Rongsong Shen
d26ff8870d
The buglog was one 27K-line JSON file whose fields covered only symptom,
root cause and fix, as free text: 2% of entries named the code version a
bug was found on, 12% the build profile, 36% a way to reproduce, and 58% of
the file was OpenWolf hook auto-detections ("Missing guard clause",
"Function not marked async" on .md files).
- .claude/skills/buglog/bug-record.md: the record standard. Eight
sections (Summary as the title, Description, Environment, Reproduction;
Root cause, Fix, Verification, References once fixed), each with the
content it must contain -- e.g. Environment needs the OS, the libc the
failing binary loads (from its ELF interpreter: glibc 2.42 in the nix
shell vs 2.40 on the host), toolchain, build profile and commit.
- .wolf/bugs/<id>.md: 500 records migrated from buglog.json; every field
verified to round-trip. They carry legacy: true and an explicit "not
recorded" where the old log had no such field; component derived from
file paths, severity from keywords (else unknown), fix_commits from
hashes in status_note/fix that git resolves (91 of 463 fixed).
- .wolf/buglog-archive.json: the 710 hook auto-detections, verbatim.
- .wolf/buglog.json is now an index generated from the records, keeping
the fields OpenWolf's hooks and `openwolf bug search` read.
- scripts/buglog.py rewritten: new (template + Environment collected from
the machine and --binary), check, status (refuses a status the record
doesn't support; fixed needs Root cause, Fix, Verification and
--commit hashes git resolves), note, recur, set, show, list, triage,
env, index. 29 unittest cases in scripts/tests/test_buglog.py.
- scripts/git-hooks/pre-commit runs `check --staged`: incomplete staged
records, a legacy record hand-changed to fixed without the fix-time
sections, a new record claiming legacy: true, or a stale index refuse
the commit.
- scripts/gitea-sync.py reads the records: issue body = the record's
sections plus linked fix commits (origin remote, GITEA_COMMIT_REPO to
override), and a body hash so an issue's text is refreshed when the
record changes.
- bug-1866 rewritten to the full standard as the worked example.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
3.5 KiB
| id | status | status_updated | status_note | component | severity | timestamp | last_seen | occurrences | files | tags | related | legacy | ||||||||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| bug-1809 | fixed | 2026-09-24 | scheduler | unknown | 2026-09-14T16:20:00Z | 2026-09-19T09:00:00Z | 1 |
|
|
|
true |
bug-1809: User-reported: live scheduler logs are flooded with WARN lines for every ordinary…
Description
User-reported: live scheduler logs are flooded with WARN lines for every ordinary XCE_CORE_EVENT_JOBS_PUSH resync redelivery: "xce_core_push_events JOBS_PUSH: sched_add_job failed for jobid (rc=-17), merging state into existing tracked job instead of discarding" (src/scheduler/src/core_events.c). rc=-17 is -EEXIST -- the expected, routine outcome of on_scheduler_join()'s full resync (bug-1734) or the array-push dedup race (bug-1714), already correctly handled by the bug-1734/bug-1738/bug-1739 merge logic in the same branch, not a functional defect.
Environment
Not recorded (migrated from buglog.json).
Reproduction
Not recorded (migrated from buglog.json).
Root cause
Not a defect -- a UX/observability request. sched_add_job() (core.c) wraps key_add()->kvtab_add() (lib/kv.c), whose only failure modes for the core->jobs table (KVMGR_KEY_INT, no keyops->assign) are -EEXIST (duplicate jobid) or -ENOMEM (allocation failure). The EEXIST case is therefore the near-universal, routine shape of this branch under a reconnect/resync storm (confirmed live in bug-1714's own investigation: 700+ of these WARNs fired from a single 8-reconnect burst) -- WARN was the wrong level for an outcome the code already handles correctly by design; it was originally added at WARN by bug-1734's own fix pass without considering steady-state log volume under the reconnect-storm conditions that same session's investigation had just captured.
Fix
FIXED (2026-09-14). Split the single WARN call in the JOBS_PUSH EEXIST branch (core_events.c) on add_rc: when add_rc == -EEXIST, log via a new DBG(...) macro (utils.h, wraps tlog_debug() -- DEBUG level, silent by default, visible with -vvv) instead of WARN -- preserves the message (still valuable forensic evidence per bug-1734/bug-1714) without flooding default-verbosity operational logs. Any OTHER negative add_rc (e.g. -ENOMEM) still logs at WARN, since that would be a genuine anomaly, not a routine resync artifact. The new macro is named DBG, not DEBUG, matching the existing TRACE/INFO/WARN/ERROR one-line wrapper convention in utils.h: a first attempt to name it DEBUG(...) collided with the build's own bare -DDEBUG command-line define (src/meson.build, set for BOTH the debug and asan buildtypes -- confirmed via build warning "DEBUG redefined ... this is the location of the previous definition " across every file that includes utils.h under either profile), so DBG was used instead per user decision after the tradeoff was surfaced. Verified: xce-build.sh scheduler debug --incremental and xce-build.sh scheduler asan --incremental both compile clean (zero DBG/DEBUG-redefinition warnings in either profile); full scheduler suite 77/77 Ok, zero regressions.
Verification
Not recorded (migrated from buglog.json).
References
- Related: bug-1734, bug-1714, bug-1739, bug-1738