TCP/IP behavior in a high-speed local ATM network environment

Modeklev, K.; Klovning, Espen; Kure, Øivind

doi:10.1109/lcn.1994.386602

Cited by 13 publications

(8 citation statements)

References 12 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The TCP implementation used in SunOS 4.1.3_Ul is a production quality implementation which has been highly optimised during the years [MKK94]. It is representative of a wide range of currently available BSD-derived implementations.…”

Section: Possible Changes To the Tcp Implementationmentioning

confidence: 99%

Is VBR a solution for an ATM LAN ?

Bonaventure

Klovning

Danthine

1997

Protocols for High-Speed Networks V

View full text Add to dashboard Cite

W e present and discuss detailed measurem!'lnts showing how TCP behaves in an A TM LAN where a Variable Bit Rate (VBR) traffic contract is enforced by a Usage Parameter Control (UPC) mechanism. These measurements show that TCP has a Iot of difficulties to adapt its behaviour to such a traffic contract. In our environment, TCP only achieved a 10% utilisation of a 10 Mbps VBR VC. We then discuss and eva!uate possible solutions to improve this low utilisation of the reserved bandwidth.

show abstract

Section: Possible Changes To the Tcp Implementationmentioning

confidence: 99%

Is VBR a solution for an ATM LAN ?

Bonaventure

Klovning

Danthine

1997

Protocols for High-Speed Networks V

View full text Add to dashboard Cite

show abstract

“…Existing studies [7,11,6,23,18] of transport protocol performance over ATM demonstrate the impact of parameters such as socket queue sizes and data buffer on performance. Therefore, our TTCP benchmarks varied these two parameters for each type of data as follows:…”

Section: Ttcp Parameter Settingsmentioning

confidence: 99%

“…These parameters influence the size of the TCP segment window, which has been shown to significantly impact CORBA-level and TCP-level performance on high-speed networks [11,23]. Since the performance of the 8 K socket queues was consistently one-half to two-thirds slower than using the 64 K queues, we omitted the 8 K results from the figures below.…”

Section: Ttcp Parameter Settingsmentioning

confidence: 99%

“…None of the systems described below are explicitly targeted for the requirements and constraints of communication middleware like CORBA. In particular, less attention has been paid to integrating the following topics related to communication middleware: 4.1 Transport Protocol Performance over ATM Networks [7,11,6] present results on performance of TCP/IP (and UDP/IP [6]) on ATM networks by varying a number of parameters (such as TCP window size, socket queue size, and user data size). This work indicates that in addition to the host architecture and host network interface, parameters configurable in software (like TCP window size, socket queue size and user data size) significantly affect TCP throughput.…”

Section: Related Workmentioning

confidence: 99%

See 1 more Smart Citation

Measuring the performance of communication middleware on high-speed networks

Gokhale

Schmidt

1996

SIGCOMM Comput. Commun. Rev.

View full text Add to dashboard Cite

Conventional implementations of communication middleware (such as CORBA and traditional RPC toolkits) incur considerable overhead when used for performance-sensitive applications over highspeed networks. As gigabit networks become pervasive, inefficient middleware will force programmers to use lower-level mechanisms to achieve the necessary transfer rates. This is a serious problem for mission/life-critical applications (such as satellite surveillance and medical imaging).This paper compares the performance of several widely used communication middleware mechanisms on a high-speed ATM network. The middleware ranged from lower-level mechanisms (such as socket-based C interfaces and C++ wrappers for sockets) to higher-level mechanisms (such as RPC, hand-optimized RPC and two implementations of CORBA -Orbix and ORBeline). These measurements reveal that the lower-level C and C++ implementations outperform the CORBA implementations significantly (the best CORBA throughput for remote transfer was roughly 75 to 80 percent of the best C/C++ throughput for sending scalar data types and only around 33 percent for sending structs containing binary fields), and the hand-optimized RPC code performs slightly better than the CORBA implementations. Our goal in precisely pinpointing the sources of overhead for communication middleware is to develop scalable and flexible CORBA implementations that can deliver gigabit data rates to applications.

show abstract

“…The CPE equipment varies greatly from customer to customer and has noticeable impact on service performance. As discussed in [4] there are many aspects 0-7803-3336-5196 $5.00 0 1996 IEEE of a system's architecture that influence performance, a faster CPU alone does not necessarily imply better performance. ISP performance is also variable and dependent on its connectivity to the backbone network (direct or indirect through subnetworks), server and router performance, and customer access to the ISP (speed and blocking).…”

Section: R E a L T I M E I N T E R N E T S E R V I C E Smentioning

confidence: 99%

Availability of real time Internet services

Garzia

Mulhotra²

Proceedings of GLOBECOM'96. 1996 IEEE Global Telecommunications Conference

View full text Add to dashboard Cite

quality of service guarantees become a key business differentiator to competitor's offers.The Internet has seen dramatic growth over the last few years aided by the increasing popularity of the World Wide Web. Businesses are rushing to provide a presence on the Web and the number of consumers sutJfig the Web grows daily. This growth in traffic and applications has imphcations on the expected quality of services. As competition for services increases and customers are asked to pay for these services, performance will become a key differentiator. In this paper we consider quality of service issues for real time Intemet applications. In particular we define service availability, the probability that a customer will be properly serviced by a given application at a given point in time, as a key measure of service qualrty. We discuss the network and service factors affecting

show abstract

TCP/IP behavior in a high-speed local ATM network environment

Cited by 13 publications

References 12 publications

Is VBR a solution for an ATM LAN ?

Is VBR a solution for an ATM LAN ?

Measuring the performance of communication middleware on high-speed networks

Availability of real time Internet services

Contact Info

Product

Resources

About