---
id: "bug-1939"
status: "fixed"
status_updated: "2026-10-10"
status_note: ""
component: "scheduler"
severity: "wrong-result"
timestamp: "2026-10-10T08:14:43Z"
last_seen: "2026-10-10T08:14:43Z"
occurrences: 1
files: ["src/scheduler/src/resource.c", "src/scheduler/src/core_session.c", "src/scheduler/src/plugins/parjob_allocator.c"]
tags: ["scheduler", "host", "admin-control"]
related: []
fix_commits: ["1668a439"]
gitea_issue: 370
gitea_synced_status: "fixed"
gitea_synced_hash: "aac8b115c3bb"
---

# bug-1939: xce_scheduler dispatches jobs to hosts closed by an admin

## Description
**Observed:** (source reading, `.kiro/reviews/project-gap-review-2026-10-10.md` (P0); to be confirmed by a failing test before the fix)
Host close sets `XCE_WORKER_CLOSED|XCE_WORKER_CLOSED_BY_ADMIN` (`src/controller/worker.c:1015-1024`), but no scheduler code checks either bit (only unavailable/reconciling hosts are skipped: `core_session.c:136`, `resource.c:704,887`); the parallel allocator's bad-host check is empty (`plugins/parjob_allocator.c:40-48`).

**Expected:** A closed host gets no new jobs; running jobs are unaffected.
**Impact:** Admin host control silently ignored: closing a host for maintenance does not stop new jobs landing on it.
**Found by:** review: `.kiro/reviews/project-gap-review-2026-10-10.md` (P0)
## Environment
- OS: no /etc/os-release; Docker container; NixOS kernel; kernel Linux 6.18.33 x86_64
- libc: glibc 2.42-47 (interpreter /nix/store/wb6rhpznjfczwlwx23zmdrrw74bayxw4-glibc-2.42-47/lib/ld-linux-x86-64.so.2)
- Toolchain: gcc (GCC) 15.2.0, meson 1.9.1
- Build: asan (xce.build)
- Code: scheduler-framework @ 94472ee3-dirty
- Deployment: any (found by source reading; the failing test that confirms it is named in Verification)

## Reproduction
1. Close a host (`xcecli hosts close <h>`), submit jobs that fit it.
2. Jobs are dispatched to the closed host.
- Frequency: deterministic (by code reading; confirmed by the RED test of the fix)
- Evidence: the code sites quoted in Description; no logs kept.
## Root cause
`sched_host_iter_next()` (`src/scheduler/src/host_iter.c`), the candidate source of both seqjob and parjob allocators, filtered only on select and free resources, never on `XCE_WORKER_CLOSED`/`CLOSED_BY_ADMIN`; parjob's `skip_bad_worker_` is unused. Pre-existing.

## Fix
`host_iter.c`: a host with either closed bit is skipped before its trace scope opens, so no new placement (serial or parallel, live or sim) lands on it; running jobs and accounting are unchanged. No new pend-reason code.

## Verification
- Test: `test_admin_closed_host_gets_no_new_jobs` (`src/scheduler/testsuite/test_host_admin_close.c`): RED `:200: FAIL: Expected 0 Was 3 (the admin-closed host must get no new usage)`, GREEN after (serial and parallel, both fixture shapes).
- Suites: `xce-build.sh all asan xce.build --incremental` on the merged branch, 2026-10-10: all C suites Ok 155 / Fail 0, frontend Go unit tests ok, xceadmin tests ok; apiserver `CGO_ENABLED=0 go test ./...` (incl. middleware, router, cmd) ok
- Live: n/a: unit/fixture-level verification only; no live deployment was run (follow-up)

## References
- Fix commit: `1668a439` (on `scheduler-framework`; worker-branch original in the merge)
- Review: `.kiro/reviews/project-gap-review-2026-10-10.md`