Fault-tolerant and scalable TCP splice and web server architecture
Research output: Contribution to conferences › Paper › Contributed › peer-review
Contributors
Abstract
This paper describes three enhancements to the TCP splic-
ing mechanism: (1) Enable a TCP connection to be simulta-
neously spliced through multiple machines for higher scala-
bility; (2) Make a spliced connection fault-tolerant to proxy
failures; and (3) Provide flexibility of splitting a TCP splice
between a proxy and a backend server for further increas-
ing the scalability of a web server system. A web server ar-
chitecture based on this enhanced TCP splicing is proposed.
This architecture provides a highly scalable, seamless ser-
vice to the users with minimal disruption during server fail-
ures. In addition to the traditional web services in which users
download webpages, multimedia files and other types of data
from a web server, the proposed architecture supports newly
emerging web services that are highly interactive, and involve
relatively longer, stateful client-server sessions. A prototype
of this architecture has been implemented as a Linux 2.6 ker-
nel module, and the paper presents important performance
results measured from this implementation.
ing mechanism: (1) Enable a TCP connection to be simulta-
neously spliced through multiple machines for higher scala-
bility; (2) Make a spliced connection fault-tolerant to proxy
failures; and (3) Provide flexibility of splitting a TCP splice
between a proxy and a backend server for further increas-
ing the scalability of a web server system. A web server ar-
chitecture based on this enhanced TCP splicing is proposed.
This architecture provides a highly scalable, seamless ser-
vice to the users with minimal disruption during server fail-
ures. In addition to the traditional web services in which users
download webpages, multimedia files and other types of data
from a web server, the proposed architecture supports newly
emerging web services that are highly interactive, and involve
relatively longer, stateful client-server sessions. A prototype
of this architecture has been implemented as a Linux 2.6 ker-
nel module, and the paper presents important performance
results measured from this implementation.
Details
Original language | English |
---|---|
Pages | 301-310 |
Number of pages | 10 |
Publication status | Published - 2006 |
Peer-reviewed | Yes |
Conference
Title | 2006 25th IEEE Symposium on Reliable Distributed Systems |
---|---|
Abbreviated title | SRDS 2006 |
Conference number | 25 |
Duration | 2 - 4 October 2006 |
Degree of recognition | International event |
City | Leeds |
Country | United Kingdom |
External IDs
Scopus | 38949166348 |
---|
Keywords
Research priority areas of TU Dresden
DFG Classification of Subject Areas according to Review Boards
Keywords
- fault tolerance, web server, service oriented architecture, splicing, Scalability, Web services, Fault tolerant systems, Web pages, Prototypes, Linux, file servers, software architecture, software fault tolerance, transport protocols, fault tolerant TCP splice, scalable TCP splice, proxy failure tolerance