Reliable Distributed Systems (PDF)
An understanding of the techniques used to make distributed computing systems and networks reliable, fault-tolerant and secure will be crucial to those who design and deploy the next generation of mission-critical applications and Web...
42 DeutschlandCard Punkte sammeln
- Lastschrift, Kreditkarte, Paypal, Rechnung
- Kostenloser tolino webreader
An understanding of the techniques used to make distributed computing systems and networks reliable, fault-tolerant and secure will be crucial to those who design and deploy the next generation of mission-critical applications and Web Services.
Reliable Distributed Systems
reviews and describes the key concepts, principles and applications of modern distributed computing systems and architectures. This self-contained book consists of five parts. The first covers introductory material, including the basic architecture of the Internet, simple protocols such as RPC and TCP, object oriented architectures, operating systems enhance-ments for high performance, and reliability issues. The second covers the Web, with a focus on Web Services technologies, Microsoft's .NET and the Java Enterprise Edition. The last three parts look at a number of reliability and fault-tolerance issues and techniques, with an emphasis on replication applied in Web Services settings.Topics and features:
* Explains fault-tolerance in clear, readily understood terms with concrete examples drawn from real-world settings
* A practical focus aimed at building "mission-critical" networked applications that keep working even when things go wrong
* Includes modern topics, such as Corba, Web Services, XML, .NET, J2EE, group communication, transactions, peer-to-peer systems, time-critical protocols, scalability and security
* Thorough coverage of fundamental mechanisms, with an emphasis on the idea of "consistent behavior" in systems that replicate critical components for availability
* Reviews more than 25 major research efforts, placing them in context with pointers to sources
* Includes 80 problems ranging from simple tests of understanding to challenging protocol and systems design topics suitable for semester-long projects
* Web-based materials for instructors, including a comprehensive slide set,available at: http://www.cs.cornell.edu/ken/book
With its well-focused approach and clarity of presentation, this new text is an excellent resource for both advanced students and practitioners in computer science, computer networks and distributed systems. Anyone seeking a solid background in distributed computing and Web Services architectures will find the book an essential and practical learning tool.
The Web is just the latest of a series of Internet technologies that have gained extremely wide acceptance. In this chapter we briefly touch on some of the other important members of this technology family, including both old technologies such as mail and file transfer and new ones such as high-speed message bus architectures and security firewalls.
11.1 File Transfer Tools
The earliest networking technologies were those supporting file transfer in distributed settings. These typically consisted of programs for sending and receiving files, commands for initiating transfers and managing file transfer queues during periods when transfers backed up, utilities for administering the storage areas within which files are placed while a transfer is pending, and policies for assigning appropriate ownership and access rights to transferred files.
The most common file transfer mechanism in modern computer systems is that associated with the File Transfer Protocol (FTP), which de.nes a set of standard message formats and request types for navigating in a file system, searching directories, and moving files. FTP includes a security mechanism based on password authentication; however, these passwords are transmitted in an insecure way over the network, exposing them to potential attack by intruders. Many modern systems employ nonreusable passwords for this reason.
Other well-known file transfer protocols include the UNIX-to-UNIX copy program (UUCP) and the file transfer protocol standardized by the OSI protocol suite. Neither protocol is widely used, however, and FTP is a de facto standard within the Internet. However, file transfer over the Web conceals the FTP protocol under a layer of browser technology, so the user doesnt see the login mechanism unless anonymous login fails, and wont normally have access to the various FTP commands for controlling the transfer type (binary or
11.2 Electronic Mail
Electronic mail was the first of the Internet applications to gain wide popularity, and remains a dominant technology at the time of this writing. Mail systems have become steadily easier to use and more sophisticated over time, and e-mail users are supported by increasingly sophisticated mail-reading and composition tools.
Indeed, in a somewhat bizarre twist, e-mail has even been used as a means of computer-to-computer communication, with the users mailbox serving as a form of message queuing middleware! Doing this can potentially allow an application to create an IP tunnel that works around firewall and NAT restrictions, although obviously at a high cost. E-mail has also become so integrated with the web that it is now possible to e-mail interactive Web pages and other forms of dynamic content to cooperative users.
Underlying the e-mail system is a small collection of very simple protocols, of which the Simple Mail Transfer Protocol, or SMTP, is the most widely used and the most standard. The architecture of a typical mailing system is as follows. The user composes a mail message, which is encoded into ASCII (perhaps using a MIME representation) and then stored in a queue of outgoing e-mail messages. Periodically, this queue is scanned by a mail daemon program, which uses SMTP to actually transmit the message to its destinations.
KENNETH P. BIRMAN is an ACM Fellow and a Professor of Computer Science at Cornell University, where his research focuses on reliable, secure, and scalable distributed computing systems. His work has been applied to the NYSE, the Swiss Stock Exchange, the US Navy's AEGIS warship, and the French air-traffic control system.
- Autor: Kenneth Birman
- 2005, 668 Seiten, Englisch
- Verlag: Springer-Verlag GmbH
- ISBN-10: 0387276017
- ISBN-13: 9780387276014
- Erscheinungsdatum: 02.07.2006
Abhängig von Bildschirmgröße und eingestellter Schriftgröße kann die Seitenzahl auf Ihrem Lesegerät variieren.
- Dateiformat: PDF
- Größe: 6.09 MB
- Ohne Kopierschutz
- Vorlesefunktion
Zustand | Preis | Porto | Zahlung | Verkäufer | Rating |
---|
Schreiben Sie einen Kommentar zu "Reliable Distributed Systems".
Kommentar verfassen