55 lines
3.1 KiB
Markdown
55 lines
3.1 KiB
Markdown
---
|
|
id: "bug-1944"
|
|
status: "fixed"
|
|
status_updated: "2026-10-10"
|
|
status_note: ""
|
|
component: "scheduler"
|
|
severity: "wrong-result"
|
|
timestamp: "2026-10-10T08:14:56Z"
|
|
last_seen: "2026-10-10T08:14:56Z"
|
|
occurrences: 1
|
|
files: ["src/scheduler/src/core_events.c", "src/scheduler/src/schedobj.c"]
|
|
tags: ["scheduler", "queue", "priority"]
|
|
related: []
|
|
fix_commits: ["7a9d75d6"]
|
|
gitea_issue: 375
|
|
gitea_synced_status: "fixed"
|
|
gitea_synced_hash: "69bd8ab4a2cc"
|
|
---
|
|
|
|
# bug-1944: xce_scheduler gives a job moved to another queue the queue's priority, so it jumps ahead
|
|
|
|
## Description
|
|
**Observed:** (source reading, `.kiro/reviews/project-gap-review-2026-10-10.md` (P0); to be confirmed by a failing test before the fix)
|
|
New jobs get priority 0 (`schedobj.c:214`, `core_events.c:353`), but on a queue change the job's priority becomes the queue's priority (`core_events.c:1360`, `schedobj.c:684`); the picker sorts by job priority within a queue (`plugins/builtin_fcfs.c:213`).
|
|
|
|
**Expected:** Moving a job does not change its job priority.
|
|
**Impact:** Wrong ordering: moved jobs overtake jobs submitted earlier to the target queue.
|
|
**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. Two jobs pending in queue B; move a third job from A to B (queue priority > 0).
|
|
2. The moved job is picked first.
|
|
- 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
|
|
On a queue change both `xce_core_push_events` (`JOB_UPDATE`, `core_events.c`) and `sched_update_existing_job` (`schedobj.c`) set `job->priority = q->priority`; new jobs and snapshot-restored jobs get 0, so moved jobs jumped ahead and live/replay disagreed. Pre-existing.
|
|
|
|
## Fix
|
|
Both sites now only re-point `job->queue` on a move; queue priority still orders the queues. `test_job_update.c`, which asserted the old behaviour, now expects the job's own priority.
|
|
|
|
## Verification
|
|
- Test: `test_moved_job_keeps_its_priority_and_place` (`src/scheduler/testsuite/test_job_move_priority.c`): RED `:208: FAIL: Expected 0 Was 10` (and `:214: Expected 2 Was 1` for the ordering), GREEN after; `test_update_existing_job_mutates_in_place` RED `Expected 1 Was 5` with the corrected expectation, GREEN after.
|
|
- 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: `7a9d75d6` (on `scheduler-framework`; worker-branch original in the merge)
|
|
- Review: `.kiro/reviews/project-gap-review-2026-10-10.md`
|