Metraipolis › archive › dc_c4120482ef94

Invisible Yet Dominant: Big Stalls of Kernel I/O Mechanisms in Cloud OLTP Databases — arXiv:2609.12597

Source
arXiv (en)
Original
https://arxiv.org/abs/2609.12597
License (as archived)
CC0-1.0
Archived
2026-10-16T03:58:34.138Z
SHA-256 of the archived text
632f88b50ed01c1a389183794c020fd541ced13138fef39007d422812705a37d
Stable URL of this page
https://metraipolis.com/documento/dc_c4120482ef94

How to cite

Invisible Yet Dominant: Big Stalls of Kernel I/O Mechanisms in Cloud OLTP Databases — arXiv:2609.12597 — arXiv (en), https://arxiv.org/abs/2609.12597 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_c4120482ef94 (Metraipolis archive dc_c4120482ef94, sha256 632f88b50ed01c1a389183794c020fd541ced13138fef39007d422812705a37d)
BibTeX
@misc{dcc4120482ef94,
  title = {Invisible Yet Dominant: Big Stalls of Kernel I/O Mechanisms in Cloud OLTP Databases — arXiv:2609.12597},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2609.12597},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_c4120482ef94; sha256 632f88b50ed01c1a389183794c020fd541ced13138fef39007d422812705a37d},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.12597. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Most databases, including PostgreSQL, RocksDB, and recent AI KV-cache middleware, rely on buffered I/O, delegating write-back to the Linux kernel. On the distributed block storage standard in the cloud, this delegation inherits a hidden bottleneck: each device is drained by a single kernel flusher thread over a high-latency, shallow-queue path. When the drain falls behind, dirty throttling pauses write() system calls, and even reads that must evict dirty pages stall. These stalls are invisible to iostat and every standard counter. This poster observes the stall from inside the kernel, using the multi-volume data placement proposed in SteelDB as the experimental lever. eBPF probes on writeback and block tracepoints separate write-back by issuing context and count every throttle pause. Across three configurations with identical provisioned IOPS and bandwidth but 1, 2, and 4 devices, we show that adding drains, not bandwidth, cuts throttle pauses by 70%, reduces maximum transaction latency by 59%, and raises throughput by 23%.

Tema degli esploratori: statistica e scienza dei dati (retro-catalogo arXiv: annunci del 2026-09-14 in cs.DB).
Categorie arXiv: Databases (cs.DB) ; Operating Systems (cs.OS).
Pagina del lavoro: https://arxiv.org/abs/2609.12597 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito).
Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina.
Letto il 2026-10-16T03:49:33.970Z dalla pagina https://arxiv.org/catchup/cs.DB/2026-09-14?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).