controller v0.118.0: backup integrity F7/F6/F5 — docs + CHANGELOG + REPORT + CONTEXT

F7 atomic volume dumps, F6 no single-copy (volume-only tier-2 + honest single-drive
signal), F5 stale-primary sweep. Part 4 locality fork = A (doc-only). Live-validated
on 9201: F7 money-shot byte-identical through mid-write cut, F6/F5 confirmed, restore
round-trip byte-identical.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_017CDMFpFx84pfviCTVuGGhf

CHANGELOG.md

@@ -1,5 +1,38 @@
 ## Changelog
 
+### v0.118.0 — backup integrity: atomic volume dumps (F7) + no single-copy (F6) + stale-primary sweep (F5) (2026-07-12) — MinAgent: 0.81.0
+
+Fixes CAMPAIGN-3 (`felhom.eu/documentation/audits/CAMPAIGN-3-2026-07-11.md`) backup findings. MinAgent
+UNCHANGED (0.81.0) — all changes are controller-local; no new agent API consumed.
+
+- **F7 (HIGH) — atomic volume dumps.** `backup.DumpAppVolumes` now writes the tar to `<vol>.tar.tmp`,
+  fsyncs it, and only atomically `os.Rename`s it over the restore point on success — the same
+  crash-safe pattern the DB-dump path already uses (`appbackup/dbdump.go` DumpOne), extended with a
+  best-effort directory fsync. Before this, tar wrote the `.tar` IN PLACE, so a mid-write NFS cut left
+  a 0-byte tar REPLACING the last good dump (tier-1 restore is replace-semantics → an empty volume).
+  Now any tar error / timeout / dead-NFS EIO removes ONLY the `.tmp`; the last good `.tar` is
+  byte-untouched. The `.tar.tmp` name (ends `.tmp`, not `.tar`) is invisible to the
+  restore-point/stale scans; orphan `.tar.tmp` from a killed run is swept. New `tarVolume` test seam.
+- **F6 (LOW) — no single-copy backups.** Volume-only apps (no HDD_PATH, backups on sys_drive) now flow
+  through the tier-2 cross-drive copy (`RunAllTier2` no longer skips non-HDD apps) — a second copy on
+  the secondary drive (the 3-2-1 intent). Their restore-point drive label is no longer blank (clear
+  "Belső SSD (rendszer)"). A single-drive box (no off-drive target) surfaces an HONEST
+  `SingleCopyWarning` banner on the backup page instead of implying a 3-2-1 guarantee it cannot keep.
+- **F5 (LOW) — stale primary-dir sweep.** After each backup cycle, `pruneStalePrimaryDirs` removes an
+  orphaned `backups/primary/<app>` dir an app left on an OLD drive when its HDD_PATH moved (invisible
+  disk residue). LOAD-BEARING GUARDS: removes only when the app is deployed AND its current namespace
+  root differs from the dir's drive; NEVER touches the app's current-drive dir (the live restore
+  point) or an undeployed app's dir; only ever operates strictly under a `backups/primary/` prefix.
+- **Part 4 (operator fork) — backup-target locality: option A (keep locality), document-only.** NAS
+  apps' tier-1 artifacts stay beside the data on the NAS; tier-2's cross-drive copy is the off-NAS
+  leg. Documented plainly (backup feature doc) so the NAS-outage window is never a surprise; no code
+  change (option B, retarget-to-local, was not selected).
+- **Live-validated (demo 9201):** F7 money-shot — a mid-write `exportfs -u` during a volume dump left
+  all 5 nas-media volume tars BYTE-IDENTICAL (sha unchanged), no 0-byte, no leftover `.tar.tmp`, run
+  `success:false`; next run produced fresh good tars. F6 — actualbudget/seerr now on
+  felhom-usb/secondary. F5 — a seeded stale dir on the wrong drive swept, current dirs kept. Restore
+  round-trip byte-identical. Tests incl. the F7 truncation red-proof + F5 guard red-proofs, all green.
+
 ### v0.117.0 — consuming-namespace NAS verification + deploy-view truth (RCA fixes 2+4) (2026-07-11) — MinAgent: 0.81.0
 
 Controller half of the RCA fix pair (agent v0.84.0 ReassertNetworkMounts). Source:

CONTEXT.md

@@ -7,7 +7,23 @@
 >
 > Ask Claude Code: "Please update CONTEXT.md with what we did today"
 
-Last updated: 2026-07-11 (v0.116.1 — observability pass; pairs with agent v0.83.0 + hub v0.46.0)
+Last updated: 2026-07-12 (v0.118.0 — backup integrity F7/F6/F5; MinAgent 0.81.0 unchanged)
+
+> **2026-07-12 — v0.118.0: BACKUP INTEGRITY (LIVE on 9201). MinAgent 0.81.0 unchanged.** Fixes
+> CAMPAIGN-3 backup findings (`documentation/audits/CAMPAIGN-3-2026-07-11.md`). **F7 (HIGH) atomic
+> volume dumps:** `DumpAppVolumes` writes `<vol>.tar.tmp` → fsync → `os.Rename` over the `.tar` only on
+> success (`atomicPromoteTar`), mirroring dbdump.go DumpOne; a mid-write NFS cut can no longer
+> truncate the last good tar to 0 bytes. **F6 (LOW) no single-copy:** `RunAllTier2` no longer skips
+> volume-only apps (they now get a cross-drive tier-2 copy); sys_drive restore-point label is clear
+> ("Belső SSD (rendszer)"); single-drive box shows an honest `SingleCopyWarning` banner. **F5 (LOW)
+> stale-primary sweep:** `pruneStalePrimaryDirs` removes an orphaned `backups/primary/<app>` dir on an
+> OLD drive after an HDD_PATH move (guarded: deployed + different-current-drive only, never a restore
+> point). **Part 4 locality fork → operator chose (A) keep locality, doc-only** (NAS tier-1 stays on
+> the NAS; tier-2 is the off-NAS leg). LIVE: F7 money-shot (all NAS tars byte-identical through a
+> mid-write cut, no 0-byte, success:false); F6 (actualbudget/seerr on felhom-usb/secondary); F5
+> (seeded stale dir swept, current kept); restore round-trip byte-identical. Seams: `tarVolume`,
+> `perAppTier2`. Task C (F8/F4) + Task D (ring/alerting) still queued; Peti reaches 0.118 + agent 0.85
+> at his next train (agentless-on-proxmox2 gap noted).
 
 > **2026-07-11 — v0.116.0/0.116.1: OBSERVABILITY PASS (LIVE on 9201; agent v0.83.0 + hub v0.46.0).
 > MinAgent: 0.81.0 unchanged.** The debug ring (`LogBuffer`) now ALWAYS exists — logger =

REPORT.md

@@ -1,58 +1,60 @@
-# REPORT — v0.117.0: consuming-namespace NAS verification + deploy-view truth (RCA fixes 2+4)
+# REPORT — v0.118.0: backup integrity (F7 atomic dumps · F6 no-single-copy · F5 stale sweep)
 
-**Date:** 2026-07-11 · **Version:** controller v0.117.0 (from v0.116.1) · **MinAgent:** 0.81.0
+**Date:** 2026-07-12 · **Version:** controller v0.118.0 (from v0.117.0) · **MinAgent:** 0.81.0 (UNCHANGED)
-UNCHANGED (all checks controller-namespace-local; no new agent API) · **Deployed:** guest 9201
+· **Deployed:** guest 9201 (`0.118.0` healthy) · **Source:** `felhom.eu/documentation/audits/CAMPAIGN-3-2026-07-11.md`.
-(`0.117.0` healthy) · **NOT published/vouched** — Peti stays 0.113.0. Pairs with agent v0.84.0
-(RCA fix 1). Commits: `c0f3e12` + `4a9c54a` (linux-suite test pin).
-**Source:** `felhom.eu/documentation/audits/AUDIT-nas-cwa-rca-2026-07-11.md`.
 
 ## What shipped
 
-- **`internal/system/fsclass*.go`** — statfs f_type classifier for THIS process's namespace:
+- **F7 (HIGH) — atomic volume dumps.** `backup.DumpAppVolumes` writes the tar to `<vol>.tar.tmp`,
-  `network` (nfs/cifs/smb2) | `autofs` (healthy idle trigger, NEVER force-mounted) | `stub` |
+  fsyncs it, then atomically `os.Rename`s over the restore point ONLY on success (+ best-effort dir
-  `unknown` (error/3 s timeout → fail open). Seams: `statfsFn`, `netProbeFSClass`,
+  fsync via `atomicPromoteTar`) — mirroring the DB-dump path (`dbdump.go` DumpOne). Any tar
-  `Server.classifyFSPath`, `Router.classifyFSPath`.
+  error / timeout / dead-NFS EIO removes only the `.tmp`; the last good `.tar` is byte-untouched. The
-- **Fix 2a — probe assertion:** `--netprobe` child creates the probe file first (legitimately
+  `.tar.tmp` is invisible to restore-point/stale scans (ends `.tmp`, not `.tar`); orphan tmps are
-  triggers the automount), then REQUIRES a mounted network fs → exit 5 → category
+  swept. New `tarVolume` seam for the red-proof.
-  `not_network_fs`, new §3.2 Hungarian message, full rollback, nothing registered.
+- **F6 (LOW) — no single-copy.** `RunAllTier2` no longer skips non-HDD apps → volume-only apps get a
-- **Fix 2b — deploy refusal:** `POST /api/stacks/{name}/deploy` → `refuseNetworkStubDeploy`: a
+  cross-drive tier-2 copy. sys_drive restore-point label is now clear ("Belső SSD (rendszer)"), never
-  registered network `HDD_PATH` classifying as stub refuses 409; idle autofs / live / unknown /
+  blank. `SingleCopyWarning` (honest HU banner) shows on a single-drive box with no off-drive target.
-  local / unregistered / empty all proceed.
+- **F5 (LOW) — stale primary sweep.** `pruneStalePrimaryDirs` (end of each backup cycle) removes an
-- **Fix 2c — stub badge:** `networkStorageWarnings` → (warnings, stubs); stub leg runs even when
+  orphaned `backups/primary/<app>` dir on an OLD drive after an HDD_PATH move. Guards: deployed +
-  the agent is unreachable; dashboard + stacks cards render "Hálózati tárhely hibás — az alkalmazás
+  different-current-drive only; never the current-drive dir or an undeployed app's dir; strictly under
-  nem a NAS-t látja"; stub WINS over unreachable; the unreachable line stays byte-identical
+  a `backups/primary/` prefix.
-  (template-asserted). Pure core `networkStorageWarningsIn` + `stubNetworkPaths`.
+- **Part 4 — locality fork: operator chose (A) keep locality, document-only.** No code change; the
-- **Fix 4 — deploy-view truth (the RCA S-C symptom):** deployed-app storage select marks `selected`
+  NAS-outage behaviour + tier-2 mitigation are documented plainly.
-  by the STORED `HDD_PATH`; absent stored path → extra disabled `<path> (nem elérhető)` option;
-  `IsDefault` selects only for NEW deploys.
 
-## Tests + red-proofs
+## Tests + red-proofs (all green)
 
-| Test | Result | Red-proof |
+- F7: happy atomic promote (no `.tmp` left); **truncation red-proof** — pre-seeded good `.tar`, tar
-|---|---|---|
+  fails mid-write → original byte-unchanged, no 0-byte (revert to in-place write → fails); leftover
-| magic table + seam + real-local-dir-is-stub (linux) | PASS | — |
+  `.tar.tmp` swept.
-| `TestNetProbeChild_StubRefused` / `_AutofsAfterCreateRefused` / `_RealLocalDirRefused` (linux) | PASS | assertion disabled → **the stub VERIFIED** (exit 0) → FAIL |
+- F6: volume-only app in the tier-2 run set (revert the skip → fails); single-drive honest signal;
-| `TestNetAdd_StubProbe_RollsBackNotRegistered` (rollback + not registered + Hungarian) | PASS | same mutation → job phase `done` → FAIL |
+  sys_drive drive-label non-blank.
-| `TestRefuseNetworkStubDeploy_Table` (7 rows) | PASS | mounted-only gate → idle-autofs + unknown rows FAIL (over-refusal) |
+- F5: redeployed residue removed; **guard red-proofs** — undeployed app's restore point kept,
-| `TestNetworkStorageWarningsIn_StubWins` / `_UnreachableAloneUnchanged` | PASS | — |
+  current-drive dir kept.
-| `TestNetworkBadgeTemplates_DistinctStrings` (unreachable line byte-identical) | PASS | — |
+- Existing `TestListRestorePoints_DriveLabel/SSD` updated (blank → clear label, the F6 behaviour).
-| `TestDeployPage_DeployedSelectShowsStoredHDDPath` (real Manager + real template render) | PASS | IsDefault-only revert → default drive selected → FAIL (both assertions) |
-| `TestDeployPage_MissingStoredPathRendersTruthOption` | PASS | — |
 
-Full suite green on Windows AND on 180 (linux); the linux run caught the pre-existing C5 child-body
+## Live validation (demo 9201, real pipeline, sim-NAS rails — exportfs only)
-suite needing the classifier pinned to `network` (fixed in `4a9c54a` — the real statfs correctly
-refuses a tempdir). Gates: `template_id_gate.py` + `emoji_gate.py` OK.
 
-## Deploy + live validation
+- **F7 money-shot:** established good tars (radarr sha `847a97ad…` +4 others), then `exportfs -u` the
+  NFS target 4 s into a backup during volume dumps → run `success:false`; **all 5 nas-media volume
+  tars BYTE-IDENTICAL** (clean host-side diff), **no 0-byte file, no leftover `.tar.tmp`**; the next
+  clean run produced fresh good tars (`success:true`). This is the exact campaign F7 repro, now safe.
+- **F6:** after a tier-2 cycle, volume-only `actualbudget` + `seerr` appear under
+  `felhom-usb/backups/secondary/` (previously none); `actualbudget` snapshot `drive_label` =
+  "Belső SSD (rendszer)"; single-copy banner correctly absent on the multi-drive demo.
+- **F5:** a seeded stale `nas-media/backups/primary/actualbudget` (actualbudget lives on sys_drive) was
+  swept by the next backup; the real sys_drive dir + radarr's nas-media dir were kept.
+- **Restore round-trip:** calibre-web tamper → restore → BYTE-IDENTICAL (53 files), container healthy —
+  F7's write change did not affect restore.
 
-- Image `0.117.0` built+pushed on 180, deployed to 9201 via the bootstrap mechanism; healthy.
+## NOT live-validated / standing items
-- **Q1c reboot validation (operator GO, with agent 0.84.0):** post-reboot, no manual action —
-  CWA auto-started healthy on the NAS, health `status=ok` on the first cycle, dashboard renders NO
-  stub badge (grep count 0 against the live page).
-- **NOT live-validated (test-level only, by design):** the stub refusal branches — producing a live
-  stub post-fix would require breaking a healthy guest (the agent now heals it at start). Covered by
-  the red-proven unit/render tests above.
 
-## Follow-ups
+- Single-drive `SingleCopyWarning` live (the demo has enrolled drives) — unit-tested only.
+- Task C (F8/F4) and Task D (ring/alerting) remain queued.
+- Peti's box (controller 0.113 / agentless-on-proxmox2) reaches 0.118 + agent 0.85 at his next train —
+  noted for the roadmap; the agentless-on-proxmox2 cluster gap stands.
+- No publish/floor movement; agent untouched (MinAgent 0.81.0).
 
-RCA fixes 3/5/6 queued as separate tasks. Peti publish train: 0.84.0 + 0.117.0.
+## Box state at wrap
+
+controller 0.118.0 healthy on 9201; all backups consistent (fresh tier-1 + tier-2 incl. volume-only
+apps); NAS re-exported; no stale test residue.

controller/README.md

@@ -363,6 +363,13 @@ self-sufficient backup** — any single tier can fully restore an app.
   No tier depends on another tier's data.
 - **Tier 2 is configurable for ALL apps** — not just apps with HDD data. Non-HDD apps
   back up config + DB dumps to the secondary drive (small but protects against drive failure).
+  **(v0.118.0, F6):** the nightly tier-2 RUN (`RunAllTier2`) now actually includes volume-only apps
+  too — it used to skip every non-HDD app, leaving them a single controller-level copy on `sys_drive`.
+  A sys_drive app's restore-point drive label is now clear ("Belső SSD (rendszer)"), never blank.
+- **3-2-1 on single vs multi drive (v0.118.0, F6):** on a box with a second physical drive, tier-2 is
+  the off-drive copy. On a **single-drive** box (no off-drive target at all) there is genuinely only
+  ONE local copy — `FullBackupStatus.SingleCopyWarning` surfaces an honest Hungarian notice on the
+  backup page ("Csak egy másolat készül…") instead of implying a 3-2-1 guarantee the box cannot keep.
 - The `AppBackupPrefs.Enabled` field in settings.json is legacy and not read by any code.
 
 **Per-app Tier 2 contents by app type:**
@@ -380,6 +387,26 @@ self-sufficient backup** — any single tier can fully restore an app.
 
 The nightly backup has two phases that run sequentially. All paths are **per-drive** — each physical drive gets its own restic repo and per-app DB dump directories.
 
+> **Atomic dump writes (v0.118.0, CAMPAIGN-3 F7).** BOTH dump paths are crash-safe: the DB dump
+> (`dbdump.go` DumpOne) and the Docker-volume dump (`DumpAppVolumes`) write to a `.tmp` sibling, fsync,
+> then `os.Rename` over the restore point ONLY on success. A mid-write failure (a NFS cut mid-tar, an
+> EIO, a timeout) removes only the `.tmp` and leaves the last good `.sql`/`.tar` **byte-untouched** —
+> a tier-1 restore is replace-semantics, so an in-place write that got truncated to 0 bytes used to
+> destroy the only restore point. `.tar.tmp` files are invisible to the restore-point/stale scans and
+> orphans are swept on the next run.
+
+> **Stale-primary sweep (v0.118.0, F5).** After each cycle, `pruneStalePrimaryDirs` removes an
+> orphaned `backups/primary/<app>` dir left on an OLD drive when an app's `HDD_PATH` moved to another
+> drive. Guarded: only for a DEPLOYED app whose CURRENT drive differs from the dir's drive; never the
+> current-drive dir (the live restore point) or an undeployed app's dir; strictly under `backups/primary/`.
+
+> **NAS backup locality (v0.118.0, CAMPAIGN-3 Part 4 — decision A).** A NAS-resident app's tier-1
+> artifacts live **on the NAS itself** (`nas-media/backups/primary/<app>`), beside the data. During a
+> NAS outage both the app data AND its freshest tier-1 dump are on the dead device — the **tier-2**
+> cross-drive copy to a local drive is the off-NAS leg that saves them (and only after it has run).
+> This locality is deliberate (kept in the fork over retargeting tier-1 to a local drive); the tier-2
+> copy is the mitigation. Stated here so the outage window is never a surprise.
+
 **Drive layout (v0.26.0):**
 ```
 <drive>/