Improving the Performance of Heterogeneous Data Centers through Redundancy
- 7 June 2021
- journal article
- meeting abstracts
- Published by Association for Computing Machinery (ACM) in ACM SIGMETRICS Performance Evaluation Review
- Vol. 49 (1), 55-56
- https://doi.org/10.1145/3543516.3456274
Abstract
We analyze the performance of redundancy in a multi-type job and multi-type server system where PS is implemented. We characterize the stability condition, which coincides with that of a system where each job type only dispatches copies into its least-loaded servers, and those copies need to be fully served. We then investigate the impact of redundancy in the stability condition by comparing that to the stability condition of a non-redundant system in which a job arrival is routed to only one randomly selected compatible server. We observe that if server loads are sufficiently heterogeneous redundancy can considerably improve the stability region of the system.Keywords
Funding Information
- Agence Nationale de la Recherche (ANR-15-CE25-0004)
This publication has 5 references indexed in Scilit:
- Improving the Performance of Heterogeneous Data Centers through RedundancyProceedings of the ACM on Measurement and Analysis of Computing Systems, 2020
- A little redundancy goes a long way: Convexity in redundancy systemsPerformance Evaluation, 2019
- When Do Redundant Requests Reduce Latency?IEEE Transactions on Communications, 2015
- Low latency via redundancyPublished by Association for Computing Machinery (ACM) ,2013
- The tail at scaleCommunications of the ACM, 2013