Combining Congested-Flow Isolation and Injection Throttling in HPC Interconnection Networks

Escudero-Sahuquillo, Jesús; Gran, Ernst Gunnar; García, Pedro J.; Flich, José; Skeie, Tor; Lysne, Olav; Quiles, Francisco J.; Duato, J.

doi:10.1109/icpp.2011.80

Cited by 14 publications

(5 citation statements)

References 23 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Since adaptive routing cannot differentiate between end-node congestion and network congestion, it does not undermine the need for high-quality deterministic routing. Instead, research focuses on techniques to either reduce the potentially harmful collateral impact of adaptive routing or reduce congestion from happening in the first place with injection-throttling [1] or better deterministic routing. The latter is the aim of this article.…”

Section: Side-note On Adaptive Routingmentioning

confidence: 99%

“…For example if using a Lustre file system, Lustre routers can be nodes of the cluster leading to an array of IO servers of which the fabric management and routing algorithm are not aware. As a concrete example, BXI 1 switches have 48 ports. Some BXI switch have only copper ports, some others have three optical ports.…”

Section: Heterogeneous Clustersmentioning

confidence: 99%

“…We use a topology with nonfull CBB because otherwise there would be no possible congestion at any top-port. 1 BXI is the interconnect technology developed by Bull/Atos. It comprises hardware (switches, links) and software (firmware, low-level and high-level development environment on which are built the fabric management and routing algorithms).…”

Section: Analysis Of a Type-based Communication Patternmentioning

confidence: 99%

See 2 more Smart Citations

Node-Type-Based Load-Balancing Routing for Parallel Generalized Fat-Trees

Gliksberg¹,

Quintin²,

García³

2018

2018 IEEE 4th International Workshop on High-Performance Interconnection Networks in the Exascale and Big-Data Era (HiPINEB)

Self Cite

View full text Add to dashboard Cite

High-Performance Computing (HPC) clusters are made up of a variety of node types (usually compute, I/O, service, and GPGPU nodes) and applications don't use nodes of a different type the same way. Resulting communication patterns reflect organization of groups of nodes, and current optimal routing algorithms for all-to-all patterns will not always maximize performance for group-specific communications. Since application communication patterns are rarely available beforehand, we choose to rely on node types as a good guess for node usage. We provide a description of node type heterogeneity and analyse performance degradation caused by unlucky repartition of nodes of the same type. We provide an extension to routing algorithms for Parallel Generalized Fat-Tree topologies (PGFTs) which balances load amongst groups of nodes of the same type. We show how it removes these performance issues by comparing results in a variety of situations against corresponding classical algorithms.

show abstract

Section: Side-note On Adaptive Routingmentioning

confidence: 99%

Section: Heterogeneous Clustersmentioning

confidence: 99%

Section: Analysis Of a Type-based Communication Patternmentioning

confidence: 99%

See 1 more Smart Citation

Node-Type-Based Load-Balancing Routing for Parallel Generalized Fat-Trees

Gliksberg¹,

Quintin²,

García³

2018

2018 IEEE 4th International Workshop on High-Performance Interconnection Networks in the Exascale and Big-Data Era (HiPINEB)

Self Cite

View full text Add to dashboard Cite

show abstract

“…Gran [10] investigated the relationship among congestion control, switch arbitration and fairness in lossless interconnection network. Jesus [11] combined the traffic injection throttling and congestion flow isolation to promptly react to the congestion.…”

Section: Introductionmentioning

confidence: 99%

IBperf: A Testing and Evaluation Model of IB Datacenter Transmission Control

Zhang

Sun

2012

AMM

View full text Add to dashboard Cite

Datacenter is the important infrastructure to provide resource and service in Internet. Due to the widely observed congestion, all kinds of transmission control mechanisms have been proposed to satisfy diverse network performance demands. Based on InfiniBand (IB) which is widely applied for datacenter construction, this paper proposed a generic testing and evaluation platform IBperf to verify transmission control protocols. We designed a central controller to instruct distributed clients to generate different communication patterns and execute commands of the protocol. The transmission performance can be derived based on the captured packets and digest. In the experiment, IB congestion control (CC) was evaluated through IBperf on our real IB platform. IBperf showed the improvement of fairness and throughput with low overhead based on CC mechanism. It proved IBperf to be an effective tool for the evaluation of IB datacenter.

show abstract

“…Network congestion is widely recognized as one of the primary causes of performance degradation, performance variability, and poor scaling in communication-heavy applications running on supercomputers [1], [2], [3], [4], [5]. However, due to the complex nature of interconnection networks, as well as message injection and routing strategies, network congestion and its root causes for network resources and hardware components are not well understood.…”

Section: Motivation and Impactmentioning

confidence: 99%

Identifying the Culprits Behind Network Congestion

Bhatelé

Titus

Thiagarajan

et al. 2015

2015 IEEE International Parallel and Distributed Processing Symposium

View full text Add to dashboard Cite

Abstract-Network congestion is one of the primary causes of performance degradation, performance variability and poor scaling in communication-heavy parallel applications. However, the causes and mechanisms of network congestion on modern interconnection networks are not well understood. We need new approaches to analyze, model and predict this critical behavior in order to improve the performance of large-scale parallel applications. This paper applies supervised learning algorithms, such as forests of extremely randomized trees and gradient boosted regression trees, to perform regression analysis on communication data and application execution time. Using data derived from multiple executions, we create models to predict the execution time of communication-heavy parallel applications. This analysis also identifies the features and associated hardware components that have the most impact on network congestion and in turn, on execution time. The ideas presented in this paper have wide applicability: predicting the execution time on a different number of cores, or different input datasets, or even for an unknown code, identifying the best configuration parameters for an application, and finding the root causes of network congestion on different architectures. I. MOTIVATION AND IMPACTNetwork congestion is widely recognized as one of the primary causes of performance degradation, performance variability, and poor scaling in communication-heavy applications running on supercomputers [5]. However, due to the complex nature of interconnection networks, as well as message injection and routing strategies, network congestion and its root causes for network resources and hardware components are not well understood. This makes the problem of mitigating and avoiding network congestion difficult. It also complicates the task of writing congestionavoiding and congestion-minimizing algorithms for communication and task mapping. Therefore, we need new approaches to understand and model network congestion in order to improve the performance of large-scale parallel applications.When a message is sent from one node to another, it is split into packets that pass through many resources and hardware components on the network. A packet starts in an injection FIFO on the source. It then passes through multiple network links and receive buffers on intermediate nodes before it finally lands in the reception FIFO on the destination. When shared by multiple packets, any or all of these network components can slow down individual flits, packets and messages. This paper aims to identify the hardware components that affect the performance of sending a message the most.Our approach is based on using supervised machine learning to build models that map from independent variables, representing different network hardware components, to a dependent variable -the execution time of the application. We only consider computationally balanced, communicationheavy parallel applications and, hence, focus on the communication fraction of the total executi...

show abstract

Combining Congested-Flow Isolation and Injection Throttling in HPC Interconnection Networks

Cited by 14 publications

References 23 publications

Node-Type-Based Load-Balancing Routing for Parallel Generalized Fat-Trees

Node-Type-Based Load-Balancing Routing for Parallel Generalized Fat-Trees

IBperf: A Testing and Evaluation Model of IB Datacenter Transmission Control

Identifying the Culprits Behind Network Congestion

Contact Info

Product

Resources

About