I/O-Aware Batch Scheduling for Petascale Computing Systems

Zhou, Zhou; Xu, Yang; Zhao, Dongfang; Rich, Paul M.; Tang, Wei; Wang, Jia; Lan, Zhiling

doi:10.1109/cluster.2015.45

Cited by 39 publications

(26 citation statements)

References 30 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…As a result, I/O awareness within batch schedulers in use for today's large centers remains largely rudimentary (e.g., Moab [2] can only hold dependent jobs when a file system is marked as being down). Our work tackles the I/O-aware scheduling problem within batch schedulers and complements existing work that manages the I/O contention problem by coordinating and optimizing the I/O performed at runtime [9,34,10,36]. While these runtime techniques have shown promising results at maximizing the use of the PFS and minimizing I/O contention, they are ultimately constrained by the I/O requests of the specific running applications.…”

Section: Introductionmentioning

confidence: 94%

“…More recent e↵orts on runtime scheduling of I/O contention include [35,10,9,36]. Specifically, work in [35] addresses I/O contention with a PFS access controller: the controller provides a single application exclusive access to the PFS for a time window.…”

Section: Related Workmentioning

confidence: 99%

“…Work in [10] analyzes the impact of congestion on applications' I/O bandwidth and assesses a variety of runtime techniques designed to either maximize system efficiency or minimize application slowdown. Finally, work in [36] presents an I/O-aware scheduling framework that coordinates I/O requests at runtime on petascale computing systems driven by either user-oriented metrics or system performance. Our work aligns with and complements these four runtime methods by targeting next-generation largescale systems and defining a method that operates at batch schedule time to mitigate I/O contention.…”

Section: Related Workmentioning

confidence: 99%

See 2 more Smart Citations

Scalable I/O-Aware Job Scheduling for Burst Buffer Enabled HPC Clusters

Herbein

Ahn

Lipari

et al. 2016

Proceedings of the 25th ACM International Symposium on High-Performance Parallel and Distributed Computing

View full text Add to dashboard Cite

The economics of flash vs. disk storage is driving HPC centers to incorporate faster solid-state burst bu↵ers into the storage hierarchy in exchange for smaller parallel file system (PFS) bandwidth. In systems with an underprovisioned PFS, avoiding I/O contention at the PFS level will become crucial to achieving high computational e ciency. In this paper, we propose novel batch job scheduling techniques that reduce such contention by integrating I/O awareness into scheduling policies such as EASY backfilling. We model the available bandwidth of links between each level of the storage hierarchy (i.e., burst bu↵ers, I/O network, and PFS), and our I/O-aware schedulers use this model to avoid contention at any level in the hierarchy. We integrate our approach into Flux, a next-generation resource and job management framework, and evaluate the e↵ectiveness and computational costs of our I/O-aware scheduling. Our results show that by reducing I/O contention for underprovisioned PFSes, our solution reduces job performance variability by up to 33% and decreases I/O-related utilization losses by up to 21%, which ultimately increases the amount of science performed by scientific workloads.

show abstract

Section: Introductionmentioning

confidence: 94%

Section: Related Workmentioning

confidence: 99%

Section: Related Workmentioning

confidence: 99%

See 1 more Smart Citation

Scalable I/O-Aware Job Scheduling for Burst Buffer Enabled HPC Clusters

Herbein

Ahn

Lipari

et al. 2016

Proceedings of the 25th ACM International Symposium on High-Performance Parallel and Distributed Computing

View full text Add to dashboard Cite

show abstract

“…Closer to this work, online schedulers for HPC systems were developed such as our previous work [14], the study by Zhou et al [31], and a solution proposed by Dorier et al [11]. In [11], the authors investigate the interference of two applications and analyze the benefits of interrupting or delaying either one in order to avoid congestion.…”

Section: Related Workmentioning

confidence: 99%

Periodic I/O Scheduling for Super-Computers

Aupy

Gainaru

Fèvre

2017

Lecture Notes in Computer Science

View full text Add to dashboard Cite

With the ever-growing need of data in HPC applications, the congestion at the I/O level becomes critical in super-computers. Architectural enhancement such as burst-buffers and pre-fetching are added to machines, but are not sufficient to prevent congestion. Recent online I/O scheduling strategies have been put in place, but they add an additional congestion point and overheads in the computation of applications.In this work, we show how to take advantage of the periodic nature of HPC applications in order to develop efficient periodic scheduling strategies for their I/O transfers. Our strategy computes once during the job scheduling phase a pattern where it defines the I/O behavior for each application, after which the applications run independently, transferring their I/O at the specified times. Our strategy limits the amount of I/O congestion at the I/O node level and can be easily integrated into current job schedulers. We validate this model through extensive simulations and experiments by comparing it to state-of-the-art online solutions, showing that not only our scheduler has the advantage of being de-centralized and thus overcoming the overhead of online schedulers, but also that it performs better than these solutions, improving the application dilation up to 13% and the maximum system efficiency up to 18%.

show abstract

“…I/O aware scheduling [23] for example, can control the status of jobs on the fly during execution based on run-time monitoring of system state and I/O activities. Another example is the COBALT scheduler [19].…”

Section: Application Areasmentioning

confidence: 99%

Seastar

Weidner

Barker

Atkinson

2017

Proceedings of the 7th International Workshop on Runtime and Operating Systems for Supercomputers ROSS 2017

View full text Add to dashboard Cite

A large number of 2nd generation high-performance computing applications and services rely on adaptive and dynamic architectures and execution strategies to run efficiently, resiliently, and at scale on today's HPC infrastructures. They require information about applications and their environment to steer and optimize execution. We define this information as telemetry data.Current HPC platforms do not provide the infrastructure, interfaces and conceptual models to collect, store, analyze, and access such data. Today, applications depend on application and platform specific techniques for collecting telemetry data; introducing significant development overheads that inhibit portability and mobility. The development and adoption of adaptive, context-aware strategies is thereby impaired. To facilitate 2nd generation applications, more efficient application development, and swift adoption of adaptive applications in production, a comprehensive framework for telemetry data management must be provided by future HPC systems and services.We introduce Seastar, a conceptual model and a software framework to collect, store, analyze, and exploit streams of telemetry data generated by HPC systems and their applications. We show how Seastar can be integrated with HPC platform architectures and how it enables common application execution strategies.

show abstract

I/O-Aware Batch Scheduling for Petascale Computing Systems

Cited by 39 publications

References 30 publications

Scalable I/O-Aware Job Scheduling for Burst Buffer Enabled HPC Clusters

Scalable I/O-Aware Job Scheduling for Burst Buffer Enabled HPC Clusters

Periodic I/O Scheduling for Super-Computers

Seastar

Contact Info

Product

Resources

About