myxce/.wolf/bugs/bug-1937.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-1937	fixed	2026-10-10		worker	wrong-result	2026-10-10T08:14:38Z	2026-10-10T08:14:38Z	1	src/frontend/apiserver/controllers/jobs.go src/worker/src/xcecd_exdata.c	apiserver worker exdata file-transfer		b9747ad2	368	fixed	9d09c2947036

bug-1937: apiserver writes job file transfers under 'transfer_files' but the worker parses 'transfer_file'

Description

Observed: (source reading, .kiro/reviews/project-gap-review-2026-10-10.md (P0); to be confirmed by a failing test before the fix) The apiserver serialises transfers as extendDataMap["transfer_files"] (src/frontend/apiserver/controllers/jobs.go:1923) while the worker looks for:

else if (strcmp(key, "transfer_file") == 0) {   /* xcecd_exdata.c:115 */

Expected: Both sides use one key, so transfers in exdata are parsed. Impact: Wrong result once exdata reaches the worker: file transfers are silently ignored. 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. Read the two sites above; or unit-test parse_exdata() with the JSON the apiserver produces: transfer_info->ops stays empty.

• 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

parse_exdata() (src/worker/src/xcecd_exdata.c) matched only "transfer_file", the apiserver writes "transfer_files". Pre-existing.

Fix

xcecd_exdata.c: accepts "transfer_files" and still "transfer_file"; when both are present the first list is kept (no leak). Stage-in/out itself remains unimplemented (future spec).

Verification

• Test: test_apiserver_transfer_files_parsed (src/worker/testsuite/test_exdata_parse.c): RED test_exdata_parse.c:73: ... FAIL: Expected 2 Was 0, GREEN after (5 tests, 0 failures).
• 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: b9747ad2 (on scheduler-framework; worker-branch original in the merge)
• Review: .kiro/reviews/project-gap-review-2026-10-10.md