The fence re-check required !persist_recovery_truncated(), a boot-time
verdict nothing cleared, so a node that booted RECOVER_TRUNCATED stayed
fenced (fenced: true, non-promotable) until restart however far it
caught up. And the "progress is possible" branch did not re-arm, so the
fence was never re-evaluated.
persist_t gains a runtime recovery_truncation_healed latch, which
persist_recovery_truncated() honours (the raw boot verdict stays
available). The re-check marks the truncation healed once the durable
floor has advanced past the fence-time baseline (entries replicated from
the live primary) and reached the HWM, and the progress branch re-arms
with the same baseline and HWM.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
A late PreVote grant after WON re-ran the WON path and started a second
real election that replaced elect_ctx, so the term-2 leader promoted no
vote endpoints. Rounds now carry an id and endpoints a round/kind; stale
or wrong-kind replies, timeouts and errors are dropped; raft ignores
replies with no round in progress. raft_cb_become_leader promotes only
with its own stamped election record (own term, majority), else steps
down and re-arms. raft_on_leader_lost keeps voted_for (one vote per
term) and the delayed election starts only as a follower in the lost
term. The failover stagger hands the PreVote start to a vtask so the
delaymgr stays trigger-only.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
vtask_handle_msg published the VTASK_IO task while ep->fsm_in.pkt still
pointed at the packet; INVOKE_MSG_HANDLER clears the slot only after the
handler returns. A vtask could decode and free the packet (xce_decode)
and wake the client, whose failover reconnect then ran endpoint_fini and
freed the dangling slot again: ASAN heap-use-after-free in free_packet,
seen in the worker and scheduler during failovers under heavy CPU load.
vtask_handle_msg now clears ep->fsm_in.pkt before vtask_enq and marks
the task the packet's sole owner (owns_io_pkt), so never-run tasks free
it at vtaskmgr_fini and an enqueue failure frees it instead of leaking;
xcectx vtask_entry frees it on its early -EINVAL exits. Same take-and-
clear discipline the controller's peer-control path already used.
Tests: test_vtask_pkt_handoff.c (3), RED reproduces the soak's ASAN stack.
libxce 733/0; xce-build all green; loaded /xce-test-ha soak (8 rounds,
40 spinners) PASS with no ASAN reports (it crashed 2/2 before).
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
After an election where the vote connections are not promoted, both
standbys register later via on_standby_connected; the own-term no-op was
appended before any standby existed, and catch-up waited for heartbeat-
driven state reports (config push one 3 s cycle, WAL stream the next), so
the no-op committed ~6 s late and every client write in that window
failed -110.
ha_send_catchup_probe() sends the ordinary heartbeat request (same fields
as primary_heartbeat_fn; encode + enqueue only; primary only) at
registration and once after the standby's first accepted config-push ACK,
so the decision loop's next action follows at ~RTT. The decision loop,
its one-action-per-report pacing, commit/majority rules and the dispatch
gate are unchanged; WAL streaming still runs only on the vtask. The
one-shot arm is cleared on disconnect and on every become_leader /
become_follower.
Tests: test_late_joiner_catchup_kick.c (6). Controller suite 1285/0.
Live: catch-up starts 0-6 ms after a late registration (was up to ~6 s);
/xce-test-ha sequential 20 rounds, chaotic 6 rounds and sequential 8
rounds under 40-spinner load all PASS with zero job loss.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
- bug-1683 (test_wire known-good capture): the client closed its TLS
socket with the server's TLS 1.3 NewSessionTickets unread, so the kernel
sent RST and the server lost the unread payload record; the server
swallowed the OSError. The client now waits for the capture to finish,
handler errors are recorded and asserted, and a 50 ms "tickets-pending"
variant makes the RST race deterministic.
- bug-1485 (live HA push, versionTag=0): the CLI's 0 was an error reply
(-110): after a late-joiner election the primary's own-term no-op only
commits ~6 s later via heartbeat catch-up, so a push in that window
misses the majority barrier. The fixture now waits until the latest
own-term no-op has committed (wait_for_primary_write_ready); versionTag
asserts print the CLI's stdout/stderr. The controller latency itself is
bug-1913.
- bug-1725 (activation line missing): "Config upgraded" and the
activation line are logged after the reply; both tests now poll for all
three commit lines (shared assert_primary_committed).
xceadmin suite 373 passed; live push tests looped (95 runs incl. heavy
CPU load, 0 failures from these mechanisms).
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
- bug-1912: xce_parse_peers() ignores whitespace around entries and skips
blank ones, matching xceadmin's resolve_controllers(), so "a=h:1, b=h:2"
names the second candidate "b" (5 libxce tests + regression case).
- bug-1911: xcecd frees the -l log dir before the -h / --version returns
(regression case: no LeakSanitizer report, exit 0).
- bug-1910: both version scripts build the apiserver with -mod=vendor
(as the meson build does) instead of -mod=readonly, so they no longer
need a Go module cache.
- bug-1909: test-version-startup-consistency.sh accepts exactly the
dirty-content-fingerprint pairing ('<sha>-<8 hex>' in --version with
'<sha>-dirty' in the startup log), with self-test cases (e)/(f).
Both version scripts now exit 0 on a dirty tree with XCE_CONTROLLER_PEERS
set; regression script 35/35; xce-build all: C 127/0, scheduler 16/0,
xceadmin 364 passed, no ASAN reports.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
18 sub-tasks in 6 groups (libxce primitives + regression script
skeleton; controller; scheduler/worker/harness; xceadmin; integration;
docs). Each C task adds its failing regression-script case before the
change. Requirements, design and tasks approved.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
Iterated design review until clean (GO):
- Controller main() calls mylog_fini() before every early return, so
log lines written by tlog's background thread survive the exit.
- cluster_init_phase1() keeps its candidates_count > 0 guard around the
new xce_peers_find_identity() call (standalone stays my_priority 0).
- xceadmin adds the "(from $VAR)" source note only to network errors,
not to local cluster-bin file errors (OSError with filename set).
- Regression script: scratch --workdir/--certdir/--logfile per daemon,
refuses to run if a local cluster is up (port 9990, worker socket),
--sim case asserts the sim branch ran, 2.3 patches follow_redirects,
2.8 malformed-entry cases use a readable cluster-bin file, wrapper
test expects PEERS_STR form.
- Two existing scripts that start xce_scheduler without --peers will
unset XCE_CONTROLLER_PEERS; --interactive wording corrected.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
Design review (NO-GO -> GO) and iterated requirements review until clean:
- Controller refuses to start (non-zero exit, stderr, before any network
I/O) when an env-supplied XCE_CONTROLLER_PEERS lacks its own name (1.6);
new libxce xce_peers_find_identity() replaces both identity loops in
controller/cluster.c so the check and my_priority/listen port agree.
- deploy-ha.sh's generated scripts/xceadmin wrapper exports
XCE_CONTROLLER_PEERS so an ambient value cannot redirect it (Req 4).
- Opt-out is unsetting/emptying the variable (1.7); "No --peers" log
lines say it was unset or empty (1.8); docs cover sudo/systemd not
inheriting it (3.1).
- Requirement 1 defines explicit candidate list / resolution point /
own name so harness (set masters) is covered precisely; error messages
name the value's source incl. validation failures (1.4).
- xceadmin: name= stripping and a clean malformed-entry error apply to
every --controller source; errors name the env var source (2.2, 2.8).
- Test plan fixed against real code (stderr vs stdout, log texts,
xcecd_virtual, --norc, free ports, failover order test).
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
- apiserver peek (parseJobIDParam): <array-id>[k] now resolves to the
element; it used to drop [k] and look up the array identity itself
(and panicked in parseArrayName on some inputs). A job number still
names that job exactly (an element by its own XCE_JOBID); ranges and
invalid selections are rejected.
- apiserver restart: the decimal-only pre-check (isValidJobId, now
unused and removed) rejected the written form c-a-ts-seq[k];
ParseIdentitySelection, already used for the lookup, validates it.
- xcecli jobs restart converts its argument with ToAPIJobID like the
control commands, so written IDs work and invalid ones fail locally.
- docs: the peek/restart caveat in the Job arrays section.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
It logged in to a hard-coded http://localhost:8080 with fixed
credentials, so 'go test ./...' failed with connection refused. It now
skips unless XCE_LOAD_TEST_URL is set, with the credentials in
XCE_LOAD_TEST_USER/XCE_LOAD_TEST_PASSWORD; 'go test ./...' is green.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
REGEX_JOBID, REGEX_JOBARRAY_ID and N_MAX_FIELDS_JOBID served the old
regex-based parse_jobid(), replaced by parse_jobsel() in
job-array-one-based-index task 1.2; nothing references them.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
Formatting only (gofmt -w on the 29 non-vendored, non-generated files
gofmt -l reported); swag-generated apiserver/docs/docs.go left as is.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
The duplicate-id numbering lives in OpenWolf's post-write hook, which
openwolf update overwrites; auto_detect=false survives updates and the
pre-commit stale-index check catches any stray hook write.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
The pc13v capture's snapshot holds format_resreq() dumps instead of
resreq text (written before 077b1213), so the simulator cannot rebuild
that array job's request. Re-verified with fresh captures from the
current scheduler, jobs in the snapshot: 402 decisions, 0 MISMATCH.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
Re-verified 2026-10-01: test_jobj_hardening and the full xcelib_test run
clean with detect_leaks=1; meson's libxce test (abort_on_error=1) OK.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
clitest_wait_for_exit/clitest_wait_for_text added CLITEST_POLL_SLICE_MS
per loop instead of measuring time. Once xce_test closes its pty after
'quit', poll() returns at once on every call, so the 5 s budget was used
up within milliseconds while the (ASAN, loaded) child was not yet
reapable -- measured: real exit within 6 ms while the counter reached
58800 ms. The waits now use a CLOCK_MONOTONIC deadline, and
clitest_pump_once naps 10 ms when poll wakes with nothing to read.
Under 16 parallel runs: 12/48 runs failed before, 0/96 after.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
update_job_cpu_stats() (misc.c), run after each usage report:
- each host's cpu_rate = CPU-seconds gained / wall seconds since its
previous report; the job's cores = the sum over hosts
- peak_cpu = highest job cores seen (peak_cpu_time = it rounded, the
wire field is an integer); peak_cpu_duration = the interval that set it
- avg_cpu_efficiency = cpu_time / elapsed / slots, peak_cpu_efficiency =
peak_cpu / slots, mem_efficiency = highest used_mem / rusage[mem]
request (job-wide total, per task by default), all in percent
aggregate_job_runusage() no longer treats cumulative cpu_time as a peak.
New fields host sample_time/cpu_rate and job peak_cpu are appended to
the capnp structs (old WAL entries still decode) and copied by
merge_job_runusage(), so a promoted standby keeps the rate baselines.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
