What Size Should Your Buffers to Disks be?

Aupy, Guillaume; Beaumont, Olivier; Eyraud-Dubois, Lionel

doi:10.1109/ipdps.2018.00075

Cited by 13 publications

(29 citation statements)

References 22 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…We report extensive simulations to evaluate our linear programming formulations, and compare them with a classic fair sharing approach. To do so, we instantiate our evaluation based on characteristics of the Intrepid platform, and based on a set of applications as described in the APEX report [27] and our previous work [22]. a) Setup: We consider a set of 4 applications described in the APEX report [27] which represent the majority of the load at LANL.…”

Section: Simulation Resultsmentioning

confidence: 99%

See 1 more Smart Citation

Sizing and Partitioning Strategies for Burst-Buffers to Reduce IO Contention

Aupy¹,

Beaumont²,

Eyraud-Dubois³

2019

2019 IEEE International Parallel and Distributed Processing Symposium (IPDPS)

Self Cite

View full text Add to dashboard Cite

Burst-Buffers are high throughput and small size storage which are being used as an intermediate storage between the PFS (Parallel File System) and the computational nodes of modern HPC systems. They can allow to hinder to contention to the PFS, a shared resource whose read and write performance increase slower than processing power in HPC systems. A second usage is to accelerate data transfers and to hide the latency to the PFS. In this paper, we concentrate on the first usage. We propose a model for Burst-Buffers and application transfers.We consider the problem of dimensioning and sharing the Burst-Buffers between several applications. This dimensioning can be done either dynamically or statically. The dynamic allocation considers that any application can use any available portion of the Burst-Buffers. The static allocation considers that when a new application enters the system, it is assigned some portion of the Burst-Buffers, which cannot be used by the other applications until that application leaves the system and its data is purged from it. We show that the general sharing problem to guarantee fair performance for all applications is an NP-Complete problem. We propose a polynomial time algorithms for the special case of finding the optimal buffer size such that no application is slowed down due to PFS contention, both in the static and dynamic cases. Finally, we provide evaluations of our algorithms in realistic settings. We use those to discuss how to minimize the overhead of the static allocation of buffers compared to the dynamic allocation.

show abstract

Section: Simulation Resultsmentioning

confidence: 99%

“…1 that will be used throughout this paper. In our previous work [22], we have studied management strategies using a very simplistic random application model. Using Markov chain, we were able to show that the bandwidth B BB to the Burst-Buffers (see Fig.…”

Section: B Algorithms To Deal With Burst-buffersmentioning

confidence: 99%

Sizing and Partitioning Strategies for Burst-Buffers to Reduce IO Contention

Aupy¹,

Beaumont²,

Eyraud-Dubois³

2019

2019 IEEE International Parallel and Distributed Processing Symposium (IPDPS)

Self Cite

View full text Add to dashboard Cite

show abstract

“…When it comes to I/O accelerators parametrization and more generally the understanding of their behavior, Reference 17 has suggested to use a Markov chain based model to describe the behavior of the burst buffer. Reference 18 proposes a polynomial time algorithms for the special case of finding the optimal buffer size, both when considering static and dynamic resource allocation.…”

Section: Related Workmentioning

confidence: 99%

A comparative study of black‐box optimization heuristics for online tuning of high performance computing I/O accelerators

Robert

Zertal²,

Vaumourin

et al. 2021

Concurrency and Computation

View full text Add to dashboard Cite

High performance computing (HPC) applications' behaviors rely on highly configurable software environments and hardware devices. Finding their optimal parametrization is a complex task, as the size of their parametric space and the non-linear behavior of HPC systems make hand-tuning, theoretical modeling or exhaustive sampling unsuitable in most cases. In this article, we propose an online auto-tuner that relies on black-box optimization to find the optimal parametrization of input/output (I/O) accelerators for a given HPC application in a limited number of iterations, without making any assumption on the behavior of the tuned system. As many heuristics are available in the literature, we need to guarantee the quality of the tuning by selecting the most appropriate one. To do so, we provide a comparative study of the efficiency of three heuristics applied to tuning two I/O accelerators developed by the Atos company: a pure software accelerator (small read optimizer) and a mixed hardware-software one (smart burst buffer). To select the most efficient heuristic for our use case, we define several new metrics to evaluate the quality of an auto-tuner, in an online and offline settings. We find that genetic algorithms provide a faster convergence rate and a faster computation time but surrogate models provide a better score in terms of both distance to the optimum and trajectory stability. Overall, the obtained results show that auto-tuning heuristics improve the execution time of applications used conjointly with both SRO and SBB accelerators.

show abstract

“…More relevant to our study is to decide how burst buffers should be shared across the applications. We refer to [48,49] for an interesting discussion of various possible strategies, ranging from the privatization of the burst buffers to given compute nodes and/or given applications, to the flat sharing of every resource.…”

Section: Related Workmentioning

confidence: 99%

Checkpointing Strategies for Shared High-Performance Computing Platforms

Hérault

Robert

Bouteiller

et al. 2019

IJNC

View full text Add to dashboard Cite

Input/output (I/O) from various sources often contend for scarcely available bandwidth. For example, checkpoint/restart (CR) protocols can help to ensure application progress in failureprone environments. However, CR I/O alongside an application's normal, requisite I/O can increase I/O contention and might negatively impact performance. In this work, we consider different aspects (system-level scheduling policies and hardware) that optimize the overall performance of concurrently executing CR-based applications that share I/O resources. We provide a theoretical model and derive a set of necessary constraints to minimize the global waste on a given platform. Our results demonstrate that Young/Daly's optimal checkpoint interval, despite providing a sensible metric for a single, undisturbed application, is not sufficient to optimally address resource contention at scale. We show that by combining optimal checkpointing periods with contention-aware system-level I/O scheduling strategies, we can significantly improve overall application performance and maximize the platform throughput. Finally, we evaluate how specialized hardware, namely burst buffers, may help to mitigate the I/O contention problem. Overall, these results provide critical analysis and direct guidance on how to design efficient, CR ready, large-scale platforms without a large investment in the I/O subsystem.

show abstract

What Size Should Your Buffers to Disks be?

Cited by 13 publications

References 22 publications

Sizing and Partitioning Strategies for Burst-Buffers to Reduce IO Contention

Sizing and Partitioning Strategies for Burst-Buffers to Reduce IO Contention

A comparative study of black‐box optimization heuristics for online tuning of high performance computing I/O accelerators

Checkpointing Strategies for Shared High-Performance Computing Platforms

Contact Info

Product

Resources

About