An On-Line Performance Introspection Framework for Task-Based Runtime Systems

Aguilar, Xavier; Jordan, Herbert; Heller, Thomas; Hirsch, A.A.; Fahringer, Thomas; Laure, Erwin

doi:10.1007/978-3-030-22734-0_18

Cited by 4 publications

(3 citation statements)

References 14 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In this aspect, it is more general than the proposed approach. However, our approach is based on internal monitoring of the MPI runtime providing more precise information in terms of exchanged data In [12], the authors proposed an introspection library for a task-based runtime system which is a different progralming model than the message passing one of MPI.…”

Section: Related Workmentioning

confidence: 99%

An introspection monitoring library to improve MPI communication time

Jeannot

Sartori

2023

J Supercomput

View full text Add to dashboard Cite

In this paper we describe how to improve communication time of MPI parallel applications with the use of a library that enables to monitor MPI applications and allows for introspection (the program itself can query the state of the monitoring system). Based on previous work, this library is able to see how collective communications are decomposed into point-to-point messages. It also features monitoring sessions that allow suspending and restarting the monitoring, limiting it to specific portions of the code. Experiments show that the monitoring overhead is very small and that the proposed features allow for dynamic and efficient rank reordering enabling up to 2-time reduction of communication parts of some program.

show abstract

Section: Related Workmentioning

confidence: 99%

An introspection monitoring library to improve MPI communication time

Jeannot

Sartori

2023

J Supercomput

View full text Add to dashboard Cite

show abstract

“…In [29], inter and intra-application resource contention are classified as sources of performance variability. On the runtime systems' side, a real-time monitoring component that assists the runtime scheduler in its decision-making process and adaptation is developed as part of the AllScale toolchain [1]. By collecting information across a cluster, the AllScale runtime can tune parameters such as thread counts or DVFS.…”

Section: Related Workmentioning

confidence: 99%

“…Continuous introspection is a step towards handling more dynamic environments. The AllScale runtime [1] collects runtime performance information and feeds it into an optimiser module that can tune thread numbers and use DVFS to optimize the execution. However, this runtime targets homogeneous architectures, and the solution is not applicable to systems exhibiting dynamic per-core performance asymmetry (such as interference).…”

Section: Introductionmentioning

confidence: 99%

Scheduling Task-parallel Applications in Dynamically Asymmetric Environments

Chen

Soomro

Abduljabbar

et al. 2020

Preprint

View full text Add to dashboard Cite

Shared resource interference is observed by applications as dynamic performance asymmetry. Prior art has developed approaches to reduce the impact of performance asymmetry mainly at the operating system and architectural levels. In this work, we study how application-level scheduling techniques can leverage moldability (i.e. flexibility to work as either single-threaded or multithreaded task) and explicit knowledge on task criticality to handle scenarios in which system performance is not only unknown but also changing over time. Our proposed task scheduler dynamically learns the performance characteristics of the underlying platform and uses this knowledge to devise better schedules aware of dynamic performance asymmetry, hence reducing the impact of interference. Our evaluation shows that both criticality-aware scheduling and parallelism tuning are effective schemes to address interference in both shared and distributed memory applications.CCS Concepts • Computer systems organization → Multicore architectures.

show abstract

Improving MPI Application Communication Time with an Introspection Monitoring Library

Jeannot

Sartori

2020

2020 IEEE International Parallel and Distributed Processing Symposium Workshops (IPDPSW)

View full text Add to dashboard Cite

In this report we describe how to improve communication time of MPI parallel applications with the use of a library that enables to monitor MPI applications and allows for introspection (the program itself can query the state of the monitoring system). Based on previous work, this library is able to see how collective communications are decomposed into point-to-point messages. It also features monitoring sessions that allow suspending and restarting the monitoring, limiting it to specific portions of the code. Experiments show that the monitoring overhead is very small and that the proposed features allow for dynamic and efficient rank reordering enabling up to 2-time reduction of communication parts of some program.

show abstract

An On-Line Performance Introspection Framework for Task-Based Runtime Systems

Cited by 4 publications

References 14 publications

An introspection monitoring library to improve MPI communication time

An introspection monitoring library to improve MPI communication time

Scheduling Task-parallel Applications in Dynamically Asymmetric Environments

Improving MPI Application Communication Time with an Introspection Monitoring Library

Contact Info

Product

Resources

About