KafkaDirect: Zero-copy Data Access for Apache Kafka over RDMA Networks
Summary: KafkaDirect extends Apache Kafka with RDMA-based zero-copy data paths, bypassing buffers for producers, replicas, and consumers. Offloading to RDMA yields up to 9× throughput and 4×–50× latency reductions, vs. TCP/IP Kafka. (summarized by gpt-5-nano on Feb 09 2026)
Incoming Non-self Citations Over Time
No non-self incoming citations found for this paper in this database.
Authors
- 1. Konstantin Taranov (ETH Zurich)
- 2. Steve Byan (Oracle)
- 3. Virendra Marathe (Oracle)
- 4. Torsten Hoefler (ETH Zurich)
BibTeX Citation
@inproceedings{taranov_sigmod22,
title = {{KafkaDirect: Zero-copy Data Access for Apache Kafka over RDMA Networks}},
author = {Taranov, Konstantin and Byan, Steve and Marathe, Virendra and Hoefler, Torsten},
series = {{SIGMOD} '22},
booktitle = {Proceedings of the {ACM} {SIGMOD} International Conference on Management of Data},
publisher = {Association for Computing Machinery},
doi = {10.1145/3514221.3526056},
url = {https://dl.acm.org/doi/10.1145/3514221.3526056},
year = {2022}
}
Incoming Citations (Sorted by Pagerank)
Showing 0 of 0 citing papers.
| Rank | Citing Paper | Year | Venue | Pagerank |
|---|
Previous
Page 1 / 1
Next
Outgoing Citations (Sorted by Pagerank)
Showing 8 of 8 cited papers.
Citations counted here include only citations to other VLDB/SIGMOD/CIDR/PODS papers in this database.
| Rank | Cited Paper | Year | Venue | Pagerank |
|---|---|---|---|---|
| 892 | Rack-Scale In-Memory Join Processing using RDMA | 2015 | SIGMOD | 0.00013376761 |
| 976 | The End of Slow Networks: It's Time for a Redesign | 2016 | VLDB | 0.00012864533 |
| 1,076 | High-Speed Query Processing over High-Speed Networks | 2016 | VLDB | 0.00012270109 |
| 1,117 | Accelerating Relational Databases by Leveraging Remote Memory and RDMA | 2016 | SIGMOD | 0.00012108907 |
| 1,342 | Designing Distributed Tree-based Index Structures for Fast RDMA-capable Networks | 2019 | SIGMOD | 0.00011098147 |
| 1,345 | The End of a Myth: Distributed Transactions Can Scale | 2017 | VLDB | 0.00011090483 |
| 3,243 | Rethinking Database High Availability with RDMA Networks | 2019 | VLDB | 7.6051655e-05 |
| 5,419 | Building a Replicated Logging System with Apache Kafka | 2015 | VLDB | 6.2255013e-05 |
Previous
Page 1 / 1
Next
Semantically Similar Papers
| # | Overall Rank | Paper | Year | Venue |
|---|---|---|---|---|
| 1 | 3,243 | Rethinking Database High Availability with RDMA Networks | 2019 | VLDB |
| 2 | 7,814 | Rethinking Stateful Stream Processing with RDMA | 2022 | SIGMOD |
| 3 | 9,562 | Kora: A Cloud-Native Event Streaming Platform For Kafka | 2023 | VLDB |
| 4 | 8,596 | Zero-sided RDMA: Network-driven Data Shuffling for Disaggregated Heterogeneous Cloud DBMSs | 2024 | SIGMOD |
| 5 | 11,018 | Ursa: A Lakehouse-Native Data Streaming Engine for Kafka | 2025 | VLDB |
| 6 | 9,668 | D-RDMA: Bringing Zero-Copy RDMA to Database Systems | 2022 | CIDR |
| 7 | 12,006 | Query-able Kafka: An agile data analytics pipeline for mobile wireless networks | 2017 | VLDB |
| 8 | 5,419 | Building a Replicated Logging System with Apache Kafka | 2015 | VLDB |
| 9 | 4,389 | Consistency and Completeness: Rethinking Distributed Stream Processing in Apache Kafka | 2021 | SIGMOD |
| 10 | 11,763 | Rethinking Message Brokers on RDMA and NVM | 2020 | SIGMOD |