Files
felhom.eu/documentation/audits/space-list-2026-11-10/TABLE.md
T
admin 449b065fbd
gates / gates (push) Successful in 1m57s
Record the operator's rulings of 2026-11-10 12:22 (D67-D70, the gap review's scope): 09 §3 decisions 286-290
D67 yes as a separate step (R4 unchanged), D68 operator only, D69 a section
inside 16, D70 yes with three changes; DooPlex is never measured, ep0 not yet.
STATUS: D67-D70 out. Rows R-1368, R-1370, R-1371 carry the ruling.

Claude-Session: https://claude.ai/code/session_0159rPz1ZhFKsS53msqPYxtS
2026-11-10 12:46:21 +01:00

52 KiB
Raw Blame History

The Space list — every place that holds data, eight questions each

Measured 2026-11-10, first reading 07:38–07:39 UTC, second reading at the end of the session (README.md gives the rates). Read-only. Sizes are allocated bytes; 1 GB = 10⁹ bytes. Boxes: hp = demo-hp (128 GB system disk), N100 = demo-felhom (512 GB, lived-in), T1 = Tester 1 (a VM, 200 GB). Guest = LXC 9201 on each.

The eight questions: 1 what is in it, who owns it · 2 what makes it grow · 3 the limit · 4 who cleans up, when · 5 what happens at the limit; is household data safe · 6 who gets an alarm, before or after · 7 does it repair itself; if not, is there a runbook · 8 how do we know.

Marks: A answered, with a pointer · D decided, not built · K known gap, with its row · N nobody asked — it survived the searches in EV/searches.txt (S1…S15; controls in the same file). Every N names its class row: C1 = R-1368 · C2 = R-1369 · C3 = R-1370 · C4 = R-1371 · C5 = R-1372.

Pointers: AG/ = felhom-agent/ · CT/ = felhom-controller/controller/ · HUB/ = felhom.eu/hub/internal/ · E/ = felhom.eu/ · AR/ = felhom.eu/documentation/architecture/ · RB/ = felhom.eu/documentation/runbooks/ · EV/ = this folder's evidence/. tools/check-pointers.py prints the text behind each one. Source lines are as of the baselines in README.md.

Named status (after the place name): NAMED = an architecture document describes the place · MENTIONED = a document names it in passing, with no limit or cleaner · UNNAMED = no architecture document names it.

A place inside a bigger place (a folder on the host system disk, on the guest data volume or on a data drive) takes question 5 and 6 from the bigger place when it has no limit of its own. Question 7 is N for each place that has no cleaner and no runbook — it is ONE missing runbook seen from many places (C2).


Host (Proxmox) — 18 places

H1 — Host system disk, root filesystem / (LV pve/root) · CUSTOMER · NAMED

AR/07-backup-architecture.md:467, AR/16-storage-physical-identity.md:146-150. Used: hp 23.97 of 41.42 GB (58 %) · N100 38.09 of 100.86 GB (38 %) · T1 21.64 of 62.90 GB (34 %). Since 2026-10-15: N100 +4.6 GiB, T1 +5.8 GiB (26 days).

# Answer Mark Pointer
1 Proxmox, kernels, the host journal, the whole-box copies when the target is local, the agent's files. The operator owns it. A EV/reading-1/hp-host.tsv
2 H3, H4, H5, H7, H9, H12 below. They hold 72 % (hp), 82 % (N100), 80 % (T1) of the used bytes. A EV/reading-1/n100-host.tsv
3 The LV size the Proxmox installer made. No budget says which part may use how much of it. N C3 · S5
4 Each part by its own rule, or nobody (H5, H9, H12, H13). The whole disk: the operator, by hand, last on 2026-11-07. K R-548
5 Under 500 MB free the OS update is refused; a local whole-box copy that cannot fit is skipped; LVM's own metadata writes can fail. Household data is not on this disk. A AG/configs/felhom-os-apply:1390-1392, AG/internal/backup/runner.go:276-287, AG/internal/vzpartial/vzpartial.go:1-12
6 The hub: host_disk_warning at 90 %, host_disk_critical at 95 % — before it is full. To the operator; the critical one also to the household by default. No architecture document names this alarm or its reader (C4, S4; the alarm ladder names only the thin-pool one). A HUB/monitor/host_disk.go:41-42, HUB/claim/engine.go:229-243
7 No self-repair. No runbook for a full host system disk. N C2 · S6
8 TEST and MEASURED. A HUB/monitor/host_disk_test.go:40, EV/reading-1/t1-host.tsv

H2 — Thin pool pve/data (data and metadata) · CUSTOMER · NAMED

AR/08-alarm-ladder.md:298-306, AR/16-storage-physical-identity.md:126-150. Data used: hp 40.88 of 57.84 GB (70.67 %) · N100 3.03 % · T1 13.10 %. Metadata: 2.85 % · 0.58 % · 1.84 %. Promised to the guest: hp 102 GiB on a 53.9 GiB pool · N100 282 on 349 · T1 102 on 113. hp was 50.75 % on 2026-10-15.

# Answer Mark Pointer
1 The guest's root disk and data volume; for a short time the restore-test's scratch guest. A EV/raw/hp-host.txt
2 Each block the guest writes. A deleted block comes back only at the weekly trim. A AG/internal/fstrim/fstrim.go:1-12
3 The pool size. Auto-extend is off. The sizing rule promises nothing beyond pool − reserve, but on a 128 GB disk the two guest disks alone (32 + 50 or 70 GiB) are larger than the 54 GiB pool. This is written down. A AR/16-storage-physical-identity.md:146-150, E/documentation/audits/storage-identity-buildB-2026-10-15/SPACE-CHECK.md:41-42, EV/os-cleaners/hp-host.txt
4 The agent trims the guest each week (decision 139). A scratch guest that stays is removed by a 10-minute retry. A AG/internal/fstrim/fstrim.go:1-12, AG/internal/reconcile/restoretest_retry.go:5-12
5 At 100 % the guest's disks go read-only and every app stops. Measured on demo-hp on 2026-09-24. Data already written stays. A AR/03-host-agent.md:513-520
6 The hub: storage_fill_warning at 85 %, storage_fill_critical at 90 %, the worse of data and metadata. Operator; the critical one also to the household by default. On hp the guest's own warning (85 % of 70 GiB) lies above the pool's real size, so this is the only alarm that can come first. A AR/08-alarm-ladder.md:300-307, HUB/monitor/storage_fill.go:36-57, HUB/claim/engine.go:229-243
7 No self-repair. The one pool runbook covers a different case (an orphan snapshot volume). No runbook for a full pool. N C2 · S6 · RB/whole-guest-backup-orphan-thin.md:1-10
8 TEST and MEASURED. A HUB/monitor/storage_fill_test.go:34, EV/reading-1/hp-host.tsv

H3 — Whole-box copies on the current local target · CUSTOMER · NAMED

AR/07-backup-architecture.md:401. hp: /var/lib/vz/dump, 1 archive, 9.10 GB (keep 1) · N100: /mnt/hdd_1/dump (the USB disk), 3 archives, 4.36 GB (keep 3), 2026-11-08…11-10 · T1: /var/lib/vz/dump, 3 archives, 9.46 GB (keep 3).

# Answer Mark Pointer
1 One packed copy of the whole guest per night. The agent makes it; Proxmox stores it. A AR/07-backup-architecture.md:401
2 The guest's size. Most of it is Docker images that could be pulled again. K R-1366
3 Keep the newest N (default 3). Proxmox removes the older ones after a good run. A AG/internal/config/config.go:540-550, AG/internal/proxmox/mutate.go:99-101
4 Proxmox, after each good run. A half-written copy: the agent, at start and before each backup. A AG/internal/backup/runner.go:227-230, AG/cmd/felhom-agent/main.go:2227-2238
5 A copy that cannot fit is skipped before it starts. On a small system disk the local copy then stops for ever. K R-548
6 The skipped run is reported to the hub and shown on the box's backup page with its reason. A AG/internal/hub/report.go:620, CT/internal/web/backup_handlers.go:149-152
7 No self-repair when the disk is too small. K R-548
8 TEST. A AG/internal/backup/r685_space_test.go:31, AG/internal/backup/prune_test.go:24

H4 — Whole-box copies on a former target · CUSTOMER · UNNAMED

N100 /var/lib/vz/dump: 3 copies of 2026-07-28, 17.59 GB (the target moved to the USB disk) · hp /mnt/hdd_1/dump: 4 copies of 2026-08-03…08-21, 5.10 GB · T1: none.

# Answer Mark Pointer
1 Old whole-box copies on the place that was the target before. A EV/raw/n100-host.txt
2 Only a move of the target. After that they stay. A EV/raw/hp-host.txt
3 None. Retention looks only at the current target. K R-1087
4 Nobody. K R-1087
5 On a small system disk they leave no room for the new copies. K R-1087
6 The host-disk alarm of H1, when they lie on the system disk. A HUB/monitor/host_disk.go:41-42
7 No. K R-1087
8 MEASURED. Second instance today: demo-hp. A EV/reading-1/hp-host.tsv

H5 — Guest template archives ("golden") in the dump folder · CUSTOMER · UNNAMED

/var/lib/vz/dump/vzdump-lxc-9100-*.tar.zst. hp 1 (0.65 GB, 2026-09-13) · N100 4 (2.56 GB, 2026-07-18…08-03) · T1 1 (0.65 GB, 2026-10-04).

# Answer Mark Pointer
1 The archive the installer downloads to make the guest. Product-owned. A E/scripts/felhom-host-install.sh:3560-3565
2 One more at each install or re-install of the box. A EV/raw/n100-host.txt
3 None. N C1 · S3
4 Nobody. Only the uninstall with --remove-golden deletes it. N C1 · S3 · E/scripts/felhom-host-install.sh:1520-1524
5 It counts against H1. A AG/configs/felhom-os-apply:1390-1392
6 The host-disk alarm of H1. A HUB/monitor/host_disk.go:41-42
7 No self-repair, no runbook. N C2 · S6
8 MEASURED. A EV/reading-1/n100-host.tsv

H6 — Backup logs and half-written copies in the dump folder · CUSTOMER · MENTIONED

AR/03-host-agent.md:518. Logs: hp 10 (9 of them from failed runs of 09-24 and 10-14, under 2 KB each) · N100 6 · T1 3. Half-written copies today: none.

# Answer Mark Pointer
1 One text log per backup run; after a power cut a half-written archive. A EV/raw/hp-host.txt
2 One log per run. A half-written copy was 12.2 GB on demo-hp on 2026-10-14. A AG/internal/vzpartial/vzpartial.go:1-12
3 The log of a failed run has no limit: it has no archive, so nothing removes it. Each is under 2 KB. N C1 · S15
4 Proxmox removes a good run's log with its archive. The agent removes a half-written copy at start and before each backup. A AG/cmd/felhom-agent/main.go:2227-2238
5 A half-written copy made each later backup skip, until the agent learned to remove it. A AG/internal/vzpartial/vzpartial.go:1-12
6 The host-disk alarm of H1. A HUB/monitor/host_disk.go:41-42
7 Yes, for the half-written copy. A AG/cmd/felhom-agent/main.go:2227-2238
8 SOURCE and MEASURED. A EV/reading-1/hp-host.tsv

H7 — Host journal /var/log/journal · CUSTOMER · MENTIONED

AR/03-host-agent.md:72. hp 0.87 GB, 16 files, oldest 2026-10-04 (cleaned by hand on 11-07) · N100 4.28 GB, 74 files, oldest 10-04 · T1 4.29 GB, 81 files, oldest 10-13. Lines in one hour: 13,946 / 13,918 / 11,558; lines written by sudo for the agent's calls are 92 % / 92 % / 90 % of them (R-1367's 61 % counts the session pair only; sudo writes a third line, the command, for each call).

# Answer Mark Pointer
1 The host's system log. Nobody set it up; it is the Debian default. A EV/os-cleaners/hp-host.txt
2 The agent's sudo calls: three lines per call, about 12,800 lines an hour. K R-1367
3 journald's default: 10 % of the disk, at most 4 GB. Two boxes sit at that cap. K R-1367
4 journald deletes its oldest file when it reaches the cap. A EV/reading-1/n100-host.tsv
5 On a small system disk the journal alone takes a tenth of it. K R-1367
6 The host-disk alarm of H1. A HUB/monitor/host_disk.go:41-42
7 It stays at its cap by itself. The cap is not ours. K R-1367
8 MEASURED, two readings. A EV/reading-1/t1-host.tsv

H8 — Other host logs, Proxmox task logs, graph data · CUSTOMER · MENTIONED

AR/03-host-agent.md:570. /var/log without the journal: 14.6 / 14.0 / 14.6 MB; task logs 914 / 706 / 877 files; /var/lib/rrdcached 65 / 21 / 5 MB.

# Answer Mark Pointer
1 Text logs, one small log per Proxmox task, fixed-size graph files. A EV/reading-1/hp-host.tsv
2 Each task the agent starts makes one log. A EV/raw/hp-host.txt
3 Rotation: logrotate; Proxmox keeps its task index short (oldest task file 2026-10-28). A EV/os-cleaners/hp-host.txt
4 logrotate.timer and Proxmox itself. A EV/os-cleaners/n100-host.txt
5 Small and bounded; it counts against H1. A EV/reading-1/n100-host.tsv
6 The host-disk alarm of H1. A HUB/monitor/host_disk.go:41-42
7 Yes, by rotation. A EV/os-cleaners/t1-host.txt
8 MEASURED. First written down today. A EV/reading-1/t1-host.tsv

H9 — Old kernels: /lib/modules and /boot · CUSTOMER · MENTIONED

AR/11-os-updates.md:402. hp 8 kernels, 8.15 GB + /boot 0.82 GB · N100 8, 8.15 GB + 0.94 GB · T1 3, 3.05 GB + 0.33 GB. /boot is a folder on the system disk. The ring-0 boxes installed 7 kernels in 37 days; no kernel was ever removed.

# Answer Mark Pointer
1 Each kernel the box ever installed: about 1.02 GB of modules and 80 MB in /boot. A EV/raw/hp-host-followup.txt
2 Each kernel set the operator approves; ring 0 takes each one first. A AR/11-os-updates.md:164-176
3 None. No rule says how many kernels a box keeps. N C1 · S1
4 Nobody. apt would remove old ones only on autoremove, which nothing runs; the update wrapper refuses each plan that removes a package. N C1 · S1 · AG/configs/felhom-os-apply:1712, EV/os-cleaners/hp-host.txt
5 Under 500 MB free the wrapper refuses the OS update, so security updates stop. After 7 days the hub says the updates are stale. A AG/configs/felhom-os-apply:1390-1392, AR/08-alarm-ladder.md:375
6 The host-disk alarm of H1; later os_update_stale to the operator. A HUB/monitor/host_disk.go:41-42, HUB/notify/dispatcher.go:840
7 No self-repair, no runbook. N C2 · S1
8 NOBODY KNOWS how many kernels a 32 GB system disk holds before updates stop. The sizes are MEASURED. N C1 · EV/reading-1/hp-host.tsv

H10 — EFI partition /boot/efi · CUSTOMER · MENTIONED

AR/11-os-updates.md:532-536. 9.5 MB of 1.07 GB on each box (1 %).

# Answer Mark Pointer
1 The boot loader, and the one-shot flag for a new kernel. A AR/11-os-updates.md:532-536
2 A boot-loader update. Kernels are not copied here (GRUB boxes). A AR/11-os-updates.md:402
3 The partition: 1.07 GB. A EV/reading-1/hp-host.tsv
4 Nothing needs to: 1 % used. A EV/reading-1/n100-host.tsv
5 Not written anywhere. N C2 · S13
6 Nobody: it is its own filesystem and nothing watches it. N C4 · S13
7 No runbook. N C2 · S13
8 MEASURED. A EV/reading-1/t1-host.tsv

H11 — Host package cache and package lists · CUSTOMER · NAMED

AR/11-os-updates.md:293. /var/cache/apt 92 MB with 0 package files on each box; lists 62 MB.

# Answer Mark Pointer
1 Downloaded packages during an OS update, and the package index. A EV/reading-1/hp-host.tsv
2 Each OS update run. A AR/11-os-updates.md:293
3 Emptied at the end of each run. A AG/configs/felhom-os-apply:1417
4 The update wrapper (apt-get clean). A AG/configs/felhom-os-apply:1417
5 A run that would not fit is refused before it starts. A AG/configs/felhom-os-apply:1390-1392
6 The host-disk alarm of H1. A HUB/monitor/host_disk.go:41-42
7 Yes: cleaned each run. A AG/configs/felhom-os-apply:1417
8 SOURCE and MEASURED (0 package files). A EV/reading-1/t1-host.tsv

H12 — The agent's self-update folder /var/lib/felhom-agent/selfupdate · CUSTOMER · UNNAMED

hp 739.9 MB, 49 binaries · N100 723.9 MB, 48 · T1 509.2 MB, 33. About 15 MB for each agent version the box ever got.

# Answer Mark Pointer
1 One downloaded agent binary per version. The agent owns it. A AG/internal/selfupdate/executor.go:106-128
2 Each agent update. A EV/raw/hp-host-followup.txt
3 None. N C1 · S2
4 Nobody. N C1 · S2
5 It counts against H1. A AG/configs/felhom-os-apply:1390-1392
6 The host-disk alarm of H1. A HUB/monitor/host_disk.go:41-42
7 No self-repair, no runbook. N C2 · S2
8 MEASURED. A EV/reading-1/hp-host.tsv

H13 — Old copies of the agent binary (.prev, .bak-*, .pre-bundle) · CUSTOMER · MENTIONED

AR/03-host-agent.md:893. /usr/local/bin/felhom-agent.bak-*: hp 1 file, 14 MB · N100 11 files, 155 MB · T1 none.

# Answer Mark Pointer
1 The previous agent binary (.prev), and one copy for each re-run of the installer (.bak-<time>). A AG/configs/felhom-selfupdate-guarded:86, E/scripts/felhom-host-install.sh:2675
2 .prev is one file, replaced each update. .bak-* gains a file at each installer re-run. A AG/configs/felhom-selfupdate-guarded:114
3 None for .bak-*. N C1 · S2
4 Nobody. N C1 · S2
5 It counts against H1. A AG/configs/felhom-os-apply:1390-1392
6 The host-disk alarm of H1. A HUB/monitor/host_disk.go:41-42
7 No self-repair, no runbook. N C2 · S2
8 MEASURED. A EV/reading-1/n100-host.tsv

H14 — Update wrapper state and previous signed bundles · CUSTOMER · MENTIONED

AR/11-os-updates.md:343. /var/lib/felhom-os-apply: 0.4 MB, 3 previous bundles on each box.

# Answer Mark Pointer
1 The update wrapper's state and the last signed config bundles. A EV/reading-1/hp-host.tsv
2 Each new bundle. A AR/11-os-updates.md:343
3 The last 3. A AG/configs/felhom-os-apply:190
4 The wrapper, at each bundle. A AG/configs/felhom-os-apply:2315-2318
5 Bounded: 0.4 MB. A EV/reading-1/n100-host.tsv
6 The host-disk alarm of H1. A HUB/monitor/host_disk.go:41-42
7 Yes: pruned each time. A AG/configs/felhom-os-apply:2315-2318
8 SOURCE and MEASURED. A EV/reading-1/t1-host.tsv

H15 — Host /tmp and /var/tmp · CUSTOMER · UNNAMED

/tmp is memory (tmpfs), 5–6 entries, 0 MB · /var/tmp 1–4 entries, 0 MB.

# Answer Mark Pointer
1 Short-lived files of the OS and of the agent's wrappers. A EV/reading-1/hp-host.tsv
2 Nothing lasting. A EV/reading-1/n100-host.tsv
3 /tmp: the tmpfs default, half of the memory. A EV/os-cleaners/hp-host.txt
4 systemd: 10 days for /tmp, 30 days for /var/tmp; /tmp is empty after each restart. A EV/os-cleaners/hp-host.txt
5 A full /tmp uses memory, not disk. Not written anywhere what then fails. N C2 · S6
6 Nobody watches /tmp. /var/tmp counts against H1. N C4 · S6
7 Yes: the clean-up timer and each restart. A EV/os-cleaners/t1-host.txt
8 MEASURED. A EV/reading-1/t1-host.tsv

H16 — The bare folder under an unmounted drive · CUSTOMER · NAMED

AR/16-storage-physical-identity.md:209, AR/02-controller-module-map.md:179. Today: 0 bytes. T1's guest shows four empty, not-mounted folders (media_a … media_d).

# Answer Mark Pointer
1 The empty folder where a drive is mounted. With the drive gone, a write there lands on the system disk. A AG/configs/felhom-backup-target-apply:21-29
2 A write while the drive is absent. A AG/internal/localapi/disks.go:414-420
3 The guards refuse the write: the backup storage is tied to the mount, and the placeholder folder is owned by host root, which the guest cannot write. A AG/configs/felhom-backup-target-apply:97, AG/internal/guesthook/heal.go:1-10
4 Nothing to clean: 0 bytes. Nobody removes the empty folder of a drive that left. A EV/raw/t1-9201.txt
5 Refused before the write; the drive's size is reported as unknown, never as the system disk's. A AG/internal/localapi/disks.go:414-420
6 storage_disconnected to the household. A AR/08-alarm-ladder.md:518-524
7 Yes: the pre-start hook makes the placeholder so the guest still starts. A AG/internal/guesthook/heal.go:1-10
8 TEST (named in the source comment). A AG/internal/localapi/disks.go:414-420

H17 — The restore-test's scratch guest · CUSTOMER · NAMED

AR/03-host-agent.md:513-520, capability row A19. None present today on any box.

# Answer Mark Pointer
1 A throwaway guest restored from last night's copy, to prove the copy works. A AR/03-host-agent.md:513-520
2 Each scheduled restore test: the full unpacked size of the guest. A AR/03-host-agent.md:513-520
3 A space check before anything is made: free ≥ restored size × 1.2 + 5 GiB, and off the tested guest's pool when possible. A AR/03-host-agent.md:513-520
4 The agent destroys it at the end; a failed destroy is retried each 10 minutes. A AG/internal/reconcile/restoretest.go:476-483, AG/internal/reconcile/restoretest_retry.go:5-12
5 It once filled the pool and the guest went read-only (2026-09-24); the space check exists since then. A AR/03-host-agent.md:513-520
6 The pool alarm of H2. A HUB/monitor/storage_fill.go:36-57
7 Yes: the retry timer. A AG/internal/reconcile/restoretest_retry.go:5-12
8 DOC with a measured incident, and SOURCE. A AR/03-host-agent.md:513-520

H18 — Swap (LV pve/swap) · CUSTOMER · NAMED

AR/08-alarm-ladder.md:399-402. 8.59 GB (hp, N100), 8.25 GB (T1); used 0 / 0 / 186 MB.

# Answer Mark Pointer
1 The host's swap space. A EV/reading-1/hp-host.tsv
2 Memory pressure. It is a fixed-size volume. A EV/reading-1/t1-host.tsv
3 The volume size. A EV/reading-1/n100-host.tsv
4 The kernel. A EV/reading-1/t1-host.tsv
5 A memory problem, not a disk problem. A AR/08-alarm-ladder.md:399-402
6 host_swap_short to the operator, when the swap is smaller than the guests' allowance. A AR/08-alarm-ladder.md:399-402
7 Not a space question. A AR/08-alarm-ladder.md:399-402
8 DOC and MEASURED. A EV/reading-1/hp-host.tsv

Guest 9201 — 17 places

G1 — Guest root disk (vm-9201-disk-0, 32 GiB) · CUSTOMER · MENTIONED

AR/16-storage-physical-identity.md:134. Used: hp 1.36 GB · N100 1.19 GB · T1 1.18 GB of 33.5 GB (4–5 %).

# Answer Mark Pointer
1 The guest's Debian, the Docker programs, /opt/docker/stacks, the package cache. A EV/reading-1/hp-9201.tsv
2 Guest OS updates. Very little else. A EV/reading-1/n100-9201.tsv
3 32 GiB. The agent can only make it larger. A AG/internal/reconcile/bringup.go:405-447
4 The update wrapper empties the package cache; logrotate; the guest journal is in memory (G15). A AG/configs/felhom-os-apply:1417, EV/os-cleaners/hp-9201.txt
5 The guest's OS update is refused when free space is too low. A AR/11-os-updates.md:302-306
6 Nobody. The controller's / is the data volume, not this disk — measured on all three boxes. The closed row R-521 says the guest root is watched; that was true only before the two volumes were merged. Capability-map row C13 says "per filesystem": this filesystem is the exception. N C4 · S9 · EV/extra/controller-fs-view.txt, CT/internal/system/dockervol.go:14-26
7 No self-repair, no runbook. N C2 · S6
8 MEASURED. A EV/extra/controller-fs-view.txt

G2 — Guest data volume (mp0, /var/lib/felhom: Docker and sys_drive) · CUSTOMER · NAMED

AR/16-storage-physical-identity.md:126-150. Used: hp 30.09 of 73.79 GB (41 %) · N100 2.47 of 263.09 GB (1 %) · T1 8.99 of 73.79 GB (12 %). hp was 19.1 GB on 2026-10-15.

# Answer Mark Pointer
1 Each app's programs (images), each app's own data (volumes), container logs, the backups of apps with no drive. A EV/reading-1/hp-9201.tsv
2 G3–G12 and G14 below. On hp the images are 81 % of it. A EV/reading-1/hp-9201.tsv
3 The volume: 50 / 100 / 130 GiB by pool size on a new install (older boxes: 70 or 250). A new app is refused below max(5 GB, 10 %) free. A backup is refused at 97 % used or under 1 GiB free. A AR/16-storage-physical-identity.md:146-150, CT/internal/system/dockervol.go:14-26, CT/internal/backup/recovery_unit.go:383-388
4 The three image rules (G3, G4, G5); the weekly trim gives the freed blocks back to the pool. A CT/internal/stacks/image_retention.go:188-193, AG/internal/fstrim/fstrim.go:1-12
5 New installs are refused; backups refuse app by app and delete nothing; running apps can fail to write. Data already written stays. A CT/internal/api/router.go:477, AR/07-backup-architecture.md:1567-1572
6 The controller, each 10 minutes: warning at 85 % or 5 GiB free, critical at 95 % or 2 GiB, to the household. Before full. The watch has two targets (/ and /mnt/sys_drive) that are ONE filesystem on all three boxes, and it de-duplicates by path — so one fill can send two warnings (C4; not proved live, a disk may not be filled in this run). A CT/internal/fillwatch/fillwatch.go:40-58, CT/cmd/controller/main.go:2745-2749, EV/extra/controller-fs-view.txt
7 No self-repair. No runbook for a full data volume. N C2 · S6
8 SOURCE and MEASURED. A CT/internal/fillwatch/fillwatch.go:40-58

G3 — App images, with one undo image per app · CUSTOMER · NAMED

AR/09-update-architecture.md:687 (decision 53). All images together (overlay2): hp 24.28 GB, 22 tagged + 21 untagged · N100 2.18 GB, 11 + 5 · T1 8.30 GB, 12 + 12.

# Answer Mark Pointer
1 The image each app runs and the image it ran before (for the undo button). A AR/09-update-architecture.md:687
2 Each app install; each update adds one image and makes the oldest free to go. A CT/internal/stacks/update.go:1063
3 A count, not a size: current + previous for each app. NONE — chosen for the total bytes (decision 53). A AR/09-update-architecture.md:687
4 The controller, after each update and each removal. A CT/internal/stacks/image_retention.go:188-193, CT/internal/stacks/update.go:1063
5 As G2. A CT/internal/api/router.go:477
6 As G2. A CT/internal/fillwatch/fillwatch.go:40-58
7 Yes: the next pass deletes what is no longer kept. A CT/internal/stacks/image_retention.go:370
8 TEST. A CT/internal/stacks/image_retention_test.go:111

G4 — Old controller images · CUSTOMER · NAMED

AR/09-update-architecture.md:710 (decision 56). Bytes are inside G3's figure.

# Answer Mark Pointer
1 The controller's own image and the one before it. A AR/09-update-architecture.md:710
2 Each controller update. A CT/internal/stacks/controller_image_retention.go:13
3 Running + previous. A CT/internal/stacks/controller_image_retention.go:25-30
4 The controller, after its own update. A CT/internal/stacks/controller_image_retention.go:49-57
5 As G2. A CT/internal/api/router.go:477
6 As G2. A CT/internal/fillwatch/fillwatch.go:40-58
7 Yes: the next pass. A CT/internal/stacks/controller_image_retention.go:49-57
8 TEST. A CT/internal/stacks/controller_image_retention_test.go:36

G5 — Infrastructure images (traefik and the other pinned helpers of the box) · CUSTOMER · NAMED

AR/09-update-architecture.md:1235 (decision 285, 2026-11-09). Bytes are inside G3's figure.

# Answer Mark Pointer
1 The images of the box's own services. A CT/internal/stacks/infra_image_retention.go:14-16
2 Each change of a pinned version. A CT/internal/stacks/infra_image_retention.go:14-16
3 The pinned one + the previous. A CT/internal/stacks/infra_image_retention.go:20-25
4 The controller. A CT/internal/stacks/infra_image_retention.go:61
5 As G2. A CT/internal/api/router.go:477
6 As G2. A CT/internal/fillwatch/fillwatch.go:40-58
7 Yes: the next pass. A CT/internal/stacks/infra_image_retention.go:61
8 TEST. A CT/internal/stacks/infra_image_retention_test.go:42

G6 — Helper images (alpine, hello-world) · CUSTOMER · MENTIONED

AR/09-update-architecture.md:2447. alpine:latest 8.42 MB on each box (4 months old on hp and N100, 7 weeks on T1) · hello-world:latest 10 kB on N100 and T1.

# Answer Mark Pointer
1 alpine: the small image the controller starts for a backup, a restore, an undo, an export (seven call sites). hello-world: left by the template bake. A CT/internal/backup/backup.go:988, AG/configs/build-golden.sh:232
2 Pulled once, at first use. Never again. A EV/extra/hp-9201.txt
3 No version is pinned: the box runs the alpine that was newest on the day of its first pull, for ever. N C5 · S7
4 Nobody. None of the three image rules can reach it: they look only at app, controller and infrastructure repositories. N C5 · S7 · CT/internal/stacks/image_retention.go:244-250
5 8 MB; as G2. A EV/extra/n100-9201.txt
6 As G2. A CT/internal/fillwatch/fillwatch.go:40-58
7 Not a space question at 8 MB. A EV/extra/t1-9201.txt
8 MEASURED and SOURCE. A EV/extra/t1-9201.txt

G7 — Named app volumes · CUSTOMER · NAMED

AR/07-backup-architecture.md:600-605. All volumes together: hp 1.61 GB · T1 0.37 GB. Named ones: hp 27 · N100 7 · T1 8.

# Answer Mark Pointer
1 Each app's own data that is not on a drive: databases, settings. The household's. A AR/07-backup-architecture.md:600-605
2 The use of the app. A EV/reading-1/hp-9201.tsv
3 No limit per app; the limit is G2's. A CT/internal/system/dockervol.go:14-26
4 Only a removal of the app deletes them; a stop keeps them. A CT/internal/stacks/delete.go:763-769, CT/internal/stacks/manager.go:1455-1459
5 As G2. A CT/internal/api/router.go:477
6 As G2. A CT/internal/fillwatch/fillwatch.go:40-58
7 No self-repair, no runbook (the same missing runbook as G2). N C2 · S6
8 SOURCE and MEASURED. A CT/internal/stacks/delete.go:763-769

G8 — Unnamed (anonymous) volumes · CUSTOMER · NAMED

AR/07-backup-architecture.md:600-605. hp 1,887 (1,885 unused) · N100 1 · T1 380 (377 unused). hp had 1,157 on 2026-10-15: about 29 more each day. T1 had 132: about 10 a day.

# Answer Mark Pointer
1 A volume an image declares by itself and the app definition does not name. Mostly empty. A AR/07-backup-architecture.md:600-605
2 Each recreate of such a container makes a new one. K R-1267
3 None. K R-1267
4 Nobody; a removal leaves them too. K R-995
5 They are small today. Nobody measured the effect of many thousands. K R-1267
6 As G2 (by bytes only; nothing counts them). A CT/internal/fillwatch/fillwatch.go:40-58
7 No. K R-995
8 MEASURED, three readings. A EV/reading-1/hp-9201.tsv

G9 — Container writable layers, with the controller's off-site cache · CUSTOMER · UNNAMED

The controller's own layer: 1.3 MB / 0.3 MB / 0.5 MB. Inside it /root/.cache/restic: 3.4 MB / 0.8 MB / 1.0 MB.

# Answer Mark Pointer
1 What a container writes outside its volumes. In the controller: the index cache of the off-site backup program. A EV/extra/hp-9201.txt
2 The app's own temp files; the off-site cache grows with the size of the off-site store. A EV/extra/n100-9201.txt
3 No limit of its own; G2's. No setting places or caps the off-site cache. A EV/searches.txt
4 Each recreate of the container drops the layer — so each controller update empties the cache. A EV/extra/t1-9201.txt
5 As G2. A CT/internal/api/router.go:477
6 As G2. A CT/internal/fillwatch/fillwatch.go:40-58
7 Yes: a recreate. A EV/extra/hp-9201.txt
8 MEASURED. No document and no source line names the cache (S10). A EV/extra/hp-9201.txt

G10 — Container logs · CUSTOMER · MENTIONED

AR/11-os-updates.md:107. hp 26.4 MB, 27 files, 26 containers · N100 9.7 MB, 9 containers · T1 7.3 MB, 10 containers.

# Answer Mark Pointer
1 What each container prints. A EV/reading-1/hp-9201.tsv
2 Each line an app prints. A EV/reading-1/n100-9201.tsv
3 10 MB × 3 files for each container: at most 780 MB (hp), 270 MB (N100), 300 MB (T1). Each container on each box carries the cap. A AG/configs/build-golden.sh:193-203, EV/raw/hp-9201.txt
4 Docker rotates the files. A AG/configs/build-golden.sh:193-203
5 Bounded; as G2. A CT/internal/api/router.go:477
6 As G2. A CT/internal/fillwatch/fillwatch.go:40-58
7 The cap is written once, when the template is baked. Nothing checks it on a running box; the one later edit of that file touches another key only. N C5 · S12 · AG/configs/felhom-os-apply:769-773
8 MEASURED on each container. A EV/raw/t1-9201.txt

G11 — Docker build cache · CUSTOMER · UNNAMED

hp 0.18 MB; the other two boxes about 0.

# Answer Mark Pointer
1 What an image build leaves. A box builds nothing. A EV/reading-1/hp-9201.tsv
2 Nothing on a household box. A EV/reading-1/n100-9201.tsv
3 No limit; it holds nothing. A EV/reading-1/t1-9201.tsv
4 Nothing to clean. A EV/reading-1/hp-9201.tsv
5 As G2. A CT/internal/api/router.go:477
6 As G2. A CT/internal/fillwatch/fillwatch.go:40-58
7 Not needed. A EV/reading-1/n100-9201.tsv
8 MEASURED. A EV/reading-1/t1-9201.tsv

G12 — The controller's own state volume · CUSTOMER · MENTIONED

AR/_recovery-inventory-2026-07-28.md:485. hp 145 MB · N100 84 MB · T1 80 MB. Largest file metrics.db: 120 / 41 / 55 MB. Catalog cache: 18 / 38 / 19 MB.

# Answer Mark Pointer
1 The box's settings, the usage graphs' database, the cached app catalog. A EV/extra/controller-state-volume.txt
2 One graph row per container per sample; the catalog. A EV/extra/controller-state-volume.txt
3 Graph rows older than 30 days are deleted. A CT/cmd/controller/main.go:1755, CT/internal/metrics/store.go:311-315
4 The controller. A CT/cmd/controller/main.go:1755
5 As G2: the install reserve exists so the box's own services can still write. A CT/internal/system/dockervol.go:14-26
6 As G2. A CT/internal/fillwatch/fillwatch.go:40-58
7 No self-repair, no runbook (the same missing runbook as G2). N C2 · S6
8 SOURCE and MEASURED. A EV/extra/controller-state-volume.txt

G13 — App definitions /opt/docker/stacks · CUSTOMER · MENTIONED

AR/_recovery-inventory-2026-07-28.md:494. hp 1.4 MB, 69 folders · T1 1.8 MB, 68. It lies on the guest root disk (G1).

# Answer Mark Pointer
1 One folder per catalog app: the definition files, and app.yaml when the app is installed. A CT/internal/stacks/delete.go:532-536
2 The catalog's size. A EV/reading-1/hp-9201.tsv
3 The catalog: under 2 MB. A EV/reading-1/t1-9201.tsv
4 A removal deletes app.yaml; the template files stay on purpose. A CT/internal/stacks/delete.go:532-536
5 As G1. A AR/11-os-updates.md:302-306
6 Nobody, as G1. N C4 · S9
7 As G1. N C2 · S6
8 MEASURED. A EV/extra/controller-fs-view.txt

G14 — Backups of apps with no drive (sys_drive) · CUSTOMER · NAMED

AR/07-backup-architecture.md:1567-1572. hp 1.33 GB, 14 apps · N100 48 MB, 5 · T1 270 MB, 4. On the data volume (G2).

# Answer Mark Pointer
1 The nightly backup unit of each app that has no data drive. The household's. A AR/07-backup-architecture.md:1567-1572
2 The apps' database and volume dumps. A EV/reading-1/hp-9201.tsv
3 One point per app. A backup is refused at 97 % used or under 1 GiB free. A AR/07-backup-architecture.md:398, CT/internal/backup/recovery_unit.go:383-388
4 Each night's backup replaces the app's unit. A unit left on a place the app moved away from is pruned. A CT/internal/backup/backup.go:1552
5 The backup refuses app by app and deletes nothing; the last good unit stays. A CT/internal/backup/recovery_unit.go:412-420, AR/07-backup-architecture.md:1567-1572
6 As G2 (this is the watch's second target). A CT/cmd/controller/main.go:2745-2749
7 No self-repair, no runbook (the same missing runbook as G2). N C2 · S6
8 SOURCE and MEASURED. A CT/internal/backup/recovery_unit.go:383-388

G15 — Guest system log · CUSTOMER · UNNAMED

The guest has no /var/log/journal, so its journal lives in memory (/run): 8 MB. It is empty after each guest restart.

# Answer Mark Pointer
1 The guest's own system log (Docker's service log is in it). A EV/extra/hp-9201.txt
2 Each log line of the guest's services. A EV/extra/n100-9201.txt
3 journald's default for a log in memory. Nobody chose "in memory": no source line and no document sets it, and a restart loses the log. N C5 · S14
4 journald; and each restart. A EV/os-cleaners/hp-9201.txt
5 The oldest lines go. A EV/os-cleaners/n100-9201.txt
6 Nobody. It is memory, not disk. N C4 · S9
7 It is bounded by itself. A EV/os-cleaners/t1-9201.txt
8 MEASURED. A EV/extra/t1-9201.txt

G16 — Guest package cache and lists · CUSTOMER · MENTIONED

AR/11-os-updates.md:293. hp 96 MB + 96 MB, 0 package files; the same shape on the other two.

# Answer Mark Pointer
1 As H11, inside the guest. A EV/reading-1/hp-9201.tsv
2 Each guest OS update. A AR/11-os-updates.md:293
3 Emptied at the end of each run. A AG/configs/felhom-os-apply:1417
4 The update wrapper. A AG/configs/felhom-os-apply:1417
5 As G1. A AR/11-os-updates.md:302-306
6 Nobody, as G1. N C4 · S9
7 Yes: cleaned each run. A AG/configs/felhom-os-apply:1417
8 MEASURED (0 package files). A EV/reading-1/t1-9201.tsv

G17 — Guest /tmp and /var/tmp · CUSTOMER · UNNAMED

/tmp is memory. Both near 0 today. An app export is written to a temp folder first.

# Answer Mark Pointer
1 Short-lived files. A EV/extra/hp-9201.txt
2 Nothing lasting on the guest. A EV/os-cleaners/hp-9201.txt
3 The tmpfs default. A EV/os-cleaners/n100-9201.txt
4 systemd's clean-up timer; each restart. A EV/extra/hp-9201.txt
5 Memory, not disk. Not written anywhere what then fails. N C2 · S6
6 Nobody. N C4 · S9
7 Yes: the timer and each restart. A EV/extra/t1-9201.txt
8 MEASURED. A EV/extra/n100-9201.txt

Data drive (/mnt/felhom-drives/<drive>/) — 10 places

Each folder below lies on a registered drive. The drive's alarm is the controller's fill watch (warning 85 %, critical 95 %, to the household). Drives today: hp hdd_1 · N100 hdd_1 (8.14 of 983 GB used) · T1 adatlemez (13.11 of 52.5 GB used).

D1 — backups/primary — the app's own backup unit · CUSTOMER · NAMED

AR/07-backup-architecture.md:398. hp 372 MB, 7 apps · N100 1.39 GB, 5 · T1 100 MB, 1.

# Answer Mark Pointer
1 The nightly backup unit of each app that has a drive. The household's. A AR/07-backup-architecture.md:398
2 The apps' database and volume dumps. A EV/reading-1/n100-9201.tsv
3 One point per app. The unit of a removed app, when the household kept it, has no limit: NONE — chosen. A AR/07-backup-architecture.md:398, AR/07-backup-architecture.md:380
4 The nightly backup replaces it; a removal with "delete the backups too" deletes it. A CT/internal/backup/backup.go:1552, CT/internal/stacks/delete.go:532-536
5 The backup of that app is refused; the last good unit stays byte for byte. A CT/internal/backup/recovery_unit.go:412-420
6 The drive's fill watch, to the household, before full. A CT/cmd/controller/main.go:2750-2757
7 No self-repair. No runbook for a full data drive. N C2 · S6
8 SOURCE and MEASURED. A CT/internal/backup/recovery_unit.go:383-388

D2 — backups/secondary — the copy on a second drive · CUSTOMER · NAMED

AR/07-backup-architecture.md:399. hp 920 MB, 8 apps · N100 294 MB, 4 · T1 13.01 GB, 8 apps (none of the 8 is installed today; one is 12.49 GB).

# Answer Mark Pointer
1 A full second copy of each app's unit and files, on another drive. A AR/07-backup-architecture.md:399
2 The apps' data. A EV/reading-1/t1-9201.tsv
3 The copy needs max(2 GB, 20 %) free on the target. A copy whose app is gone is kept with no limit: NONE — chosen (D25). A CT/internal/backup/tier2.go:80-84, AR/07-backup-architecture.md:1731-1735
4 The mirror deletes what the source deleted. A copy of a removed app: only the household, from the backup page. A CT/internal/backup/tier2.go:433, CT/internal/backup/removed_units.go:114
5 The copy is refused under the headroom. A CT/internal/backup/tier2.go:80-84
6 The drive's fill watch. A CT/cmd/controller/main.go:2750-2757
7 No self-repair, no runbook (the same missing runbook as D1). N C2 · S6
8 SOURCE and MEASURED. A EV/raw/t1-9201.txt

D3 — backups/offsite-restore — copies fetched from the off-site store · CUSTOMER · MENTIONED

AR/_recovery-inventory-2026-07-28.md:126. N100 1.65 GB, 2 apps (1.49 GB + 163 MB; neither app is installed today) · hp 0 · T1 0.

# Answer Mark Pointer
1 The full copy an off-site restore or check downloads before it is used. A CT/internal/backup/offbox_verify_copies.go:10-14
2 Each off-site restore or check of an app. A EV/raw/n100-9201.txt
3 None: no age, no size. (The restore's undo copy next to it has 30 days — D5.) N C5 · S8
4 Only the household's delete button. A CT/internal/backup/offbox_verify_copies.go:111-115
5 As the drive: a new download is refused when it does not fit. A CT/internal/backup/offbox_proof.go:163-170
6 The drive's fill watch. A CT/cmd/controller/main.go:2750-2757
7 No self-repair, no runbook (the same missing runbook as D1). N C2 · S6
8 MEASURED: two copies stayed on the N100 after their apps were removed. A EV/reading-1/n100-9201.tsv

D4 — backups/offsite-proof — the nightly proof's scratch · CUSTOMER · MENTIONED

AR/07-backup-architecture.md:2020. 0 bytes on each box.

# Answer Mark Pointer
1 The copy the nightly proof downloads to show the off-site store still holds something. A AR/07-backup-architecture.md:2020
2 One app's unit, each night, for minutes. A CT/internal/backup/offbox_proof.go:163-170
3 A headroom check runs before the download. A CT/internal/backup/offbox_proof.go:171-183
4 The proof itself, at its end; it also removes a stale copy from an interrupted run. A CT/internal/backup/offbox_proof.go:171-183
5 The proof is refused before any download. A CT/internal/backup/offbox_proof.go:171-183
6 The drive's fill watch. A CT/cmd/controller/main.go:2750-2757
7 Yes. A CT/internal/backup/offbox_proof.go:171-183
8 SOURCE and MEASURED (empty). A EV/reading-1/hp-9201.tsv

D5 — backups/restore-undo — what a restore replaced · CUSTOMER · NAMED

AR/07-backup-architecture.md:764, AR/09-update-architecture.md:1203 (decision 281). The folder does not exist on any of the three boxes today.

# Answer Mark Pointer
1 The data a restore replaced, kept so the restore can be undone. A CT/internal/backup/offbox_restore_undo.go:55
2 Each restore: a full copy of the app's data. A CT/internal/backup/restore_copies.go:160
3 30 days when unused; no limit while a support hold exists (NONE — chosen, decision 281). A CT/internal/backup/restore_copy_sweep.go:27-37
4 The controller's sweep. A removal of the app with "delete the backups too" leaves it behind. K R-1268
5 As the drive. A CT/internal/backup/recovery_unit.go:412-420
6 The drive's fill watch; the household is told when a copy is deleted. A CT/internal/notify/notifier.go:1604
7 Yes: the sweep. A CT/internal/backup/restore_copy_sweep.go:27-37
8 SOURCE. A CT/internal/backup/restore_copy_sweep.go:27-37

D6 — kept — data kept at "remove the app, keep my data" · CUSTOMER · NAMED

AR/07-backup-architecture.md:990-1009, AR/09-update-architecture.md:545 (decision 40). hp 208 MB (one app, three dates of 2026-10-16) · none on the other two.

# Answer Mark Pointer
1 The removed app's data folder, with its backup unit. The household's. A AR/07-backup-architecture.md:1008
2 Each "remove, keep my data". A CT/internal/stacks/kept.go:45
3 NONE — chosen: kept data is never deleted by the product (decision 40). A AR/09-update-architecture.md:545
4 Only the household. A AR/09-update-architecture.md:545
5 As the drive. A CT/internal/backup/recovery_unit.go:412-420
6 The drive's fill watch. A CT/cmd/controller/main.go:2750-2757
7 No self-repair, no runbook (the same missing runbook as D1). N C2 · S6
8 DOC and MEASURED. A EV/reading-1/hp-9201.tsv

D7 — appdata — the apps' files on the drive · CUSTOMER · NAMED

AR/07-backup-architecture.md:996. hp 30 MB, 2 apps · N100 441 MB, 4 apps · T1 45 KB, 1.

# Answer Mark Pointer
1 The files an app keeps on the drive (photos, documents). The household's. A AR/07-backup-architecture.md:996
2 The use of the app. A EV/reading-1/n100-9201.tsv
3 The drive. No limit per app. A EV/reading-1/n100-9201.tsv
4 Only a removal "with data" deletes it. A AR/07-backup-architecture.md:1002-1005
5 The app's own write fails. The drive's backups refuse app by app and delete nothing. A AR/07-backup-architecture.md:1567-1572, CT/internal/backup/recovery_unit.go:412-420
6 The drive's fill watch, before full. A CT/cmd/controller/main.go:2750-2757
7 No self-repair, no runbook (the same missing runbook as D1). N C2 · S6
8 DOC and MEASURED. A EV/reading-1/hp-9201.tsv

D8 — userdata — the household's own folders · CUSTOMER · NAMED

AR/07-backup-architecture.md:1002-1005, AR/09-update-architecture.md:810-811 (decision 67). hp 6 MB · N100 0.6 MB · T1 0.1 MB (test boxes hold almost nothing here).

# Answer Mark Pointer
1 The household's films, music, documents, import folders. A AR/07-backup-architecture.md:1002-1005
2 The household. A EV/reading-1/n100-9201.tsv
3 NONE — chosen: it stays on every removal (decision 67). A AR/09-update-architecture.md:810-811
4 Only the household. A removal leaves the app's empty import folder. K R-995
5 As D7. A CT/internal/backup/recovery_unit.go:412-420
6 The drive's fill watch. A CT/cmd/controller/main.go:2750-2757
7 No self-repair, no runbook (the same missing runbook as D1). N C2 · S6
8 DOC and MEASURED. A EV/reading-1/t1-9201.tsv

D9 — shares — folders shared on the home network · CUSTOMER · MENTIONED

AR/_recovery-inventory-2026-07-28.md:432. N100 one share, 32 KB · none on the other two.

# Answer Mark Pointer
1 The folders the household shares to its computers. A CT/internal/web/sharing_handlers.go:390-400
2 The household's copies. A EV/reading-1/n100-9201.tsv
3 The drive. No limit per share. A CT/internal/web/sharing_handlers.go:390-400
4 Only the household. A CT/internal/web/sharing_handlers.go:390-400
5 As D7. A CT/internal/backup/recovery_unit.go:412-420
6 The drive's fill watch. A CT/cmd/controller/main.go:2750-2757
7 No self-repair, no runbook (the same missing runbook as D1). N C2 · S6
8 SOURCE and MEASURED. A EV/reading-1/n100-9201.tsv

D10 — backups/set-aside — the backup of an earlier install of the same app · CUSTOMER · NAMED

AR/07-backup-architecture.md:1793-1800, AR/09-update-architecture.md:1344-1346 (decision 256). The folder does not exist on any of the three boxes today.

# Answer Mark Pointer
1 The backup unit of an earlier install, moved aside when the app is installed again. A AR/07-backup-architecture.md:1793-1800
2 Each re-install of an app that still had a unit. A AR/07-backup-architecture.md:1793-1800
3 NONE — chosen: kept until the household removes it (decision 256). A AR/09-update-architecture.md:1344-1346
4 Only the household, or the app's removal with "delete its backups too". A AR/07-backup-architecture.md:1793-1800
5 A move that cannot be done refuses the write and leaves the old unit as it was. A AR/07-backup-architecture.md:1793-1800
6 The drive's fill watch. A CT/cmd/controller/main.go:2750-2757
7 No self-repair, no runbook (the same missing runbook as D1). N C2 · S6
8 DOC only: no box holds one today. A AR/07-backup-architecture.md:1793-1800

TEST-ONLY places — measured, not counted

Place Box Size today Why it is not a customer place
Scratch guest 9202: data volume, images, volumes hp 13.99 GB; 12.79 GB images; 368 volumes (363 unnamed, all unused) the test guest (EV/reading-1/hp-9202.tsv)
Drive scratch_hdd hp — the scratch guest's drive
Test VM disks and installer images on /mnt/hdd_1 hp the filesystem holds 322.7 GB Tester 1 and the other test VMs live here
Proxmox template cache /var/lib/vz/template hp, N100 130 MB each (12 KB on T1) only a box that bakes the guest template downloads one
Sampler scripts in /var/tmp hp 4 small files of 2026-10-14 left by a test session
traefik/whoami image T1 6.6 MB a test tool
alpine:3.20, curlimages/curl images hp removed by hand on 2026-11-09 test tools, not the product (see README.md, question 1)

DOC-ONLY places — DooPlex: not measured — operator ruling 2026-11-10 · ep0 and the off-site box: not yet

Not measured in this run (no login to ep0, no inventory of DooPlex, no read of the hub database). Operator ruling 2026-11-10 12:22 (AR/09-update-architecture.md §3 decision 290): the gap review does not measure DooPlex at all (O4, O5, O6), and does not measure ep0 yet (O2, O3). The marks are candidates from documents and source; they are NOT in MATRIX.md. A helper filled them; the main session did not read each pointer.

Place What the documents say Candidate gaps
O1 — the shared off-site box (Hetzner Storage Box, restic) — not yet keep 7 daily, 4 weekly, 6 monthly; hub warns at 80 % / 90 % of the box (AR/07-backup-architecture.md:400) limit and "what when it fills": R-540; a weekly clean-up refused by its own guard: R-1294
O2 — ep0's backup store (PBS datastore) — ep0: not yet a 100 GB volume, 13 % used on 2026-07-27; the server prunes what happens when it is full; a runbook for that
O3 — ep0's own system disk — ep0: not yet no document found grow, limit, full, alarm, runbook — all open; clean-up: R-91
O4 — DooPlex's copy of ep0's store — DooPlex: not measured — operator ruling 2026-11-10 8 weekly copies (decision 71) what happens when full; who is alarmed
O5 — the hub's database and its snapshots — on DooPlex: not measured — operator ruling 2026-11-10 rows 90 days; the newest 2 snapshots; a 4 GiB volume (E/manifests/hub.yaml:58) what happens when full; who is alarmed
O6 — the image registry and the release copy — on DooPlex: not measured — operator ruling 2026-11-10 one image per version for ever; pruned only by hand (decision 62 — decided, not built as a job) what happens when full; who is alarmed; R-1235, R-1259