2018
DOI: 10.1016/j.jpdc.2017.10.021
|View full text |Cite
|
Sign up to set email alerts
|

Harnessing sliding-window execution semantics for parallel stream processing

Abstract: According to the recent trend in data acquisition and processing technology, big data are increasingly available in the form of unbounded streams of elementary data items to be processed in real-time. In this paper we study in detail the paradigm of sliding windows, a well-known technique for approximated queries that update their results continuously as new fresh data arrive from the stream. In this work we focus on the relationship between the various existing sliding window semantics and the way the query p… Show more

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
1
1
1
1

Citation Types

0
4
0

Year Published

2018
2018
2022
2022

Publication Types

Select...
6
2
1

Relationship

2
7

Authors

Journals

citations
Cited by 13 publications
(4 citation statements)
references
References 36 publications
0
4
0
Order By: Relevance
“…e authors name this pa ern the "split/aggregate/join" pa ern. e work of Mencagli et al [105] propose a novel execution model, "agnostic worker" for multi-core shared memory environments. An agnostic worker parallelizes operators with window policies that require forward context (cf.…”
Section: Parallelization For General Streammentioning
confidence: 99%
“…e authors name this pa ern the "split/aggregate/join" pa ern. e work of Mencagli et al [105] propose a novel execution model, "agnostic worker" for multi-core shared memory environments. An agnostic worker parallelizes operators with window policies that require forward context (cf.…”
Section: Parallelization For General Streammentioning
confidence: 99%
“…Integrating key-based load splitting into Jarvis may further reduce network transfer costs by minimizing the number of output keys sent from the data source. Recent studies investigate assigning subsequences of records or windows to operator instances based on the currently assigned compute load in the host node [91]- [93]. These approaches do not assume that compute resources in each stream processor node are shared by multiple data sources (see Figure 4).…”
Section: Related Workmentioning
confidence: 99%
“…A large part of the articles that referred to languages for HPC do not include authors of the cHiPSet ICT COST Action but there are some exceptions like [39,40,43].…”
Section: )mentioning
confidence: 99%