PSockets: The Case for Application-level Network Striping for Data Intensive Applications using High Speed Wide Area Networks

Sivakumar, H.; Bailey, Stephen; Grossman, Robert L.

doi:10.1109/sc.2000.10040

Cited by 125 publications

(128 citation statements)

References 5 publications

Supporting

Mentioning

124

Contrasting

Order By: Relevance

“…This extends prior work on application-level flow control or data striping [3,42], as well as approaches that focus on the system-level provision of network status in- formation [30]. It also distinguishes our work from the many, middleware-level adaptation infrastructures developed in previous research, including BBN's Quo [51] or other object-based infrastructures [50].…”

Section: Discussion and Related Worksupporting

confidence: 62%

IQ‐Services: network‐aware middleware for interactive large‐data applications

Cai¹,

Eisenhauer²,

He³

et al. 2005

Concurrency and Computation

View full text Add to dashboard Cite

IQ-Services are application-specific, resource-aware code modules executed by data transport middleware. They constitute a 'thin' layer between application components and the underlying computational and communication resources that implements the data manipulations necessary to permit wide-area collaborations to proceed smoothly, despite dynamic resource variations. IQ-Services interact with the application and resource layers via dynamic performance attributes, and end-to-end implementations of such attributes also permit clients to interact with data providers. Joint middleware/resource and provider/consumer interactions implement a cooperative approach to data management for the large-data applications targeted by our research. Experimental results in this paper demonstrate substantial performance improvements attained by coordinating network-level with service-level adaptations of the data being transported and by permitting end users to dynamically deploy and use application-specific services for manipulating data in ways suitable for their current needs.

show abstract

Section: Discussion and Related Worksupporting

confidence: 62%

IQ‐Services: network‐aware middleware for interactive large‐data applications

Cai¹,

Eisenhauer²,

He³

et al. 2005

Concurrency and Computation

View full text Add to dashboard Cite

show abstract

“…Alternatively, the application can be modified to initiate multiple TCP connections in parallel 2,3) to increase throughput by aggregating multiple TCP connections. This approach effectively multiplies the AWnd and congestion window size (CWnd) by the number of TCP flows and so can mitigate the AWnd limitation.…”

Section: Sender-receiver-based Approachesmentioning

confidence: 99%

TCP-SuperCharger : A New Approach to High-Throughput Satellite Data Transfer

Lee

Yum

Wan

2010

AEROSPACE TECHNOLOGY JAPAN

View full text Add to dashboard Cite

It is well-known that TCP does not perform well in modern satellite-based communications due to the channel's very large bandwidth-delay product. The conventional solution is to make use of TCP's Large Window Scale (LWS) extension as defined in RFC1323 to increase TCP's windows size. However this can only be done if either the operating system or the application can be modified to explicitly make use of TCP's LWS extension during connection setup. This work proposes a novel TCP-SuperCharger (TCP-SC) to enable TCP to fully utilize the underlying network bandwidth without modification to the applications or the operating system. TCP-SC is implemented as a gateway along the path between the sender and the receiver. By estimating the receiver's processing capacity, TCP-SC allows the sender to transmit more data than the amount of receiver buffer available. As modern computers are very fast compared to network speed, this prevents the receiver's buffer limit from restricting the throughput achievable by TCP. Experimental results show that TCP-SC can enable existing TCP with default buffer size (64 KB) to achieve 98% link utilization over a satellite link such as the WINDS communications satellite with 1,000 ms RTT and 24 Mbps bandwidth.

show abstract

“…Heyman and Lucantoni, (2003) show that by using rate limiting to lower the effective bandwidth across links the aggregated bandwidth used can be maximized across all the streams (and links) to create a total bandwidth much larger than any single instance on these links. Experiments in how TCP performs when split across multiple steams has been performed (Sivakumar, Bailey and Grossman, 2000) but these results haven't been scaled to study results across multiple, distinct, paths. Most tests suffer from the congestion caused by their own streams, while we are looking at the results obtained using separate paths.…”

Section: Related Workmentioning

confidence: 99%

Improving the bulk data transfer experience

Guok

Lee

Berket

2008

IJIPT

View full text Add to dashboard Cite

Scientific computations and collaborations increasingly rely on the network to provide high-speed data transfer, dissemination of results, access to instruments, support for computational steering, etc. The Energy Sciences Network is establishing a science data network to provide user driven bandwidth allocation. In a shared network environment, some reservations may not be granted due to the lack of available bandwidth on any single path. In many cases, the available bandwidth across multiple paths would be sufficient to grant the reservation. In this paper we investigate how to utilize the available bandwidth across multiple paths in the case of bulk data transfer.

show abstract

PSockets: The Case for Application-level Network Striping for Data Intensive Applications using High Speed Wide Area Networks

Cited by 125 publications

References 5 publications

IQ‐Services: network‐aware middleware for interactive large‐data applications

IQ‐Services: network‐aware middleware for interactive large‐data applications

TCP-SuperCharger : A New Approach to High-Throughput Satellite Data Transfer

Improving the bulk data transfer experience

Contact Info

Product

Resources

About