myxce/.wolf/bugs/bug-1941.md

id	status	status_updated	status_note	component	severity	timestamp	last_seen	occurrences	files	tags	related	fix_commits	gitea_issue	gitea_synced_status	gitea_synced_hash
bug-1941	fixed	2026-10-10		controller	wrong-result	2026-10-10T08:14:48Z	2026-10-10T08:14:48Z	1	src/controller/job.c	controller job-control host-status		f244661b	372	fixed	a1c9a6520cb1

bug-1941: xce_controller refuses kill/stop/resume of running jobs on hosts whose status is not exactly OK

Description

Observed: (source reading, .kiro/reviews/project-gap-review-2026-10-10.md (P0); to be confirmed by a failing test before the fix)

/* FIXME: better method to check host status */
if (w->ep && w->host.status == 0) {      /* src/controller/job.c:2145-2156 */

Any other status bit (CLOSED_BY_ADMIN, BUSY, UNAVAIL...) makes the control request fail.

Expected: Jobs on a closed (or otherwise flagged but connected) host can be killed, stopped and resumed; a host that is not connected gives a clear error. Impact: Users and admins cannot kill jobs on a host closed for maintenance. 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. Run a job on host h, close h, kill the job.
2. The kill is refused.

• 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

job_ctrl() (src/controller/job.c) sent control only if w->host.status == 0 and returned -errno otherwise. Pre-existing FIXME.

Fix

job.c: new job_ctrl_to_host() sends whenever the worker is connected (endpoint present, not UNAVAIL), so CLOSED/CLOSED_BY_ADMIN/BUSY/locked hosts work; an unknown or unconnected host fails with -E_XCE_BAD_WORKER and an error log naming it.

Verification

• Test: test_jc_connected_flagged_host_is_controllable and test_jc_unconnected_host_fails_with_bad_worker (src/controller/testsuite/test_job_ctrl_hosts.c): RED :239: FAIL: Expected 0 Was -2 (status 0x100000) and Expected -10002 Was -22, GREEN after (asserts the decoded request in the worker send queue).
• 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: f244661b (on scheduler-framework; worker-branch original in the merge)
• Review: .kiro/reviews/project-gap-review-2026-10-10.md