{
  "slug": "sfs",
  "term": "SFS",
  "aliases": [],
  "definition": "SFS is a research log-structured filesystem that groups file blocks by estimated update likelihood before writing segments.",
  "note": "Min et al.'s FAST 2012 prototype modifies NILFS2. Grouping aims to leave nearly empty victims and nearly full cold segments; hotness uses update count and recency. Its write-cost metric includes bandwidth-weighted cleaning reads as well as writes. [sfs2012]",
  "status": "research",
  "advantages": [],
  "limitations": [
    {
      "text": "The paper's performance and erase results are historical experiments and simulations, not current SSD guarantees.",
      "citationIds": [
        "sfs2012"
      ]
    }
  ],
  "related": [
    "garbage-collection-host",
    "f2fs"
  ],
  "supersedes": [],
  "supersededBy": [],
  "firstSeen": null,
  "lastSeen": null,
  "citations": [
    {
      "id": "sfs2012",
      "title": "Min, Kim, Cho, Lee and Eom. SFS: Random Write Considered Harmful in Solid State Drives. FAST 2012, sections 3.1-3.6 and 4.1-4.4.",
      "url": "https://www.usenix.org/system/files/conference/fast12/min.pdf",
      "accessed": "2026-11-14"
    },
    {
      "id": "hesela",
      "title": "Hesela: lifetime grouping and the cost of cleaning",
      "url": "https://hesela.com/analyses/nand-erase-block-size-gc-cost/#sfs",
      "accessed": "2026-11-14"
    }
  ],
  "updated": "2026-11-14",
  "provenance": [
    "Primary-source synthesis. No SFS prototype or benchmark reproduced."
  ]
}