A Transparent Accelerating Software Architecture for Network Storage Based on Multi-core Heterogeneous Systems

Shang, Qiuli; Wang, Jinlin; Chen, Xiao

doi:10.1007/978-3-319-42085-1_43

Cited by 2 publications

(4 citation statements)

References 4 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…On basis of the mentioned transparent accelerating software architecture [5], the implementation of TCP pacing algorithm (e.g. TCP-SRP [4]) based on multi-core heterogeneous systems is shown in Fig.1.…”

Section: Methodsmentioning

confidence: 99%

“…Based on the P-SPL multi-core topology proposed by the previous work [5], it implements the TCP/IP stack of TCP-SRP and the original heterogeneous TCP/IP stack on separate group of cores in parallel. In this way, TCP-SRP can be an optional alternative to the traditional TCP algorithms.…”

Section: Methodsmentioning

confidence: 99%

“…By the features of parallel processing and hardware acceleration, multi-core network processor (NP) can support data processing for network storage protocols and TCP/IP in high-speed network environment. Thus, multi-core NPs are becoming the mainstream platform of next generation network storage systems, high-performance network servers, and network edge devices [4,5].…”

Section: Introductionmentioning

confidence: 99%

“…For the demands on both protocol acceleration and transparent development, a transparent accelerating software architecture for network storage based on multi-core heterogeneous systems has been presented in previous works [5]. The design of heterogeneous system divides the processing cores into two sets: one carries a general-purpose operating system which is Linux operating system (OS) in this paper; the other carries the network processing operating system which is the simple execution environment (SE) provided by Cavium OCTEON.…”

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

An Implementation of TCP Pacing Algorithm based on Multi-core Heterogeneous Systems

Shang¹,

Tang²,

Liu³

et al. 2016

Proceedings of the 2016 4th International Conference on Electrical &Amp; Electronics Engineering and Computer Science (ICEEECS

Self Cite

View full text Add to dashboard Cite

By the features of parallel processing and hardware acceleration, multi-core network processor (NP) are becoming the mainstream platform of next generation network edge devices. In order to take advantage of multi-core NP platform and improve the performance of TCP pacing algorithm, this paper proposes an implementation of TCP pacing algorithm based on multi-core heterogeneous systems. The implementation provides a timing method that is more accurate than the clock interrupt-based timing in Linux kernel. And it is an optional alternative to the traditional TCP implementations. The experimental test based on prototype platform shows that, the smoothness performance of the proposed implementation is improved by 27.78%, and the resource utilization is also better than the traditional scheme.

show abstract

Section: Methodsmentioning

confidence: 99%

Section: Methodsmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations