of Wisconsin-Madison USA Motivation Scientific experiments are generating large amounts of data Education

More information

IO-Lite: A Unified I/O Buffering and Caching System

IO-Lite: A Unified I/O Buffering and Caching System IO-Lite: A Unified I/O Buffering and Caching System Vivek S. Pai, Peter Druschel and Willy Zwaenepoel Rice University (Presented by Chuanpeng Li) 2005-4-25 CS458 Presentation 1 IO-Lite Motivation Network

More information

vcloud Automation Center Reference Architecture vcloud Automation Center 5.2

vcloud Automation Center Reference Architecture vcloud Automation Center 5.2 vcloud Automation Center Reference Architecture vcloud Automation Center 5.2 This document supports the version of each product listed and supports all subsequent versions until the document is replaced

More information

Discrete-Approximation of Measured Round Trip Time Distributions: A Model for Network Emulation

Discrete-Approximation of Measured Round Trip Time Distributions: A Model for Network Emulation Discrete-Approximation of Measured Round Trip Time Distributions: A Model for Network Emulation Jay Aikat*, Shaddi Hasan +, Kevin Jeffay*, and F. Donelson Smith* *University of North Carolina at Chapel

More information

Validation of Router Models in OPNET

Validation of Router Models in OPNET Validation of Router Models in OPNET B. Van den Broeck, P. Leys, J. Potemans 1, J. Theunis, E. Van Lil, A. Van de Capelle Katholieke Universiteit Leuven (K.U.Leuven) Department of Electrical Engineering

More information

Why Study Multimedia? Operating Systems. Multimedia Resource Requirements. Continuous Media. Influences on Quality. An End-To-End Problem

Why Study Multimedia? Operating Systems. Multimedia Resource Requirements. Continuous Media. Influences on Quality. An End-To-End Problem Why Study Multimedia? Operating Systems Operating System Support for Multimedia Improvements: Telecommunications Environments Communication Fun Outgrowth from industry telecommunications consumer electronics

More information

MiAMI: Multi-Core Aware Processor Affinity for TCP/IP over Multiple Network Interfaces

MiAMI: Multi-Core Aware Processor Affinity for TCP/IP over Multiple Network Interfaces MiAMI: Multi-Core Aware Processor Affinity for TCP/IP over Multiple Network Interfaces Hye-Churn Jang Hyun-Wook (Jin) Jin Department of Computer Science and Engineering Konkuk University Seoul, Korea {comfact,

More information

Evaluation of Performance of Cooperative Web Caching with Web Polygraph

Evaluation of Performance of Cooperative Web Caching with Web Polygraph Evaluation of Performance of Cooperative Web Caching with Web Polygraph Ping Du Jaspal Subhlok Department of Computer Science University of Houston Houston, TX 77204 {pdu, jaspal}@uh.edu Abstract This

More information

Summary Cache based Co-operative Proxies

Summary Cache based Co-operative Proxies Summary Cache based Co-operative Proxies Project No: 1 Group No: 21 Vijay Gabale (07305004) Sagar Bijwe (07305023) 12 th November, 2007 1 Abstract Summary Cache based proxies cooperate behind a bottleneck

More information

Measuring the Processing Performance of NetSniff

Measuring the Processing Performance of NetSniff Measuring the Processing Performance of NetSniff Julie-Anne Bussiere *, Jason But Centre for Advanced Internet Architectures. Technical Report 050823A Swinburne University of Technology Melbourne, Australia

More information

Initial Evaluation of a User-Level Device Driver Framework

Initial Evaluation of a User-Level Device Driver Framework Initial Evaluation of a User-Level Device Driver Framework Stefan Götz Karlsruhe University Germany sgoetz@ira.uka.de Kevin Elphinstone National ICT Australia University of New South Wales kevine@cse.unsw.edu.au

More information

Operating System Support for Multimedia. Slides courtesy of Tay Vaughan Making Multimedia Work

Operating System Support for Multimedia. Slides courtesy of Tay Vaughan Making Multimedia Work Operating System Support for Multimedia Slides courtesy of Tay Vaughan Making Multimedia Work Why Study Multimedia? Improvements: Telecommunications Environments Communication Fun Outgrowth from industry

More information

Swapping. Operating Systems I. Swapping. Motivation. Paging Implementation. Demand Paging. Active processes use more physical memory than system has

Swapping. Operating Systems I. Swapping. Motivation. Paging Implementation. Demand Paging. Active processes use more physical memory than system has Swapping Active processes use more physical memory than system has Operating Systems I Address Binding can be fixed or relocatable at runtime Swap out P P Virtual Memory OS Backing Store (Swap Space) Main

More information

A Predictable RTOS. Mantis Cheng Department of Computer Science University of Victoria

A Predictable RTOS. Mantis Cheng Department of Computer Science University of Victoria A Predictable RTOS Mantis Cheng Department of Computer Science University of Victoria Outline I. Analysis of Timeliness Requirements II. Analysis of IO Requirements III. Time in Scheduling IV. IO in Scheduling

More information

Quality of Service Mechanism for MANET using Linux Semra Gulder, Mathieu Déziel

Quality of Service Mechanism for MANET using Linux Semra Gulder, Mathieu Déziel Quality of Service Mechanism for MANET using Linux Semra Gulder, Mathieu Déziel Semra.gulder@crc.ca, mathieu.deziel@crc.ca Abstract: This paper describes a QoS mechanism suitable for Mobile Ad Hoc Networks

More information

Implementation and Evaluation of Prefetching in the Intel Paragon Parallel File System

Implementation and Evaluation of Prefetching in the Intel Paragon Parallel File System Implementation and Evaluation of Prefetching in the Intel Paragon Parallel File System Meenakshi Arunachalam Alok Choudhary Brad Rullman y ECE and CIS Link Hall Syracuse University Syracuse, NY 344 E-mail:

More information

PCnet-FAST Buffer Performance White Paper

PCnet-FAST Buffer Performance White Paper PCnet-FAST Buffer Performance White Paper The PCnet-FAST controller is designed with a flexible FIFO-SRAM buffer architecture to handle traffic in half-duplex and full-duplex 1-Mbps Ethernet networks.

More information

Multiprocessor and Real- Time Scheduling. Chapter 10

Multiprocessor and Real- Time Scheduling. Chapter 10 Multiprocessor and Real- Time Scheduling Chapter 10 Classifications of Multiprocessor Loosely coupled multiprocessor each processor has its own memory and I/O channels Functionally specialized processors

More information

IP SLAs Overview. Finding Feature Information. Information About IP SLAs. IP SLAs Technology Overview

IP SLAs Overview. Finding Feature Information. Information About IP SLAs. IP SLAs Technology Overview This module describes IP Service Level Agreements (SLAs). IP SLAs allows Cisco customers to analyze IP service levels for IP applications and services, to increase productivity, to lower operational costs,

More information

Announcement. Exercise #2 will be out today. Due date is next Monday

Announcement. Exercise #2 will be out today. Due date is next Monday Announcement Exercise #2 will be out today Due date is next Monday Major OS Developments 2 Evolution of Operating Systems Generations include: Serial Processing Simple Batch Systems Multiprogrammed Batch

More information

Chapter 9: Virtual Memory

Chapter 9: Virtual Memory Chapter 9: Virtual Memory Silberschatz, Galvin and Gagne 2013 Chapter 9: Virtual Memory Background Demand Paging Copy-on-Write Page Replacement Allocation of Frames Thrashing Memory-Mapped Files Allocating

More information

Chapter 6: CPU Scheduling. Operating System Concepts 9 th Edition

Chapter 6: CPU Scheduling. Operating System Concepts 9 th Edition Chapter 6: CPU Scheduling Silberschatz, Galvin and Gagne 2013 Chapter 6: CPU Scheduling Basic Concepts Scheduling Criteria Scheduling Algorithms Thread Scheduling Multiple-Processor Scheduling Real-Time

More information

Availability and Utility of Idle Memory in Workstation Clusters. Anurag Acharya, UC-Santa Barbara Sanjeev Setia, George Mason Univ

Availability and Utility of Idle Memory in Workstation Clusters. Anurag Acharya, UC-Santa Barbara Sanjeev Setia, George Mason Univ Availability and Utility of Idle Memory in Workstation Clusters Anurag Acharya, UC-Santa Barbara Sanjeev Setia, George Mason Univ Motivation Explosive growth in data intensive applications Large-scale

More information

Introduction to OpenOnload Building Application Transparency and Protocol Conformance into Application Acceleration Middleware

Introduction to OpenOnload Building Application Transparency and Protocol Conformance into Application Acceleration Middleware White Paper Introduction to OpenOnload Building Application Transparency and Protocol Conformance into Application Acceleration Middleware Steve Pope, PhD Chief Technical Officer Solarflare Communications

More information

Chapter 13: I/O Systems

Chapter 13: I/O Systems Chapter 13: I/O Systems Chapter 13: I/O Systems I/O Hardware Application I/O Interface Kernel I/O Subsystem Transforming I/O Requests to Hardware Operations Streams Performance 13.2 Silberschatz, Galvin

More information

The Impact of Write Back on Cache Performance

The Impact of Write Back on Cache Performance The Impact of Write Back on Cache Performance Daniel Kroening and Silvia M. Mueller Computer Science Department Universitaet des Saarlandes, 66123 Saarbruecken, Germany email: kroening@handshake.de, smueller@cs.uni-sb.de,

More information

Introduction to TCP/IP Offload Engine (TOE)

Introduction to TCP/IP Offload Engine (TOE) Introduction to TCP/IP Offload Engine (TOE) Version 1.0, April 2002 Authored By: Eric Yeh, Hewlett Packard Herman Chao, QLogic Corp. Venu Mannem, Adaptec, Inc. Joe Gervais, Alacritech Bradley Booth, Intel

More information

Performance Enhancement for IPsec Processing on Multi-Core Systems

Performance Enhancement for IPsec Processing on Multi-Core Systems Performance Enhancement for IPsec Processing on Multi-Core Systems Sandeep Malik Freescale Semiconductor India Pvt. Ltd IDC Noida, India Ravi Malhotra Freescale Semiconductor India Pvt. Ltd IDC Noida,

More information

Improve Web Application Performance with Zend Platform

Improve Web Application Performance with Zend Platform Improve Web Application Performance with Zend Platform Shahar Evron Zend Sr. PHP Specialist Copyright 2007, Zend Technologies Inc. Agenda Benchmark Setup Comprehensive Performance Multilayered Caching

More information

Statement of Research for Taliver Heath

Statement of Research for Taliver Heath Statement of Research for Taliver Heath Research on the systems side of Computer Science straddles the line between science and engineering. Both aspects are important, so neither side should be ignored

More information

Securing Grid Data Transfer Services with Active Network Portals

Securing Grid Data Transfer Services with Active Network Portals Securing Grid Data Transfer Services with Active Network Portals Onur Demir 1 2 Kanad Ghose 3 Madhusudhan Govindaraju 4 Department of Computer Science Binghamton University (SUNY) {onur 1, mike 2, ghose

More information

Virtual Memory Outline

Virtual Memory Outline Virtual Memory Outline Background Demand Paging Copy-on-Write Page Replacement Allocation of Frames Thrashing Memory-Mapped Files Allocating Kernel Memory Other Considerations Operating-System Examples

More information

Measurement-based Analysis of TCP/IP Processing Requirements

Measurement-based Analysis of TCP/IP Processing Requirements Measurement-based Analysis of TCP/IP Processing Requirements Srihari Makineni Ravi Iyer Communications Technology Lab Intel Corporation {srihari.makineni, ravishankar.iyer}@intel.com Abstract With the

More information

Performance of UMTS Radio Link Control

Performance of UMTS Radio Link Control Performance of UMTS Radio Link Control Qinqing Zhang, Hsuan-Jung Su Bell Laboratories, Lucent Technologies Holmdel, NJ 77 Abstract- The Radio Link Control (RLC) protocol in Universal Mobile Telecommunication

More information

Using Time Division Multiplexing to support Real-time Networking on Ethernet

Using Time Division Multiplexing to support Real-time Networking on Ethernet Using Time Division Multiplexing to support Real-time Networking on Ethernet Hariprasad Sampathkumar 25 th January 2005 Master s Thesis Defense Committee Dr. Douglas Niehaus, Chair Dr. Jeremiah James,

More information

The latency of user-to-user, kernel-to-kernel and interrupt-to-interrupt level communication

The latency of user-to-user, kernel-to-kernel and interrupt-to-interrupt level communication The latency of user-to-user, kernel-to-kernel and interrupt-to-interrupt level communication John Markus Bjørndalen, Otto J. Anshus, Brian Vinter, Tore Larsen Department of Computer Science University

More information

Client-Server Semantic Binary Database: Design and Development

Client-Server Semantic Binary Database: Design and Development Client-Server Semantic Binary Database: Design and Development Konstantin Beznosov High Performance Database Research Center Florida International University http://www.cs.fiu.edu/ beznosov December 9,

More information

Appendix B. Standards-Track TCP Evaluation

Appendix B. Standards-Track TCP Evaluation 215 Appendix B Standards-Track TCP Evaluation In this appendix, I present the results of a study of standards-track TCP error recovery and queue management mechanisms. I consider standards-track TCP error

More information

Web File Transmission by Object Packaging Performance Comparison with HTTP 1.0 and HTTP 1.1 Persistent Connection

Web File Transmission by Object Packaging Performance Comparison with HTTP 1.0 and HTTP 1.1 Persistent Connection Web File Transmission by Performance Comparison with and Hiroshi Fujinoki, Murugesan Sanjay, and Chintan Shah Department of Computer Science Southern Illinois University at Edwardsville Edwardsville, Illinois

More information

Evaluation Strategies. Nick Feamster CS 7260 February 26, 2007

Evaluation Strategies. Nick Feamster CS 7260 February 26, 2007 Evaluation Strategies Nick Feamster CS 7260 February 26, 2007 Evaluation Strategies Many ways to evaluate new protocols, systems, implementations Mathematical analysis Simulation (ns, SSFNet, etc.) Emulation

More information

One Server Per City: Using TCP for Very Large SIP Servers. Kumiko Ono Henning Schulzrinne {kumiko,

One Server Per City: Using TCP for Very Large SIP Servers. Kumiko Ono Henning Schulzrinne {kumiko, One Server Per City: Using TCP for Very Large SIP Servers Kumiko Ono Henning Schulzrinne {kumiko, hgs}@cs.columbia.edu Goal Answer the following question: How does using TCP affect the scalability and

More information

Traffic Characteristics of Bulk Data Transfer using TCP/IP over Gigabit Ethernet

Traffic Characteristics of Bulk Data Transfer using TCP/IP over Gigabit Ethernet Traffic Characteristics of Bulk Data Transfer using TCP/IP over Gigabit Ethernet Aamir Shaikh and Kenneth J. Christensen Department of Computer Science and Engineering University of South Florida Tampa,

More information

CA Single Sign-On. Performance Test Report R12

CA Single Sign-On. Performance Test Report R12 CA Single Sign-On Performance Test Report R12 Contents CHAPTER 1: OVERVIEW INTRODUCTION SUMMARY METHODOLOGY GLOSSARY CHAPTER 2: TESTING METHOD TEST ENVIRONMENT DATA MODEL CONNECTION PROCESSING SYSTEM PARAMETERS

More information

Process- Concept &Process Scheduling OPERATING SYSTEMS

Process- Concept &Process Scheduling OPERATING SYSTEMS OPERATING SYSTEMS Prescribed Text Book Operating System Principles, Seventh Edition By Abraham Silberschatz, Peter Baer Galvin and Greg Gagne PROCESS MANAGEMENT Current day computer systems allow multiple

More information

Application Layer Switching: A Deployable Technique for Providing Quality of Service

Application Layer Switching: A Deployable Technique for Providing Quality of Service Application Layer Switching: A Deployable Technique for Providing Quality of Service Raheem Beyah Communications Systems Center School of Electrical and Computer Engineering Georgia Institute of Technology

More information

Zilog Real-Time Kernel

Zilog Real-Time Kernel An Company Configurable Compilation RZK allows you to specify system parameters at compile time. For example, the number of objects, such as threads and semaphores required, are specez80acclaim! Family

More information

Commercial Real-time Operating Systems An Introduction. Swaminathan Sivasubramanian Dependable Computing & Networking Laboratory

Commercial Real-time Operating Systems An Introduction. Swaminathan Sivasubramanian Dependable Computing & Networking Laboratory Commercial Real-time Operating Systems An Introduction Swaminathan Sivasubramanian Dependable Computing & Networking Laboratory swamis@iastate.edu Outline Introduction RTOS Issues and functionalities LynxOS

More information

Chapter 1. Introduction

Chapter 1. Introduction Chapter 1 Introduction In a packet-switched network, packets are buffered when they cannot be processed or transmitted at the rate they arrive. There are three main reasons that a router, with generic

More information

Operating- System Structures

Operating- System Structures Operating- System Structures 2 CHAPTER Practice Exercises 2.1 What is the purpose of system calls? Answer: System calls allow user-level processes to request services of the operating system. 2.2 What

More information

Improving TCP Performance over Wireless Networks using Loss Predictors

Improving TCP Performance over Wireless Networks using Loss Predictors Improving TCP Performance over Wireless Networks using Loss Predictors Fabio Martignon Dipartimento Elettronica e Informazione Politecnico di Milano P.zza L. Da Vinci 32, 20133 Milano Email: martignon@elet.polimi.it

More information

Computer-System Organization (cont.)

Computer-System Organization (cont.) Computer-System Organization (cont.) Interrupt time line for a single process doing output. Interrupts are an important part of a computer architecture. Each computer design has its own interrupt mechanism,

More information

Operating System Performance and Large Servers 1

Operating System Performance and Large Servers 1 Operating System Performance and Large Servers 1 Hyuck Yoo and Keng-Tai Ko Sun Microsystems, Inc. Mountain View, CA 94043 Abstract Servers are an essential part of today's computing environments. High

More information

Configuring Cisco IOS IP SLAs Operations

Configuring Cisco IOS IP SLAs Operations CHAPTER 50 This chapter describes how to use Cisco IOS IP Service Level Agreements (SLAs) on the switch. Cisco IP SLAs is a part of Cisco IOS software that allows Cisco customers to analyze IP service

More information

Kernel Korner. Analysis of the HTB Queuing Discipline. Yaron Benita. Abstract

Kernel Korner. Analysis of the HTB Queuing Discipline. Yaron Benita. Abstract 1 of 9 6/18/2006 7:41 PM Kernel Korner Analysis of the HTB Queuing Discipline Yaron Benita Abstract Can Linux do Quality of Service in a way that both offers high throughput and does not exceed the defined

More information

Chapter 5 - Input / Output

Chapter 5 - Input / Output Chapter 5 - Input / Output Luis Tarrataca luis.tarrataca@gmail.com CEFET-RJ L. Tarrataca Chapter 5 - Input / Output 1 / 90 1 Motivation 2 Principle of I/O Hardware I/O Devices Device Controllers Memory-Mapped

More information

Configuring Cisco IOS IP SLA Operations

Configuring Cisco IOS IP SLA Operations CHAPTER 58 This chapter describes how to use Cisco IOS IP Service Level Agreements (SLA) on the switch. Cisco IP SLA is a part of Cisco IOS software that allows Cisco customers to analyze IP service levels

More information

L41: Lab 2 - Kernel implications of IPC

L41: Lab 2 - Kernel implications of IPC L41: Lab 2 - Kernel implications of IPC Dr Robert N.M. Watson Michaelmas Term 2015 The goals of this lab are to: Continue to gain experience tracing user-kernel interactions via system calls Explore the

More information

Chapter 6. Design and Implementation of the Experiment

Chapter 6. Design and Implementation of the Experiment Chapter 6. Design and Implementation of the Experiment 6.1. Physical Test Setup The test set-up included two ATM switches (Olicom CrossFire 9100 and 9200), two PCs (450 MHz Pentium III processors with

More information

CS555: Distributed Systems [Fall 2017] Dept. Of Computer Science, Colorado State University

CS555: Distributed Systems [Fall 2017] Dept. Of Computer Science, Colorado State University CS 555: DISTRIBUTED SYSTEMS [THREADS] Shrideep Pallickara Computer Science Colorado State University Frequently asked questions from the previous class survey Shuffle less/shuffle better Which actions?

More information

Switched FC-AL: An Arbitrated Loop Attachment for Fibre Channel Switches

Switched FC-AL: An Arbitrated Loop Attachment for Fibre Channel Switches Switched FC-AL: An Arbitrated Loop Attachment for Fibre Channel Switches Vishal Sinha sinha@cs.umn.edu Department of Computer Science and Engineering University of Minnesota Minneapolis, MN 55455 7481

More information

Performance and Scalability with Griddable.io

Performance and Scalability with Griddable.io Performance and Scalability with Griddable.io Executive summary Griddable.io is an industry-leading timeline-consistent synchronized data integration grid across a range of source and target data systems.

More information

CSMA based Medium Access Control for Wireless Sensor Network

CSMA based Medium Access Control for Wireless Sensor Network CSMA based Medium Access Control for Wireless Sensor Network H. Hoang, Halmstad University Abstract Wireless sensor networks bring many challenges on implementation of Medium Access Control protocols because

More information

The Journal of Systems and Software

The Journal of Systems and Software The Journal of Systems and Software 81 (28) 244 258 Contents lists available at ScienceDirect The Journal of Systems and Software journal homepage: www.elsevier.com/locate/jss Design and implementation

More information

Performance Analysis of a WWW Server

Performance Analysis of a WWW Server Boston University OpenBU Computer Science http://open.bu.edu CAS: Computer Science: Technical Reports 1996-8-5 Performance Analysis of a WWW Server Almeida, Virgílio Boston University Computer Science

More information

Tuning RED for Web Traffic

Tuning RED for Web Traffic Tuning RED for Web Traffic Mikkel Christiansen, Kevin Jeffay, David Ott, Donelson Smith UNC, Chapel Hill SIGCOMM 2000, Stockholm subsequently IEEE/ACM Transactions on Networking Vol. 9, No. 3 (June 2001)

More information

The control of I/O devices is a major concern for OS designers

The control of I/O devices is a major concern for OS designers Lecture Overview I/O devices I/O hardware Interrupts Direct memory access Device dimensions Device drivers Kernel I/O subsystem Operating Systems - June 26, 2001 I/O Device Issues The control of I/O devices

More information

Oscillation of RED with 2way TCP bulk data traffic

Oscillation of RED with 2way TCP bulk data traffic First Author: Oscillation of RED with way TCP bulk data traffic Thomas Ziegler, Salzburg Polytechnic University; Thomas.Ziegler@fh-sbg.ac.at Second Author: Bernhard Hechenleitner, Salzburg Polytechnic

More information

Impact of bandwidth-delay product and non-responsive flows on the performance of queue management schemes

Impact of bandwidth-delay product and non-responsive flows on the performance of queue management schemes Impact of bandwidth-delay product and non-responsive flows on the performance of queue management schemes Zhili Zhao Dept. of Elec. Engg., 214 Zachry College Station, TX 77843-3128 A. L. Narasimha Reddy

More information

A WWW Server Benchmark System in IPv6 Environment. Takao Nakayama. Graduate School of Information Science Nara Institute of Science and Technology

A WWW Server Benchmark System in IPv6 Environment. Takao Nakayama. Graduate School of Information Science Nara Institute of Science and Technology A WWW Server Benchmark System in IPv6 Environment Takao Nakayama Graduate School of Information Science Nara Institute of Science and Technology Background(/2) With the spread of IPv6 technology, we can

More information

An AIO Implementation and its Behaviour

An AIO Implementation and its Behaviour An AIO Implementation and its Behaviour Benjamin C. R. LaHaise Red Hat, Inc. bcrl@redhat.com Abstract Many existing userland network daemons suffer from a performance curve that severely degrades under

More information

This project must be done in groups of 2 3 people. Your group must partner with one other group (or two if we have add odd number of groups).

This project must be done in groups of 2 3 people. Your group must partner with one other group (or two if we have add odd number of groups). 1/21/2015 CS 739 Distributed Systems Fall 2014 PmWIki / Project1 PmWIki / Project1 The goal for this project is to implement a small distributed system, to get experience in cooperatively developing a

More information

Managing NFS and KRPC Kernel Configurations in HP-UX 11i v3

Managing NFS and KRPC Kernel Configurations in HP-UX 11i v3 Managing NFS and KRPC Kernel Configurations in HP-UX 11i v3 HP Part Number: 762807-003 Published: September 2015 Edition: 2 Copyright 2009, 2015 Hewlett-Packard Development Company, L.P. Legal Notices

More information

An FPGA-Based Optical IOH Architecture for Embedded System

An FPGA-Based Optical IOH Architecture for Embedded System An FPGA-Based Optical IOH Architecture for Embedded System Saravana.S Assistant Professor, Bharath University, Chennai 600073, India Abstract Data traffic has tremendously increased and is still increasing

More information

Problem Set: Processes

Problem Set: Processes Lecture Notes on Operating Systems Problem Set: Processes 1. Answer yes/no, and provide a brief explanation. (a) Can two processes be concurrently executing the same program executable? (b) Can two running

More information

The MOSIX Scalable Cluster Computing for Linux. mosix.org

The MOSIX Scalable Cluster Computing for Linux.  mosix.org The MOSIX Scalable Cluster Computing for Linux Prof. Amnon Barak Computer Science Hebrew University http://www. mosix.org 1 Presentation overview Part I : Why computing clusters (slide 3-7) Part II : What

More information

What Operating Systems Do An operating system is a program hardware that manages the computer provides a basis for application programs acts as an int

What Operating Systems Do An operating system is a program hardware that manages the computer provides a basis for application programs acts as an int Operating Systems Lecture 1 Introduction Agenda: What Operating Systems Do Computer System Components How to view the Operating System Computer-System Operation Interrupt Operation I/O Structure DMA Structure

More information

Systems Architecture II

Systems Architecture II Systems Architecture II Topics Interfacing I/O Devices to Memory, Processor, and Operating System * Memory-mapped IO and Interrupts in SPIM** *This lecture was derived from material in the text (Chapter

More information

Virtualization, Xen and Denali

Virtualization, Xen and Denali Virtualization, Xen and Denali Susmit Shannigrahi November 9, 2011 Susmit Shannigrahi () Virtualization, Xen and Denali November 9, 2011 1 / 70 Introduction Virtualization is the technology to allow two

More information

10 MONITORING AND OPTIMIZING

10 MONITORING AND OPTIMIZING MONITORING AND OPTIMIZING.1 Introduction Objectives.2 Windows XP Task Manager.2.1 Monitor Running Programs.2.2 Monitor Processes.2.3 Monitor System Performance.2.4 Monitor Networking.2.5 Monitor Users.3

More information

Large-Scale Network Simulation Scalability and an FPGA-based Network Simulator

Large-Scale Network Simulation Scalability and an FPGA-based Network Simulator Large-Scale Network Simulation Scalability and an FPGA-based Network Simulator Stanley Bak Abstract Network algorithms are deployed on large networks, and proper algorithm evaluation is necessary to avoid

More information

WANMon: A Resource Usage Monitoring Tool for Ad Hoc Wireless Networks

WANMon: A Resource Usage Monitoring Tool for Ad Hoc Wireless Networks WANMon: A Resource Usage Monitoring Tool for Ad Hoc Wireless Networks Don Ngo, Naveed Hussain, Mahbub Hassan School of Computer Science & Engineering The University of New South Wales Sydney, Australia

More information

Evaluation Report: Improving SQL Server Database Performance with Dot Hill AssuredSAN 4824 Flash Upgrades

Evaluation Report: Improving SQL Server Database Performance with Dot Hill AssuredSAN 4824 Flash Upgrades Evaluation Report: Improving SQL Server Database Performance with Dot Hill AssuredSAN 4824 Flash Upgrades Evaluation report prepared under contract with Dot Hill August 2015 Executive Summary Solid state

More information

Operating Systems. Lecture 09: Input/Output Management. Elvis C. Foster

Operating Systems. Lecture 09: Input/Output Management. Elvis C. Foster Operating Systems 141 Lecture 09: Input/Output Management Despite all the considerations that have discussed so far, the work of an operating system can be summarized in two main activities input/output

More information

Performance Analysis of iscsi Middleware Optimized for Encryption Processing in a Long-Latency Environment

Performance Analysis of iscsi Middleware Optimized for Encryption Processing in a Long-Latency Environment Performance Analysis of iscsi Middleware Optimized for Encryption Processing in a Long-Latency Environment Kikuko Kamisaka Graduate School of Humanities and Sciences Ochanomizu University -1-1, Otsuka,

More information

The design and implementation of the NCTUns network simulation engine

The design and implementation of the NCTUns network simulation engine Simulation Modelling Practice and Theory 15 (2007) 57 81 www.elsevier.com/locate/simpat The design and implementation of the NCTUns network simulation engine S.Y. Wang *, C.L. Chou, C.C. Lin Department

More information

IBM MQ Appliance HA and DR Performance Report Model: M2001 Version 3.0 September 2018

IBM MQ Appliance HA and DR Performance Report Model: M2001 Version 3.0 September 2018 IBM MQ Appliance HA and DR Performance Report Model: M2001 Version 3.0 September 2018 Sam Massey IBM MQ Performance IBM UK Laboratories Hursley Park Winchester Hampshire 1 Notices Please take Note! Before

More information

Web File Transmission by Object Packaging Performance Comparison with HTTP 1.0 and HTTP 1.1 Persistent Connection

Web File Transmission by Object Packaging Performance Comparison with HTTP 1.0 and HTTP 1.1 Persistent Connection Web File Transmission by Performance Comparison with HTTP 1. and Hiroshi Fujinoki, Murugesan Sanjay, and Chintan Shah Department of Computer Science Southern Illinois University at Edwardsville Edwardsville,

More information

File Server Comparison: Executive Summary. Microsoft Windows NT Server 4.0 and Novell NetWare 5. Contents

File Server Comparison: Executive Summary. Microsoft Windows NT Server 4.0 and Novell NetWare 5. Contents File Server Comparison: Microsoft Windows NT Server 4.0 and Novell NetWare 5 Contents Executive Summary Updated: October 7, 1998 (PDF version 240 KB) Executive Summary Performance Analysis Price/Performance

More information

Scalability of Linux Event-Dispatch Mechanisms

Scalability of Linux Event-Dispatch Mechanisms Scalability of Linux Event-Dispatch Mechanisms Abhish ek Chandra, David Mosberger Internet and Mobile Systems Laboratory HP Laboratories Palo Alto HPL-2-174 December 14 th, 2* E-mail: abh ish ek@cs.umass.edu,

More information