2016
DOI: 10.1109/tpds.2015.2485980
|View full text |Cite
|
Sign up to set email alerts
|

Using Formal Grammars to Predict I/O Behaviors in HPC: The Omnisc'IO Approach

Abstract: The increasing gap between the computation performance of post-petascale machines and the performance of their I/O subsystem has motivated many I/O optimizations including prefetching, caching, and scheduling. In order to further improve these techniques, modeling and predicting spatial and temporal I/O patterns of HPC applications as they run has become crucial. In this paper we present Omnisc'IO, an approach that builds a grammar-based model of the I/O behavior of HPC applications and uses it to predict when… Show more

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
3
1
1

Citation Types

0
9
0

Year Published

2018
2018
2024
2024

Publication Types

Select...
4
2
1

Relationship

0
7

Authors

Journals

citations
Cited by 22 publications
(9 citation statements)
references
References 26 publications
0
9
0
Order By: Relevance
“…For representing symbolic sequences, context-free grammars have been widely used in text compression, natural language processing, music processing, and macromolecular sequence modeling [11]. A context-free grammar provides a simple and mathematically precise mechanism for describing the methods by which phrases in language are built from smaller blocks, capturing the block structure of sentences in a natural way.…”
Section: Intra-process Compressionmentioning
confidence: 99%
“…For representing symbolic sequences, context-free grammars have been widely used in text compression, natural language processing, music processing, and macromolecular sequence modeling [11]. A context-free grammar provides a simple and mathematically precise mechanism for describing the methods by which phrases in language are built from smaller blocks, capturing the block structure of sentences in a natural way.…”
Section: Intra-process Compressionmentioning
confidence: 99%
“…Based on that, the applications are scheduled avoiding I/O conflicts. Other works that use models to predict and avoid the application I/O interference are [1,10,26].…”
Section: Related Workmentioning
confidence: 99%
“…• Their number of iterations n i is chosen uniformly at random between 250 and 1000. • w i is chosen uniformly at random in [10,100].…”
Section: Synthetic Workload Generationmentioning
confidence: 99%
“…Characterizing I/O patterns is key as it allows the developers to identify performance bottlenecks, and allows the system administrator to better configure the platforms. A complementary path is to predict I/O performances during execution [4]. Such instrumentation efforts allow for a better use of the scarce communication resources.…”
Section: Related Workmentioning
confidence: 99%