Files

T

admin 7714f1f6e4 docs(tests): Campaign 2 report + evidence — DR primitives, self-update, storage, faults, endurance

Full-system validation on demo 9201 (controller v0.102.0 / agent v0.73.0). Headline:
ONE genuine pilot-critical finding — the non-root agent cannot pct-restore the
standard Felhom guest because it has bind-mount mountpoints (mp8 /mnt/felhom-drives,
mp9 bootstrap) and pct restore refuses bind mounts unless root. This breaks BOTH
--selftest=restore-test AND the offsite DR bring-up (R1), linked to a missing
VM.Audit/VM.Allocate on the agent token (R2). The large raw FAIL count is dominated
by one harness bug (auth-session expiry mis-scoring P3-P9) plus D1-injection/soak-probe/
dryrun-journal/escrow-blob calibration issues; the product where actually exercised was
healthy with ZERO resource leaks over ~7h. Demo node healed to pre-campaign state
(auth off, notifications restored, no scratch leaks). No production code changed.

Evidence snapshot + README under campaign2-evidence-2026-07-07/ (cookie jars scrubbed).

🤖 Generated with [Claude Code](https://claude.com/claude-code)

Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01PSK5g6qYLknKj8u3QAFEr6

2026-07-07 14:26:52 +02:00

evidence

docs(tests): Campaign 2 report + evidence — DR primitives, self-update, storage, faults, endurance

2026-07-07 14:26:52 +02:00

state

docs(tests): Campaign 2 report + evidence — DR primitives, self-update, storage, faults, endurance

2026-07-07 14:26:52 +02:00

README.md

docs(tests): Campaign 2 report + evidence — DR primitives, self-update, storage, faults, endurance

2026-07-07 14:26:52 +02:00

README.md

Campaign 2 — evidence snapshot (2026-07-07)

Raw evidence for ../CAMPAIGN-2-2026-07-07.md (read that first — it triages everything here). Collected from the throwaway harness on DooPlex (~/campaign2-nomercy/); session cookie jars were deleted before commit (no secrets). Times are CEST.

Layout

state/ — run-level artifacts:
- overnight.out — the full run log (every phase, every JOURNAL line, all agent errors).
- journal.tsv — per-test verdicts (incl. the 170 T-SOAK-BREAK-* — all one artifact, H2).
- flightrec.log — 351 samples (~7 h): pct/docker/cc-latency/load/free + the leak sampler (c_rss/c_fd/c_sock controller, a_rss/a_fd/a_sock agent). The leak headline lives here.
- r0_snapshot.txt, p9_restore_snap.txt, d1_gate_ok.
evidence/<TEST-ID>/ — per-test evidence (61 dirs).

The pointers that matter

R1 (headline finding — bind-mount restore needs root): evidence/T-SELFRESTORE/restore-test.out (pass:false, mp8 … only possible for root), evidence/T-DR-BRINGUP/bringup.out (same on the DR path).
R2 (agent VM.Audit/VM.Allocate gaps): same two files (the 403 VM.Audit config-read + the 403 VM.Allocate scratch-teardown lines).
H1 (dominant harness bug — auth-session expiry): evidence/T-STG-ENROLL/init1.txt (CSRF token missing or invalid), evidence/T-RST-VW/{restore_resp.txt,canary_check.txt} (302→login, empty canary).
H3 (D1 not delivered): overnight.out around the [D1] block (sqlite3: … not found).
Positives: evidence/T-PBS-VERIFY/, evidence/T-RESTORE-TEST-CRASH/, evidence/T-FLT-{KILL9,AGENT,CF}/, evidence/T-BAK-FULL/, evidence/T-AUTH-1/, evidence/T-CSRF/.

Verdict in one line

One genuine pilot-critical product finding (R1/R2 — non-root can't restore a bind-mount guest); the rest of the FAIL count is a single dominant harness bug (H1) plus a few calibration issues; the product, where actually exercised, was healthy with zero resource leaks.