{
  "slug": "crash-recovery",
  "term": "Crash-recovery model",
  "aliases": [
    "Crash recovery model",
    "Crash-restart model"
  ],
  "definition": "The crash-recovery model allows a process to stop and later resume under an explicit contract for which state survives the crash.",
  "note": "Unlike crash-stop, recovery permits the same process to return. In the 1998 model, identities are fixed, recovery is locally recognized, volatile state is lost, and explicitly saved stable state can survive. [act98]\n\nAn unstable process crashes and recovers infinitely often; a finite restart trace cannot establish that whole-execution classification. Fair-loss links do not promise delivery of each message. [act98]\n\nStable-state writes and message sends are separate actions. Recovery therefore needs an ordering contract, not merely persistent media. [act98]\n\nRaft's currentTerm, votedFor and log are persistent state; relevant changes precede RPC responses. Recovered participation still requires the protocol's rules. [raft-paper]\n\nThe author's errata qualify lastApplied: its lifetime must match the state machine's. A durable state machine needs durable applied progress; treating the index as always volatile is unsafe. [raft-errata]\n\nA valid local file checksum need not establish freshness. Paxos Made Live explicitly excludes old-state rollback from that check. Its separately ordered database-checksum comparison addresses replica divergence. Neither check alone restores voting authority after lost protocol state. [paxos-made-live]",
  "status": "unknown",
  "advantages": [],
  "limitations": [],
  "supersedes": [],
  "supersededBy": [],
  "firstSeen": null,
  "lastSeen": null,
  "updated": "2026-11-21",
  "related": [
    "stable-storage",
    "consensus",
    "raft",
    "crash-stop",
    "paxos"
  ],
  "citations": [
    {
      "id": "act98",
      "title": "Aguilera, Chen and Toueg. Failure Detection and Consensus in the Crash-Recovery Model. DISC 1998, pp. 231-245, sections 2.1-2.4.",
      "url": "https://www.microsoft.com/en-us/research/wp-content/uploads/1998/09/disc98_recovery.pdf",
      "accessed": "2026-11-20"
    },
    {
      "id": "raft-paper",
      "title": "Ongaro and Ousterhout. In Search of an Understandable Consensus Algorithm, extended version (2014), Figure 2 and sections 5.5-5.6.",
      "url": "https://raft.github.io/raft.pdf",
      "accessed": "2026-11-20"
    },
    {
      "id": "raft-errata",
      "title": "Diego Ongaro. Consensus: Bridging Theory and Practice, author errata for Figure 3.1 (Chapter 3), repository revision 0c10ea1b53d73bf6d163bdde8e4141a1cbd8f36a.",
      "url": "https://github.com/ongardie/dissertation/blob/0c10ea1b53d73bf6d163bdde8e4141a1cbd8f36a/README.md#chapter-3-basic-raft-algorithm",
      "accessed": "2026-11-20"
    },
    {
      "id": "applied-progress",
      "title": "Hesela. Constructed application-progress recovery controls: consistent recovery, replay and skipped effects.",
      "url": "https://hesela.dev/research/crash-recovery/README.md",
      "accessed": "2026-11-20"
    },
    {
      "id": "paxos-made-live",
      "title": "Chandra, Griesemer and Redstone. Paxos Made Live. PODC 2007, sections 5.1 and 6.2, footnotes 2-3; DOI 10.1145/1281100.1281103.",
      "url": "https://15799.courses.cs.cmu.edu/fall2013/static/papers/p398-chandra.pdf",
      "accessed": "2026-11-21"
    }
  ],
  "provenance": [
    "Primary-source model synthesis, not a protocol implementation, failure experiment or proof of the paper's consensus bounds."
  ]
}
