Adaptive Speculation for Efficient Internetware Application Execution in Clouds
- Submitting institution
-
The University of Leeds
- Unit of assessment
- 11 - Computer Science and Informatics
- Output identifier
- UOA11-1588
- Type
- D - Journal article
- DOI
-
10.1145/3093896
- Title of journal
- ACM Transactions on Internet Technology
- Article number
- 15
- First page
- -
- Volume
- 18
- Issue
- 2
- ISSN
- 1533-5399
- Open access status
- Compliant
- Month of publication
- January
- Year of publication
- 2018
- URL
-
-
- Supplementary information
-
-
- Request cross-referral to
- -
- Output has been delayed by COVID-19
- No
- COVID-19 affected output statement
- -
- Forensic science
- No
- Criminology
- No
- Interdisciplinary
- No
- Number of additional authors
-
5
- Research group(s)
-
E - DSS (Distributed Systems and Services)
- Citation count
- 7
- Proposed double-weighted
- No
- Reserve for an output with double weighting
- No
- Additional information
- The straggler problem is common in modern Cloud systems, i.e. a small number of slow tasks significantly impede parallel job completion. It is extremely difficult in practice to prevent stragglers or reduce the number of them since the stragglers are created by a large variety of and highly dynamic causes. This substantial piece of work presents an adaptive method for dynamically detecting and treating potential stragglers, thereby speeding up parallel job execution (up to 20%). Implemented into YARN and assessed against state-of-the-art approaches. Clear impact for researchers and cloud service providers (Keynote speech: IEEE AIMS 2018; Ricardo Bianchini (Microsoft) DOI:10.1145/3302424.3303973).
- Author contribution statement
- -
- Non-English
- No
- English abstract
- -
