admin
cee8f70e98
gates / gates (push) Failing after 17s
Overnight soak 22:39->06:10 CEST. demo-hp the victim, demo-felhom the untouched observer.
No production code, no golden, no version bump. Report at
documentation/audits/DRILL-soak-2026-08-31/REPORT.md.
VERDICTS: 1 lock-collision FAIL, 2 guard-interactions PASS-with-one-defect, 3 R-357 PASS,
4 proof-edges PASS, 5 mutated-cycle PASS, 6 observer FAIL, 7 teardown PASS.
R-414 - THE MOST VALUABLE FINDING, AND ONLY AN UNTOUCHED BOX COULD HAVE FOUND IT. On
demo-felhom the nightly proof fired for the first time unattended at 05:30 and REFUSED:
"nowhere to restore to - nincs regisztralt adatmeghajto". Cause established, not inferred:
storage_paths is EMPTY, so there is no path to put a scratch on. It will fail this way every
night forever with only a WARN, and because the error path reaches no verdict,
last_proof_result stays ABSENT - which is also what a pre-0.231.0 controller sends. The hub
cannot tell "never ran" from "not deployed": the StatsKnown trap one level up. The box is
NOT unprotected; its off-site backup ran fine in 46.9s. It is the PROOF that cannot run.
R-411 - measured, not reasoned: restic stats TAKES A LOCK; a customer full-restore runs it
while holding no acquireRunning; the integrity check is therefore not blocked, meets that
lock and escalates to unlock --remove-all. The sampler caught "restore ..." and
"unlock --remove-all" in the SAME sample. Contained: the check was classified unreachable,
not damage, so no false alarm.
R-412 - CORRECTED from HIGH to LOW. I filed it on a mechanism I had not finished measuring.
The off-site run has its own pre-push dump leg, so a hollow unit is REPAIRED before it
ships - proven on two apps and confirmed by pulling the snapshot back out of the store.
What survives is a narrow race, plus a success line over a backup holding no data.
R-413 - the R-87 proof caught a product-produced hollow snapshot unattended, and the
nightly job fired on its own schedule at 05:30 for the first time (bentopdf PASSED on
9d002b38 in 2.315s). Both were listed "not yet live-validated" yesterday.
R-403 mirror guard PROVEN live, with a negative control: it fired when a unit was hollow
("The copy was PRESERVED rather than replaced with an empty one") and skipped 0 legs at
teardown when every unit was sound.
R-357 PASS at last, six days owed: a real full filesystem, refused BEFORE StopStack, app
never stopped, live data byte-identical, and it worked once the space came back.
Phase 4 built the false-alarm control the whole R-87 design rests on: bentopdf is the only
template of 53 with neither a database nor a named volume. It passes silently.
EIGHT of my own instrument errors are named in the report, each caught by its own control -
including a time guard that fired an injection four hours early, and filing R-412 at the
wrong severity.
Teardown clean on all three layers of both boxes; both healthy on 0.231.0.
OWED: a golden for 0.231.0, and a decision on keeping bentopdf.
2.4 KiB
2.4 KiB
Phase 7 — restore the world. PASS
| layer | demo-hp |
demo-felhom |
|---|---|---|
PVE host /root |
.cache .ssh only |
soak files: 0 |
guest /root |
soak files: 0 | soak files: 0 |
container /tmp |
empty | n/a (never touched) |
scratches (offsite-restore, offsite-proof) |
both empty | n/a |
| ballast | 0 files; 949 GB free on hdd_1 |
n/a |
| controller | 0.231.0, healthy | 0.231.0, healthy |
| apps | 18 containers, all healthy, none exited or restarting | all healthy |
Local credential copies shred -u'd.
One clean cycle of each tier, checked for the RIGHT reason
- Tier-2: 3 apps copied, and 0 unit legs skipped — the R-403 guard's negative control. It fired earlier tonight when a unit was hollow and correctly does not fire now that every unit is sound.
- Proof:
bookstack→pass. - Off-site: the real 04:15 run completed in 2 m 57 s. Not re-run.
Every primary unit, final
bentopdf dumps=0 tars=0 <- correct: it legitimately has neither (the R-87 control)
bookstack dumps=1 tars=2 opengist dumps=0 tars=1
docmost dumps=1 tars=3 privatebin dumps=0 tars=1
kimai dumps=1 tars=2 calibre-web dumps=0 tars=1
paperless-ngx dumps=1 tars=3 romm dumps=1 tars=3
All Tier-2 copies present and sized as expected.
Two dispositions stated rather than left silent
bentopdf— KEEP (recommendation). It is the only template of 53 with neither a database nor a named volume, so it is the only possible live subject for R-87's false-alarm control. Removing it removes the control. Viktor's call./mnt/sys_drive/felhom-data/backups/primary/paperless/— a stale, manifest-less directory for a stack that is not deployed, dated 21–22 August. Not mine, established by mtime after my own baseline grep failed its control. Left in place: I did not create it and do not know why it is there. Harmless (no manifest ⇒ treated as hollow ⇒ the R-403 guard protects any copy), but it reads as an error in any unit sweep.
The hub layer — the one that is invisible from the box
The hub's host page mentions proof zero times: the controller publishes last_proof_* and no
hub surface reads them. That is R-402's shape, already allowlisted with that reason in
wire_contract_gate.py. No scratch customers, appliances or configs were created tonight, so there is
nothing to discard there.