Part 4 (ships): `david` — a prospective customer with hosts=0, host_deletions=0,
reports=0 — e-mailed an expected_dbdump_missed ERROR at 03:00 UTC three mornings
running. The existing down-skip could never cover it: it reads the staleness
checker's state, which is seeded from a query over the `reports` table, so a
customer that never reported has no state at all and GetState() returns "" rather
than "down". store.HasEverBoundHost (hosts row OR host_deletions tombstone) is
consulted once per customer at the top of the deadline loop. The discriminator is
"was a host EVER bound", never "has a report arrived" — a box installed and never
heard from is a real fault and keeps alarming. Fail-OPEN on a read error. Red-proof
observed: removing the guard fails with `got [expected_dbdump_missed]`, verbatim the
event david sent.
Parts 0-3 (spike, NO production code for R-193/R-192):
audits/SPIKE-offsite-credential-recovery-2026-08-04.md establishes that the one-shot
provider password is the RECOVERABLE secret and the restic repository password is the
irreplaceable one — and that a guest rebuild mints a fresh one, orphaning the previous
off-site history. Measured without touching a box, by comparing
host_escrow.restic_pw_sha256 against host_escrow_superseded: BOTH demo boxes changed
(demo-hp 15 snapshots / 40.9 MB, demo-felhom 36 snapshots / 1.14 GB). demo-felhom's
"lucky" 76-second recovery restored delivery and not the repository, silently, for 13h.
ReissueCredentials does NOT rotate the restic password (R-39's record and two hub
comments are wrong -> R-196); candidate (b) is not implementable against a
zero-knowledge escrow; candidate (a) already exists as F3 and is wired to the wrong
event. Ends in ranked options and an unanswered question for the operator.
R-195 SHIPPED; R-196 + R-197 filed; R-192 + R-193 updated, neither closed.
The 20-minute latch expired at 10:20:29 and the hub logged degraded -> ok
(agent_capability_recovered) at 10:30:40. Final state on both boxes: agent
0.124.1, two ACL rows on /storage/felhom-backup.
CI run 28 failed and run 29 passed on the same sha. The tag was created locally at
07:38:36Z; run 28 started 07:38:31Z (commit push) and run 29 at 07:38:41Z (tag
push). The old race was 'tag before package'; the new one is 'package before its
tag is pushed' — the window the converse probe exists to detect, and it did.
Not confirmed from the run log (the endpoint 404s), so recorded as the
timing-consistent explanation rather than a proven one. The gate passes locally
with all 7 versions installable and none untagged. Both obvious closures are the
'make the gate blind for a window' shape this row refused, so the options are to
accept a seconds-wide window or to make releases quiet to CI another way.
Part 0's gate PASSED — ep0 prunes both namespaces daily since 2026-07-27 (18
tasks, all OK) — but three of my own queries said the opposite and all three were
broken instruments. Acting on them would have disabled the only pruning attempt
while filing a finding that nothing prunes.
Also records that v0.124.0's transition record failed in production with a green
test suite, that two red-proofs did not fail on the first attempt (one could not
compile, one asserted a helper rather than the path), and that two hollow tests
were caught in one file.
R-191: the gate was run first and it mattered — ep0's prune jobs have run daily
since 2026-07-27 (18 tasks, all OK). THREE of my own queries said the opposite and
all three were broken instruments (worker-type vs worker_type, prune vs prunejob,
proxmox-backup vs proxmox-backup-proxy); acting on any would have disabled the only
pruning attempt while reporting that nothing prunes.
R-190: the row now reflects the MITIGATION, not the cause, and says so. Includes
that v0.124.0's transition record failed in production while every unit test
passed, and Part 3's single pass — the token-rotation class is real and documented
in our own installer but does not fit this box (same token still authenticates, the
other three grants survived, no installer run evidenced); ruled out on this box, not
fleet-wide.
R-194: PVE caches permissions (~40s and ~16min measured), so a single permission
read is a lagging indicator and may explain when R-190's refusal surfaced.
07-backup-architecture: the offsite retention row was RIGHT and the configuration
disagreed with it — noted, since a doc that states a contract does not enforce it.
Every weekly off-site run uploaded successfully and then failed the job on a prune
the box's token is deliberately refused — R-89 moved off-site pruning server-side
to ep0 and box tokens stay write-only. The 2026-07-26 'two weeks' ruling was not
reversed; where it is enforced moved, and keep_last: 2 did not follow. Now 0,
which the agent's existing guard already reads as 'never prune from the box'.
Verified read-only on ep0 before changing it: both namespaces have a prune job at
03:30 keep-last 2 that has run every day since 2026-07-27 — 18 tasks, all OK, the
newest keeping exactly two. Without that check this would have traded a weekly
false alarm for unbounded growth.
A gate asserts the offsite tier carries no client-side prune. The local tier is
untouched.
Re-issue through the designed endpoint: fresh password staged 07:11:51, settle-gate
GO, consumed 15s later at 07:12:06, offsite configured at 07:12:09 against the SAME
sub-account (275124) and repo path. Both boxes now escrowed.
CORRECTION to this session's own reading: the escrow did NOT recover by itself.
escrow_state went pending -> escrowed 15s after the apply and I inferred automation;
the operator had run the ceremony. It needed a human on both boxes — demo-hp at
07:16:02, demo-felhom at 07:20:28 (its escrow had been pending since its own rebuild
the day before). A 15-second state change is not evidence of automation, which is the
same class as reading an absent log line as success.
Snapshot continuity is NOT established: both boxes report snapshot_count 0, but the
run-history keys are ABSENT rather than zeroed — a controller that has never run an
offbox backup, not an empty repo. The 04:15 run decides it.
Operator confirms no hub-side offsite config change, so the regression was not an
action. Evidence: demo-hp's controller went 0.187.0 -> 0.192.0 at 06:12:18 with a
new config hash and the agent re-keyed its leaf three minutes earlier — a guest
rebuild. The last pre-rebuild report shows the tier fully healthy: escrowed, last
success 02:16:39Z, 15 snapshots, 40.9 MB. No offsite object in the 108 reports
since.
Mechanism: the restic credential is delivered once. demo-hp consumed its secret on
2026-07-23; the rebuilt controller has no copy and no way to request another.
demo-felhom survived the SAME rebuild only because its secret was still unconsumed
— it consumed it four seconds after its config hash changed and was reporting
offsite again 76 seconds later. That difference was luck, not design.
Also sharpens R-192: the self-heal's guard refuses when any report since the
consume carried an offbox target, but that query reads the OLDEST 500 reports —
all of which predate the rebuild. Healthy history before a rebuild is not evidence
the credential still works, which is why the automation that exists for this case
declined to act.
demo-hp has been emailing daily since 2026-08-03 06:12 UTC that '500 reports
since carry no offbox target — the credential is likely burned'. Measured against
the hub's own data: all 500 of those reports DO carry one. The credential was
consumed 2026-07-23 and applied successfully; the controller reported an offsite
object continuously until 2026-08-03 05:59:21, then it vanished at 06:12:19 and
has been absent for 108 reports. A regressed apply, not a burn.
Two defects. maybeEmitStuck builds its message from ReportsSinceConsume while
hardcoding 'carry no offbox target' and never consults OffsiteReportsSinceConsume,
so the recommended action targets a failure that did not occur. And maybeHeal
refuses SILENTLY on exactly that field ('regressed-apply shape -> operator's
call', a bare return), so there is a daily email with the wrong story, no heal,
and nothing saying why the heal declined.
The underlying condition is real: that box currently has no off-site copy of its
customer app data, since 08:12 CEST on 2026-08-03. What removed it is not
established.
Four SCHEDULED runs, none triggered by hand: demo-felhom host 83.8s / offsite
540.4s; demo-hp host 109.3s / offsite 300.1s. Each restored into a scratch guest,
booted, verified and destroyed itself; zero 990000 guests or volumes afterwards
and both local-lvm figures returned to their pre-run values.
Both boxes had BOTH tiers due at once, so R-86's ordering was observed live for
the first time: never-proven sorted first, each box took its HOST tier, deferred
the offsite one, and picked it up on the next evaluation six hours later. The
host-tier proofs reached the hub through R-189's merge — demo-felhom's report
carries two tiers, and the local one can only have come from disk.
The capability map's optimistic half is cashed, with its scope stated: these two
boxes, not the fleet.
Surfaced and filed rather than fixed:
- R-190: a storage ACL that demonstrably worked at 04:44 was gone by 09:24, with
a reinstall, any logged pveum activity and any cluster-log entry ruled out.
- R-191: every weekly offsite backup uploads successfully and then fails the job
on a prune the box is deliberately not allowed to do (R-89 moved it
server-side; both boxes still arm keep_last=2).
Two corrections to yesterday's record: the R-185 drift DID surface as 403s on the
write path (six, with the hub raising whole_guest_backup_failed at the first), and
my earlier "no restore_test_* events" was produced by grepping a 404 page.
CORRECTION 1 — the runbook annotation and the R-185 row both said the drift did
not surface as a 403 because writes go through a root path. That is WRONG.
demo-felhom's local-api backup jobs 403'd six times between 09:24 and 17:34 CEST
on exactly that storage and privilege, and the hub raised whole_guest_backup_failed
at the first with edge-triggering suppressing the rest. The impact was not only an
unreadable tier: the agent's own whole-guest backups to it were failing.
CORRECTION 2 — on this box the grant was LOST, not never issued. A vzdump by the
agent's token to that storage completed OK at 04:44:50 the same morning; the first
403 is 09:24:56. Ruled out by measurement: a host reinstall (uptime 12 days), any
pveum/ACL/user.cfg activity in syslog 04:00-10:00, any ACL entry in the cluster
log. Correlated but not established: guest 9201 was reprovisioned nine minutes
before the first failure.
R-190 files the unexplained disappearance, and notes that the new store-grant
probe detects the STATE but says nothing about the TRANSITION.
- OPEN-ITEMS: R-185 closed with the measurement, the corrected root cause (the
installer's Scenario-F reuse arm, not PVE_STORAGES), and the live sequence.
Records that demo-hp carried the same drift and was fixed too.
- capability map: the whole-guest row's HOST-tier half was OPTIMISTIC and now
says so — that tier was not merely unproven, it was unprovable on both demo
boxes, and every live proof cited was on the offsite tier.
- vzdump-target-move runbook: its item 5 predicted this; annotated (not
rewritten) with what actually happened — the create arm did grant, the reuse
arm did not, and it surfaced as a silent unreadable tier rather than the 403
the item expected, because vzdump writes through a root path.
- CONTEXT: S-21 (an empty listing cannot distinguish forbidden from newborn; the
measured trap that an ungranted path answers with INHERITED privileges) and
S-22 (the Scenario-F arm must finish the job).
- STATUS: rewritten for the operator, back to one screen.
configure_backup_target has two arms and only one granted. Case A creates the
storage and grants in the same breath; the Scenario-F arm ('the target already
exists') returned without granting. A box whose felhom-backup pre-dated the
install therefore pointed local_backup_target at a storage its own token could
not read — measured on BOTH demo boxes: {"data":[]} through the token while root
lists three archives. That tier was never restore-tested and nothing said so,
because an empty listing is also what a brand-new tier returns.
The reuse arm now ensures the ACL through the same guarded wrapper. Scenario F is
unviolated: the storage DEFINITION is untouched, and pveum acl modify is
idempotent. BACKUP_TARGET_ID is deliberately NOT added to PVE_STORAGES — that
list is granted a step before the target is resolved, and --acl-storages entries
are preflight-checked for existence; the comment now says so.
A gate asserts it: every arm that resolves the target must also grant on it.
Red-proved by reverting the arm.
- OPEN-ITEMS: all three rows closed with their measurements. R-189 records the
migration consequence seen live (a pre-R-189 record has no tier and is not
reportable, so upgrading does not retroactively make an old proof visible).
R-188 records both directions demonstrated and the fixture cleaned up. R-186
carries the before/after sha values including the CGO_ENABLED discrepancy.
The priority list states explicitly that R-185 is the one left open from that
group and that this session did not touch it.
- capability map: the restore-proof row now says the evidence path itself had a
gap — the proof reached the hub only because no restart intervened — and what
closed it.
- CONTEXT: S-19 (a proof is durable and reportable, a failure is neither, and
why) and S-20 (the release ordering, what each step protects, and how a binary
is verified).
- STATUS: rewritten for the operator, back to one screen.
The agent reports restore_tests[] from an IN-MEMORY store ('lost on restart; the
cadence re-populates'), and the hub report interval is 900s. A restart inside
that window drops the result. That used to self-heal within 24h because the next
cadence re-tested the tier; under per-archive due-ness the agent will NOT re-test
a proven archive, so the hub can stay ignorant until the next archive generation
— a week on the offsite tier.
Observed today: the 15:25:14 PASS on felhom-pbs (a real 14.5 GB offsite restore)
reached no host-report at all, because the agent was restarted 2m43s later for
the v0.121.1 deploy.
The persisted proof already exists — RestoreTestState.Snapshot() even carries the
comment 'for the host-report gauge' and has NO production caller: a seam built and
never wired, and an invariant asserted in a comment with no test, in one method.
on: [push] fires the gates workflow on the TAG push too, and release-agent.sh
pushes the tag before publishing (deliberately). CI can therefore run the
published-versions gate inside the window where the tag exists and the package
does not, and correctly reports 'every released agent version must be
INSTALLABLE' for a release that completes seconds later.
Measured across two releases in one session: v0.121.0 -> runs 12/13 (success,
failure) on the same sha; v0.121.1 -> runs 17/18 (failure, success) on the same
sha; and one pair both green — a race, not a rule.
It matters because R-168 made CI email on failure so a red gate cannot be missed.
A signal that cries wolf on every second correct release is how that mail becomes
something you archive unread.
- OPEN-ITEMS: R-86 CLOSED with the trap in its own wording recorded (the literal
reading is never true on a daily tier); R-87 re-ranked UP because R-86 built
most of what it waited for; R-185 (the agent cannot list demo-felhom's host
backup tier — a missing storage ACL, pre-existing), R-186 (a released binary's
sha is not reproducible from its tag), R-187 (R-115's publish leg had never
actually run) filed. R-184 was the highest ID in use.
- ROADMAP: R-86 collapsed, keeping the reasoning and correcting the shape the row
itself proposed — which would have been the never-fires version.
- 07-backup-architecture: new contract section — restore-testing is per ARCHIVE
GENERATION, with the trap and what did not change (S-1).
- 00-capability-map: the unattended restore-proof row upgraded to PROVEN-LIVE on
the 635 s due-triggered offsite run, with the restart and teardown evidence.
- CONTEXT: S-17 (the rule, the trap, the config key, the hub's derivation) and
S-18 (ep0 is Tier 2 — extends D-d's protected list to three machines).
Numbered 17/18 because S-14 and S-15 were already duplicated in the file.
- STATUS: rewritten for the operator, trimmed back to one screen.
Found by checking v0.91.0 against the live box, not by review. demo-felhom's two
retained PBS snapshots sit 8h54m apart (one is a healing artefact), so the
mean-gap estimator reads a WEEKLY tier as nine-hourly: x4 = 36h, the 7-day floor
lifts it to 168h, and a weekly tier proved weekly reaches ~8.25d of proof age.
The false alarm this task exists to prevent would have returned within a week, on
the box it had just shipped to.
restoreProvenWindow now takes max(observed, declared). A gap SHORTER than the
declared rhythm is routine and means nothing (a retry, a manual run, a heal, a
catch-up); a gap LONGER than it is real information. Cost stated: a tier running
faster than its declared rhythm gets a slower stale signal — the right direction
for a signal that means 'unverified', since 'broken now' is a different event.
Ships WITH agent v0.121.0, not after it. The agent now proves a tier once per
ARCHIVE GENERATION, so a weekly tier is proved weekly — in perfect health. The
flat 7-day restoreProvenStaleAfter derived its number from the 24h cadence R-86
removes, and a healthy weekly tier's proof age reaches EXACTLY 168h just before
its next proof: it sat ON the line, so any ordinary delay tipped it into a
nightly alarm about a working system.
restoreProvenWindow(tier, observed, ok):
- the tier's own archive interval, OBSERVED from reports the hub already holds
(pbs_snapshots + successful backups attributed by TARGET TYPE, slice A.4)
- x4 generations = the same tolerance the flat constant expressed
- floored at 7d (never tighter than before), capped at 12d (strictly inside the
2-week offsite retention)
- falls back to the DECLARED rhythm (26h host / 8d offsite — the thresholds the
backup-freshness checker already uses) when history is too short to observe
one; falling back to the FLOOR would recreate the false alarm on a fresh box
Kept: absence is UNKNOWN until the anchored window passes; the signal stays
edge-triggered; failed and stale remain distinct events. Every reason string now
states the window it was judged against (R-100's corollary).
Also backfills the missing v0.90.1 CHANGELOG entry (deployed since f21e7ca), and
records the operator's 2026-08-03 ruling that ep0 is Tier 2 / protected.
R-182 CLOSED (controller v0.194.0 + hub v0.90.0/.1), proven live on demo-hp. The
hub's notification_log for the run reads: two per-app failures RECORDED, one
digest SENT naming both, and the customer channel SKIPPED with operator_only.
Against the measured previous behaviour — two failures, one email naming one
app, one leaving no trace anywhere.
Scenario D proved itself on an event I had not planned: disk_critical alarmed on
two filesystems, the second was collapsed by the cooldown, and that collapse is
now visible WITH ITS KEY. Yesterday it would have left nothing at all.
A gap the spec did not anticipate is recorded with its fix: the per-app event
also fires from the periodic sweep, outside any run, so making it record-only
would have created a NEW silence. The sweep emits a digest too, with no run_id,
so it stays under the ordinary hourly cooldown.
ep0: MEASURED on the box — 7757 MB (8 GB), 4 vCPU, and the 4 GiB swapfile
SURVIVED the resize and is active (checked, because a resize is a stop/start).
The 40 GB local disk is UNCHANGED, so no disk figure was touched anywhere.
Five documents corrected — three of which the task's list did not name, found by
searching. Two audit/evidence documents ANNOTATED, body untouched: they record
what was true when written and that is their value.
R-90 CLOSED. R-86 unblocked and re-ranked, stated honestly: 8 GB is comfortable,
not unbounded — the original OOM was a 14.46 GB restore — so the restore-test
cadence should still be paced, just not by fear of the endpoint.
target-selection.md's "D-d did not name ep0 either way" is deliberately left
standing. It is the operator's question, not CC's.
STATUS.md 127 -> 83 lines, items rather than sentences.
Found by reading the first REAL digest, not by design. Every app row ended with
the same usage clause the mail already prints once on its own Filesystem line.
On a two-app box that is untidy; down a list of a dozen it is the same forty
characters twelve times, pushing the part that DIFFERS off a phone screen at
07:00 — the only moment this mail has to work.
The reserve's refusal message is authored for a single-app alert where naming
the filesystem is right, so the message is unchanged; the digest trims the
duplicate when rendering. trimRepeatedUsage removes ONLY an exact
"— <target path>:" suffix, so an unrelated reason is untouched and a reason that
is nothing but the usage clause is left alone rather than emptied.
Also updates TestRecoveryUnitCaptureFailed_NeverReachesTheCustomer, which
required the OPERATOR to be emailed a per-app capture failure. That was correct
when the event was the only signal and is wrong now that it is the record and
the digest is the notification. Its customer-safety claim is unchanged and is
why the test still exists; the operator assertion is inverted with the reasoning
written in place, and R-158's guarantee is shown to have MOVED, not weakened.
processOperator's cooldown no longer returns bare. It dropped the event BEFORE
LogNotification, so a suppressed operator alert and an event that never happened
were indistinguishable — from the operator's side and from the hub's own records.
Measured 2026-08-03: nine recovery_unit_capture_failed events arrived, two were
mailed, seven left no row anywhere. That is why the defect took a day to get the
right way round: there was nothing to read.
A suppressed operator event now writes a `suppressed` row carrying the message
and the key that suppressed it. This applies to EVERY operator event, not only
the one that exposed it. It does NOT change the cooldown's duration or semantics.
backup_run_failures: the per-run digest. In allowedEventTypes AND in
operatorOnlyEvents — allowlisting alone does not make an event operator-only,
and FormatCustomerEmail falls back to the raw English message rather than
blocking. A test demonstrates a customer with the type enabled receiving nothing.
recordOnlyEvents: a third routing class — stored and recorded, never mailed.
recovery_unit_capture_failed moves here: it is the record, the digest is the
notification. A register rather than downgrading severity to info, which would
relabel a genuine failure as informational everywhere it is queried.
cooldownRunSuffix: a sibling of cooldownTierSuffix, not a branch inside it, so
tier keeps byte-identical semantics and R-97a's tests are untouched. It makes
the cooldown effectively inert for the digest, which is the intent — a digest is
already rate-limited by construction; the refresh sweep sends no run_id and so
stays under the ordinary hourly cooldown.
The email renders as a list, not a JSON blob. An absent space reading renders as
unavailable, never as zeros.
R-182's direction REVERSED by Part 0's measurement. Filed yesterday as "the
reserve re-alerts on every status refresh" — too many alerts, seen at the
sending end. Measured at the receiving end: 9 events received today, 2 operator
emails sent. When two apps are refused in the same second the operator is told
about ONE; the other is dropped before LogNotification, so it leaves no row on
any channel and cannot be audited. The operator cooldown key is
customerID:eventType(+tier) and the capture-failed event carries `app` but no
`tier`, so the key has no app identifier. Same failure mode as R-97a, in a
second event type that never opted into the narrow fix. Nothing changed —
Part 0 was investigation only.
Correction owed: yesterday's report said "one recovery_unit_capture_failed per
app, HTTP 200". True of what the CONTROLLER pushed; a reader would take it as
"the operator was told about each app", which is false.
R-110 CLOSED (installer v1.23.0). Both channels moved. The spec's mechanism for
channel 2 rested on a factual error — the run-time fetches are sixteen, not
nine, and come from felhom-agent, not this repo — so no tag here could cover
them; pinned to the agent version being installed instead, on the operator's
ruling. Channel 3 needed no change: the URL never carried a ref, so no hub
change and no hub bump.
R-115 CLOSED. release-agent.sh builds, tags, publishes and verifies by an
independent download; check-published-versions.py refuses a tag with no package;
CI now runs the full gate set so it actually runs.
R-183 NEW+CLOSED: a fresh install fetched the vouched agent binary and its
sixteen config files from two different refs, and nothing compared them.
R-184 NEW: nothing stops the hub vouching a version that was never released.
The R-115 gate cannot see it — measured, the hub manifest and Gitea's package
listing are both 401 anonymously.
capability map: new PROVEN-LIVE row for the published installer channel.
STATUS.md 138 -> 127 lines.
Scenario A: pushed to main without moving the tag; the served bytes stayed
byte-identical (sha 2f859555…) and the marker never appeared, while the website
tree advanced to the new commit — both halves of the split in one observation.
Scenario B: moving the tag published it (~40s, sha ea2b4aa9…, marker present);
moving it back rolled it back to exactly the pre-publish sha with the marker
gone. The website returned 200 throughout.
A comment only. It exists to be LOOKED FOR at the served URL: if it appears
there, /scripts/ is still tracking main and the publish channel is still
floating. It must not appear until installer-v1.23.0 is moved.
Two channels moved off main in the same change, because either one left behind
makes the other cosmetic.
Channel 1 — the served script. webpage.yaml git-synced /scripts/ from
--branch=main every 30s and nginx served that tree, so pushing this file WAS
publishing it: within half a minute it was what every new machine downloaded and
ran as root, with no staging and no rollback but another push. The sync is now
SPLIT: the website keeps tracking main at the same cadence (a copy edit must
never need a release) and /scripts/ tracks the tag installer-v<SCRIPT_VERSION>.
PROVEN before the manifest was touched: git-sync v4.4.0 follows a tag AND
notices a MOVED one — measured on a throwaway sync against this repo,
"update required ... local:<old> remote:<new>" -> "updated successfully",
within one period. The moved-tag half is what the publish model rests on.
Channel 2 — the sixteen files fetched at run time. fetch_raw pulled from
$AGENT_REPO/raw/branch/main; it now pulls raw/tag/v$ART_AGENT_VER. That is a
correctness fix, not only a channel one (R-183): a fresh install fetched the
vouched agent BINARY while taking its unit file, sudoers and guarded wrappers
from whatever main held. Two refs, one install, nothing compared them. Their
correct ref was never SCRIPT_VERSION — they do not live in this repo.
No fallback to a branch: a vouched version whose tag is missing fails loudly
rather than quietly serving main.
Channel 3 — the URL — needed no change, recorded rather than left silent:
https://felhom.eu/scripts/felhom-host-install.sh never carried a ref, so both
producers follow the tag with no edit. No hub change, no hub version bump.
Gate 6 in hostinstall_gates.py pins all three structurally with no network, so
it stays in --fast and runs in CI. It deliberately does NOT assert "a tag exists
for the current SCRIPT_VERSION": that would go red on the very push that bumps
the version, before publishing — and publishing being separate is the ruling.
Syncs documentation/runbooks/workspace-CLAUDE.md with the workspace root file.
Row 7: the B2 refusal claimed the previous unit was untouched; nothing-deleted
held, untouched was measured false.
Also records WHY it survived review: it passed a full green suite AND three of
its own red-proofs, because every one of them asserted the mechanism inside
captureAllRecoveryUnits and none asserted the consequence across the whole
backup run. The test that would have caught it is the one the fix ships —
fingerprint the tree before and after, and compare.
R-181 CLOSED (controller v0.193.0 + v0.193.1) and proven live on demo-hp for
BOTH reserve terms. The reserve is now a per-app, per-run ADMISSION decision
taken before the app's first write and covering all three write legs, and it
gained a size term. The refusal's wording was not weakened; the behaviour moved
so it became true, verified by sha256 tree fingerprint.
R-156 CLOSED — papra's template mounts the app's own data root. Precondition
re-measured rather than inherited (both boxes were wiped today).
Part 4, documentation only, nothing built:
- R-110 WAITING-ON-OPERATOR -> READY. Ruling: option (b), the installer's
publish channel moves to a TAG. Recorded with the condition that decides
whether it works at all — it must cover BOTH the /scripts/ git-sync and the
nine files the installer fetches from raw/branch/main.
- R-115 WAITING-ON-OPERATOR -> READY. Ruling: mechanism (b), a build-side gate
refusing to deploy or vouch an unpublished version. The third instance (agent
v0.120.0) would have silently downgraded both demo boxes while succeeding.
R-182 NEW: the periodic status refresh has no admission scope, so a refused app
re-alerts on every poll (measured: a second alert pair 13s after the run's).
Pre-existing in v0.192.0; deliberately not fixed in the R-181 task.
capability map: the local-backup row moves to PROVEN-LIVE in BOTH halves.
ROADMAP: R-165 collapses to CLOSED; R-181 collapsed into it.
07-backup-architecture.md: the reserve's contract stated as what the code
provides (S-1 — an architectural contract changed in the same session).
STATUS.md trimmed 150 -> 111 lines, "What's broken" no longer holds shipped
work, and the stale "After:" line (pointing at work that shipped on 2 August)
is fixed.
The skill's own description claimed 'ANY Felhom artifact' and 'publish', and had no ISO section —
a description asserting coverage that did not exist. Description corrected and a section added.
POINTERS, NOT COPIES. The 13-criterion release gate stays in
documentation/runbooks/iso-release-gate.md and the measurements stay in the four spike audits;
duplicating them into a skill guarantees drift (the R-94/R-128 class). What the skill adds is the
ROUTING that was missing: nothing told anyone the gate exists, which is R-29's exact shape.
Records the two modes (--release public vs --pairing appliance) because picking the wrong one ships
the wrong product, the build and publish commands (rclone env-only, so no credential file is ever
written), and the round-trip verification.
The traps it carries existed only in commit messages until now, and each cost a wrong diagnosis:
- 'qm set --scsi0 ... --boot order=' in ONE call silently yields boot: order=net0;ide2
- after install the CD must be detached, or a COMPLETED install looks exactly like a stuck one
- verify focus by screendump before every Enter (GTK Enter lands in fields, not Next)
- proof installs register appliances; the verb is POST /appliances/<id>/discard, not /delete
- scratch storage at the /mnt/nvme-1tb ROOT (a subdirectory reads disconnected forever)
Also flags that the ISO gate is NOT wired into repo_gates.py, so nothing reminds you to run it.
Docs only. python3 scripts/repo_gates.py --fast: all gates OK (rc=0).
Two boxes, two DIFFERENT supply paths, so the session proved the disk shape and
the delivery route rather than one of them twice.
demo-hp (layout proof, --golden <local volid>): mp0 at /var/lib/felhom,
backup=1, 70G, no mp1; /var/lib/docker and /mnt/sys_drive both real mounts of
its subdirectories via fstab; one df figure and one device id (64519) on all
three paths; reboots 3/3 with the binds surviving each.
demo-felhom (pipeline proof, --force-gitea-golden): fetch_verify succeeding
against the vouched manifest for BOTH artifacts -- 'verified sha256
54e2a4c431daf580... matches the hub manifest' for the golden, a7763d31... for
the agent. 250G single volume, grep -c '^mp1:' = 0, reboots 3/3.
Journey proven on both, endpoint-level: claim -> deploy -> back up -> restore,
with a planted marker returning byte-identical on each box. Ceiling measured
gone: 65 GiB and 233 GiB available to a recovery unit, against 19 and 45.
R-165 -> IMPLEMENTED, not PROVEN-LIVE, on the operator's ruling. B2, which that
row records as the bulkhead's replacement, fired live for the first time and
does refuse per app, delete nothing and alert -- but it is checked only in
captureAllRecoveryUnits while runVolumeDumps writes the bulk unguarded, and its
'the previous unit is untouched' claim was measured false (182,272 B dump
replaced by 2,147,666,432 B under a manifest still dated 06:34:26). -> R-181.
New: R-179 (uninstall leaves NAS network-storage units), R-180 (--archive-storage
not cross-checked against the ACL grant; 403 at step 8/8 after root@pam is
rotated), R-181. Third instance of R-115 recorded (agent 0.120.0 unpublished).
No code written, no version bumps -- this was a runbook.
07-backup-architecture.md gains §7.5.1 (S-1: the contract changed in the same
session): the ceiling §7.5 describes no longer exists for a box built from
golden >= 0.192.0, the bulkhead's replacement is recorded, and R-175 is FIXED
here rather than left standing — the bound is restated as a function of mp1
and scoped to split-layout boxes, naming all three real shapes.
Capability map: new row as IMPLEMENTED, deliberately NOT proven-live, with
the missing leg named — no box has been reinstalled from the golden, and
"the golden baked" is not "a box built from it works".
R-163 CLOSED: the ceiling it recorded stops existing. R-176(a) answered by
P1; (b) WITHDRAWN, since every node is reinstalled rather than migrated.
R-178 filed for the reinstalls, which were not done this session.
CONTEXT S-13 (the variant chosen on measurement; pruning rejected with its
reason) and S-14 (prove first, then vouch — the golden is published but
deliberately unvouched, because vouching is what makes a fresh install pick
up a layout no box has been proven from).
STATUS: plain-language section; both operator questions now answered, so the
waiting-on-you item is cleared. Two older entries trimmed so the page did
not grow.
felhom-agent v0.120.0 merges the two data volumes into one, and step_grows
computed two numbers while the install call passed both — so this had to
change with the agent or every install would have provisioned a half-sized
box. The 80/20 split is summed (226 = 184+42), so a standard appliance keeps
exactly the 250 G it had, no longer split by a wall.
The size still comes from the physical disk: step_grows already read the
thin pool's free space, and the merge only collapsed its two outputs into
one. --sysdata-grow is deprecated but still honoured, because the agent
folds a hand-passed value in rather than dropping it.
P1 PASS: a pre-merge archive (mp0+mp1, confirmed from its own vzdump log)
restore-tests clean on demo-hp with mount_parity ok in 84s. mountParity was
not touched. Limit stated: run with the current agent because the merged one
does not exist until after the STOP; the comparison is archive-vs-its-own-
restore and never consults the host layout, so it carries provided Part 2
honours its constraint not to touch the restore path. Re-run after Part 2.
P2: all three probed variants are mechanically clean — both paths writable,
ONE df figure, dockerd 3/3 reboots, /mnt propagation, and a container's
statfs(/) reporting the merged volume. The task's flagged ordering risk for
V-b did not materialise. They are separated by SCOPING instead:
V-a container sees /mnt = 8.0K but customer data sits inside Docker's
data-root, so clearing /var/lib/docker destroys every local unit
V-b container sees Docker's ENTIRE data-root under /mnt (17.9M on an
empty box), making the bootstrap's own scoping comment false
V-c neutral mount at /var/lib/felhom, both paths binds — breaks neither
V-c was probed because the measurements showed each named variant violates
a different documented invariant. It is offered as a measured option for the
operator, NOT adopted.
Teardown all three layers: 9401/9402/9403 destroyed, 5.19 GB returned, and
the hub registers verified unchanged (5 customers, 4 hosts).
R-167/R-158 shipped and proven live (controller v0.191.x, hub v0.89.0):
two new capability-map rows PROVEN-LIVE with live citations, and
07-backup-architecture.md §7.5's closing claim "nothing warns when an app
crosses the line" is now false and rewritten (S-1: an architectural
contract changed in the same session). §7.5 also gains the caveat that its
size bound is ONE BOX'S, not the fleet's.
Part 3 SPIKE (audits/SPIKE-r165-mp1-merge-2026-08-02.md): M1-M5 measured,
NO layout touched. Three findings the merge session must not re-derive:
"the layout" is not one thing (200G/50G vs 50G/20G vs 16G/8G); mp1 is a
BULKHEAD and not only a ceiling, so after the merge an overflow reaches
/var/lib/docker; the golden fails closed on the split in four places.
D-a's condition (1) is currently SATISFIED — no external box is in the
hub's register, and both demo boxes are Tier 0 and reinstallable.
Recommendation given, choice NOT made — it ends at the operator's ruling.
CONTEXT.md S-11 (D-c's routing, and why R-158's own backup_failed proposal
was overruled) and S-12 (the monitoring landed BEFORE the merge).
STATUS.md gains the plain-language section and the merge decision, with two
older entries trimmed so the page did not grow.
New rows R-174 (closed same session), R-175, R-176, R-177; each ID grepped
free before minting.
New OPERATOR-ONLY event type recovery_unit_capture_failed (controller
v0.191.0, R-158): in allowedEventTypes AND notify.operatorOnlyEvents.
Deliberately not a reuse of backup_failed, which carries customer copy and
sits in the controller's DefaultEnabledEvents — reusing it would email the
customer in Hungarian about a failure they cannot act on. R-158's own
proposal said backup_failed; D-c overrides it.
disk_warning/disk_critical lose their generic customerMessages entries.
Both were allowlisted, copy'd, default-enabled and checkbox'd with NO
producer anywhere; controller v0.191.0 becomes that producer and sends a
DYNAMIC Hungarian message naming the drive and its free space.
FormatCustomerEmail prefers the entry over the message, so keeping a static
entry would discard the label and the byte figures — the same reason
offbox_enlarge_blocked and disk_health_degraded have none. The deletion is
pinned by a test.
New notify.IsOperatorOnly so the api package can pin BOTH registers of a new
event type in ONE test; allowlisted-but-not-operator-only is invisible when
they are checked separately, and it is the defect v0.78.0 shipped. The
register itself stays unexported.
REUSE.md's "new event type" extension point rewritten: it told readers to
always add a customerMessages entry, which is wrong for operator-only types
and harmful for dynamic-message ones.
Tests 574 -> 579. Red-proof: removing the operatorOnlyEvents entry shows the
customer being emailed; the skipped/operator_only row is asserted as a
positive observable.
admin