Tuple switching network—When slower may be better

Shi, Justin Y.; Taifi, Moussa; Khreishah, Abdallah; Wu, Jie

doi:10.1016/j.jpdc.2012.01.014

Cited by 8 publications

(2 citation statements)

References 22 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Each node is a standalone computer with any number of cores, local memory, storage and multiple network interfaces. UVR (Unidirectional Virtual Ring) implements a Service Content Addressable Network (or Tuple Switching Network (TSN) [16]) using all available processing and networking components. This allows zero single point failures regardless the number of networks and processors for a given application.…”

Section: Single-sided Statistic Multiplexingmentioning

confidence: 99%

Scalability Dilemma and Statistic Multiplexed Computing — A Theory and Experiment

Shi¹

2017

Adv. sci. technol. eng. syst. j.

View full text Add to dashboard Cite

show abstract

Section: Single-sided Statistic Multiplexingmentioning

confidence: 99%

Scalability Dilemma and Statistic Multiplexed Computing — A Theory and Experiment

Shi¹

2017

Adv. sci. technol. eng. syst. j.

View full text Add to dashboard Cite

show abstract

“…Previous attempts include the Linda project [21] and more recently Synergy [14]. While these projects have added a set of innovations to SPP the current implementation improves on these projects by providing two main improvements: 1) A Distributed Compute Space to reduce the IO bottleneck on the master when scaling the application [20] and to decouple the compute space data from the master process. 2) Master Spatial Redundancy to provide master process fault tolerance, contrary to the viable but less portable checkpoint restart mechanism that was designed previously in [12] A.…”

Section: Implementation and Failures Considerationsmentioning

confidence: 99%

JENERGY: A Fault Tolerant Stateless Architecture for High Performance Computing

Taifi

Shi

Celik

2015

2015 IEEE Symposium on Service-Oriented System Engineering

Self Cite

View full text Add to dashboard Cite

Large scale HPC (high performance computing) applications require thousands of nodes for computing parallel scientific applications. At this scale, hardware and software failures, network congestion or disconnections are frequent faults experienced by compute nodes. This introduces high levels of volatility which reduces the Mean Time between Failures (MTBF) of the whole system down to hours or minutes. To deal with this kind of failure rates, traditional point-topoint transmission semantics can be ill-fitted and cumbersome to re-engineer to support distributed partial failures. In this paper, we propose an application dependent network design that focuses on the sustainability of High Performance Computing (HPC) applications using packet-switching-inspired statistical multiplexing of semantic data tuples and decoupled computations. We report the design and implementation of a distributed tuple space using Cassandra and Zookeeper for tunable spatial and temporal redundancies without negative impact on application performance. We detail the various failure scenarios that can be handled seamlessly by our system and provide a description of the advantages of Stateless Parallel Processing for HPC applications. We report the preliminary results on performance, reliability and overall application scalability. We found that our system can provide high levels of sustained performance, while providing a reliable computing architecture that can withstand a range of failure types without manual checkpoint-restart, in a portable and non-intrusive manner.

show abstract

Use HDLC Principles for Distributed Applications

Shi

2015

2015 IEEE International Conference on Computer and Information Technology; Ubiquitous Computing and Communications; Dependable,

View full text Add to dashboard Cite

Tuple switching network—When slower may be better

Cited by 8 publications

References 22 publications

Scalability Dilemma and Statistic Multiplexed Computing — A Theory and Experiment

Scalability Dilemma and Statistic Multiplexed Computing — A Theory and Experiment

JENERGY: A Fault Tolerant Stateless Architecture for High Performance Computing

Use HDLC Principles for Distributed Applications

Contact Info

Product

Resources

About