When Server Joins INA: The Resource-Aware Repair Acceleration for Erasure-Coded Storage Systems
IEEE Transactions on Parallel and Distributed Systems · 2026-10 · 37 (10) · 2122-2136
DOI: 10.1109/tpds.2026.3704880 ↗Kurz gefasst
Fachartikel in der Zeitschrift IEEE Transactions on Parallel and Distributed Systems, 2026-10. Verfasst von Geyao Cheng und 5 weiteren Autor:innen. Die Arbeit nennt 51 Literaturangaben. Gefördert von: Guangdong Major Project of Basic Research, National Natural Science Foundation of China, Research Funding of NUDT.
Abstract (Originalsprache)
In the erasure-coded storage systems, multiple related blocks have to be retrieved from other surviving nodes to repair a failed block. This incurs significant communication overhead with the surging scale of distributed storage systems. To mitigate the bandwidth bottleneck, in-network repair (INR) has emerged as a promising transport paradigm, which migrates the aggregation operations from the repair node to the programmable hardware, such as Intel Tofino switches. However, due to the limited on-chip memory size of these switches, the INR can degrade to the most primitive incast-type transmission, leading to massive traffic volume and hindered repair throughput. While we notice that, there are spare CPU cores in the storage servers that can be leveraged as alternative computing resources. With this intuition, we propose SINA, a Server-assisted In-Network repair Acceleration framework in this paper, which leverages the spare servers to assist aggregation operations when the programming switches’ memory size is scarce for failure repair. By carefully adjusting the aggregation modes as well as allocating the computing and bandwidth resources, SINA accelerates failure repair without impacting running applications. We prove that solving the SINA problem is NP-hard and then provide efficient heuristic algorithms with complexity analyses. For all we know, this is the first work exploring spare servers for assisting the memory-scarce INR in erasure-coded storage systems. We have implemented SINA on an FPGA-based prototype system and also conducted large-scale simulations. The experimental results show that SINA can ensure fault tolerance and accelerate failure repair by 6.5× compared to the conventional methods.
Förderung
- Guangdong Major Project of Basic Research
- National Natural Science Foundation of China
- Research Funding of NUDT
- Major Key Project of Peng Cheng Laboratory
Zitieren
APA
Cheng, G., Xia, J., Fan, H., Liu, F., Mi, H., & Guo, D. (2026). When Server Joins INA: The Resource-Aware Repair Acceleration for Erasure-Coded Storage Systems. IEEE Transactions on Parallel and Distributed Systems, 37(10), 2122-2136. https://doi.org/10.1109/tpds.2026.3704880
Vancouver
Cheng G, Xia J, Fan H, Liu F, Mi H, Guo D. When Server Joins INA: The Resource-Aware Repair Acceleration for Erasure-Coded Storage Systems. IEEE Transactions on Parallel and Distributed Systems. 2026;37(10):2122-2136. doi:10.1109/tpds.2026.3704880
BibTeX
@article{Cheng2026,
author = {Cheng, Geyao and Xia, Junxu and Fan, Hao and Liu, Fengzeng and Mi, Haibo and Guo, Deke},
title = {When Server Joins INA: The Resource-Aware Repair Acceleration for Erasure-Coded Storage Systems},
journal = {IEEE Transactions on Parallel and Distributed Systems},
year = {2026},
volume = {37},
number = {10},
pages = {2122--2136},
doi = {10.1109/tpds.2026.3704880}
}
EUVIMED