A cleanup.run parked on startup readiness can start a run after a mid-startup stop #1001

Activity

1. V3RON

   added

   bug:newReported. An agent may triage it (rule 8), and nothing more.Reported. An agent may triage it (rule 8), and nothing more.

   on Oct 25, 2026
2. V3RON

   self-assigned this

   on Oct 25, 2026
3. V3RON commented on Oct 25, 2026

   

   V3RON

   on Oct 25, 2026

   Last edited by V3RON

   ContributorAuthor

   More actions

   Reproduction

   Defect, reproduced as the issue describes. A cleanup.run parked on readiness runs after stop() has finished, and shuts down a device.

   src/daemon/server.test.ts, "DaemonServer startup readiness" › "refuses a cleanup.run parked on startup with DAEMON_STOPPING once a stop lands mid-converge, and shuts down no device" (branch bug/1001-repro). The test seeds an idle ready device, parks cleanup.run {rule: "idle-shutdown"} through DaemonServer.dispatch (the HTTP seam), awaits stop(), then lets convergence resolve.

   pnpm exec vitest run --project unit src/daemon/server.test.ts -t "refuses a cleanup.run parked on startup"
   
   AssertionError: promise resolved "[ { action: 'shutdown', …(3) } ]" instead of rejecting
    ❯ src/daemon/server.test.ts:1537  await expect(parked).rejects.toMatchObject({ code: "DAEMON_STOPPING" })
   

   Root cause

   src/daemon/server.ts:609 (#awaitReady) is the cause. It resolves when convergence succeeds and never checks #stopping, so runDispatch (src/daemon/dispatch.ts:327) goes on to the handler. The only stopping gates are at the entry points (server.ts:425 in dispatch(), :819 in #dispatchLine), and a parked request has already passed them. The symptom shows in CleanupReaper.run (src/core/reaper.ts:76): it starts a run after reaper.settle() (server.ts:642) has already taken its snapshot. This affects every parked operation, not just cleanup: lease.request, device.exec and nuke.run would also run on a stopped daemon. On the socket path the client never gets an answer, because stop() closes the connection, but the work still runs.

   Simplest fix

   At the end of #awaitReady, add if (this.#stopping) throw new DispatchError("DAEMON_STOPPING", "Daemon is stopping"). I tried this locally: the test passes, and so do all 910 tests in src/daemon/ and src/http/.

   Alternatives rejected

   • Make CleanupReaper.run refuse after dispose(): this fixes cleanup only, and other parked operations still run.
   • Make stop() await #readyPromise before it settles: a slow or hung convergence would then block daemon stop.

   Risk

   • src/daemon/AGENTS.md Readiness should add "a stop during converge → parked requests get DAEMON_STOPPING".
   • The convergence-failed path should still answer DAEMON_STARTUP_FAILED. The catch runs before #stopping is checked, so it should, but a reviewer should confirm it.

   Written by an agent.
4. V3RON

   removed their assignment

   on Oct 25, 2026
5. V3RON

   added

   bug:readyAn agent may fix it.An agent may fix it.

   and removed

   bug:newReported. An agent may triage it (rule 8), and nothing more.Reported. An agent may triage it (rule 8), and nothing more.

   on Oct 25, 2026
6. V3RON

   self-assigned this

   on Oct 25, 2026
7. V3RON

   closed this as completedin #1003on Oct 25, 2026
8. V3RON commented on Oct 25, 2026

   

   V3RON

   on Oct 25, 2026

   ContributorAuthor

   More actions

   Delivery run

   Merged: #1003 (#1001)

   Written by an agent.