Agent 0.159.1, the catalog allow-list (controller 0.328.0), break-glass rotation: report, register, design
gates / gates (push) Successful in 8m49s
gates / gates (push) Successful in 8m49s
Closed R-971, R-954, R-1124, R-1012 (174 -> 175 rows); opened R-1139..R-1143; R-895 READY with the design and hub slice 1; R-972 waits on the vouch. [DESIGN] "The catalog allow-list" in 02; the catalog trust line in 01; FELHOM_THINCHECK and unit-forget in 03. Evidence audits/agent-allowlist-rotation-2026-10-17/. Claude-Session: https://claude.ai/code/session_0159rPz1ZhFKsS53msqPYxtS
This commit is contained in:
1 parent
d968f8490e
commit
5a5486104a
36 files changed
+816
-16
No files matched your search
@@ -2,7 +2,7 @@
|
||||
|
||||
**Ready for the first real tester (Tester-2): yes. Tester 2 (a laptop) is off; nothing was sent to it.**
|
||||
|
||||
**Updated 2026-10-17 (afternoon, later).** Controller 0.327.1 is the floor for every box since 12:06 UTC: FileBrowser is published only after its default password is checked (your D11 = A), and the install check uses real memory amounts with a clear message (your D10 = A) — `REPORT-filebrowser-memory-ci-2026-10-17.md`. **Earlier:** Controller 0.326.0 is the floor for every box since 09:55 UTC: FileBrowser's default password is checked at every FileBrowser start. Tester 1 is back on the live app list (test apps and drives removed); six memory limits raised — `REPORT-tester1-limits-filebrowser-2026-10-17.md`. **Earlier today:** controller 0.325.0 was the floor since 08:21 UTC (after the three test boxes): two more password doors count a try before checking it, family sign-ins are stored as fingerprints, a file-share rename always applies; five apps' Redis saves snapshots now, which a full disk cannot damage — `REPORT-burndown-3-2026-10-17.md`. **Earlier (2026-10-17 night):** controller 0.324.0 was the floor for every box since 02:46 UTC (all three test boxes made tonight's off-site backup on it first): a restore now brings the database password back with the data (your 1A), restic 0.19.1 (your 2A), dashboard hardening — `REPORT-night-2026-10-17.md`. **Earlier (2026-10-16 evening):** 0.323.0, Debian 13 — `REPORT-debian13-restic-2026-10-16.md`. **Earlier (midday):** Controller 0.319.0 (the app page in panels, the login in one panel) is the floor, MinAgent 0.158.1 — `REPORT-app-page-tuning-2026-10-16.md`. **Earlier (morning):** Storage identity build B shipped: agent 0.158.1 and golden 0.318.0 vouched by you, controller 0.318.0 is the floor (MinAgent 0.158.1), installer 1.33.0 published — new one-disk boxes get their own data area. Open items: see `documentation/backlog/OPEN-ITEMS.md`. Older notes, day by day:
|
||||
**Updated 2026-10-17 (evening).** Controller 0.328.0 is the floor for every box (17:22 local): the box refuses a catalog change that asks for dangerous container rights (your D2 = A). Agent 0.159.1 (a gone, decommissioned drive is forgotten; an orphan thin snapshot is named in the backup failure; new Go) — runs on demo-hp, Tester 1 and demo-felhom; **your vouch in the hub is still open**. New emergency root passwords on demo-hp, Tester 1, demo-felhom (your D1 = A; read them in the hub). Hub 0.157.0 deployed on your yes — `REPORT-agent-allowlist-rotation-2026-10-17.md`. **Earlier:** **Updated 2026-10-17 (afternoon, later).** Controller 0.327.1 is the floor for every box since 12:06 UTC: FileBrowser is published only after its default password is checked (your D11 = A), and the install check uses real memory amounts with a clear message (your D10 = A) — `REPORT-filebrowser-memory-ci-2026-10-17.md`. **Earlier:** Controller 0.326.0 is the floor for every box since 09:55 UTC: FileBrowser's default password is checked at every FileBrowser start. Tester 1 is back on the live app list (test apps and drives removed); six memory limits raised — `REPORT-tester1-limits-filebrowser-2026-10-17.md`. **Earlier today:** controller 0.325.0 was the floor since 08:21 UTC (after the three test boxes): two more password doors count a try before checking it, family sign-ins are stored as fingerprints, a file-share rename always applies; five apps' Redis saves snapshots now, which a full disk cannot damage — `REPORT-burndown-3-2026-10-17.md`. **Earlier (2026-10-17 night):** controller 0.324.0 was the floor for every box since 02:46 UTC (all three test boxes made tonight's off-site backup on it first): a restore now brings the database password back with the data (your 1A), restic 0.19.1 (your 2A), dashboard hardening — `REPORT-night-2026-10-17.md`. **Earlier (2026-10-16 evening):** 0.323.0, Debian 13 — `REPORT-debian13-restic-2026-10-16.md`. **Earlier (midday):** Controller 0.319.0 (the app page in panels, the login in one panel) is the floor, MinAgent 0.158.1 — `REPORT-app-page-tuning-2026-10-16.md`. **Earlier (morning):** Storage identity build B shipped: agent 0.158.1 and golden 0.318.0 vouched by you, controller 0.318.0 is the floor (MinAgent 0.158.1), installer 1.33.0 published — new one-disk boxes get their own data area. Open items: see `documentation/backlog/OPEN-ITEMS.md`. Older notes, day by day:
|
||||
`documentation/backlog/STATUS-HISTORY.md`. Reports: `REPORT-night-2026-10-14.md`, `REPORT-immich-rollout-2026-10-14.md`, `REPORT-seo-felho-tarhely-2026-10-14.md`, `REPORT-immich-repair-2026-10-14.md`, `REPORT-power-cut-fixes-2026-10-14.md`, `REPORT-night-power-2026-10-13.md`, `REPORT-poweroff.md`.
|
||||
|
||||
## What works
|
||||
@@ -34,7 +34,7 @@
|
||||
- **One-disk boxes get their own data area** (2026-10-16): a new install makes a „Belső tárhely" on the disk, outside every whole-box copy, and media apps install there with no extra step (proven on a fresh 512 GB box: 213 GiB area, Jellyfin with no extra step). A disk split into parts can be used part by part. The backup offer no longer promises disk-failure safety when it cannot check the disk, and `/mnt/sys_drive` is no longer offered.
|
||||
|
||||
## What is broken
|
||||
- Several root-level weaknesses wait on your four decisions below (R-954, R-971, R-982, R-985).
|
||||
- Two root-level weaknesses wait on your decisions below (R-982, R-985). Tester 2 still has its old emergency password (R-1139, when it is back).
|
||||
- DooPlex: a failed disk with its alarms silenced (R-934); CI jobs still get lost (R-887); something sets the shared
|
||||
checkout to „bare" (R-935); a Longhorn volume at 91 %.
|
||||
- Tester 2 has never made an off-site check (R-862). The scratch box cannot read the drill catalog (R-992).
|
||||
@@ -42,27 +42,29 @@
|
||||
## What we work on
|
||||
- **CC's secrets** (R-1118): done 2026-10-17 — the store, `sec`, the guards in every CC session (option B: `bash -x` and `curl -v` allowed inside `sec run`), the nightly check (linger on). The old credentials file and its three copies are deleted.
|
||||
- **After 2026-10-18 14:16 UTC** (demo-hp's next clean-up window): check that the old restic (image 0.322.0) still lists and restores demo-hp's store after its first real 0.19.1 prune (`audits/night-restore-2026-10-17/tools/rx.sh`). Everything else of R-1105/R-1107 is done.
|
||||
- The power-cut findings still open (R-1004, R-1005, R-1010, R-1012, R-1013).
|
||||
- The power-cut findings still open (R-1004, R-1005, R-1010, R-1013). R-1012 closed: agent 0.159.1 names an orphan thin snapshot in the backup failure.
|
||||
- Media apps on every drive (R-1023): live with the 0.313.0 floor — every drive in every media app, a new drive made ready by the box, a pulled drive loses nothing, a safe removal lets go first and the drive comes back by itself.
|
||||
|
||||
## Waiting on you — decide
|
||||
|
||||
| # | Row | Question | Pick | The other option costs | If you do nothing | Blocked |
|
||||
|---|---|---|---|---|---|---|
|
||||
| D1 | R-954 | Change every box's break-glass root password, since we cannot see if the old key was used? | **Yes** — about one hour, one box at a time, through the vault | Accept: no work, a password that may have been read stays valid | the old root passwords stay; if they were read, they still open the boxes from the LAN or console | — |
|
||||
| D2 | R-971 | Should the box refuse a catalog change that asks for dangerous container rights? | **Yes — an allow-list on the box** (one controller release) | Keep trusting the catalog repo: whoever can push to it reaches every box | a catalog push keeps reaching every box's apps with no check | — |
|
||||
| D1 | R-954 | **DECIDED 2026-10-17: A** — new random emergency root passwords on demo-hp, Tester 1 and demo-felhom (your answer: random); each new one works, each old one is refused. Tester 2: R-1139. | done | — | — | — |
|
||||
| D2 | R-971 | **DECIDED 2026-10-17: A** — the box refuses a catalog change that asks for dangerous container rights (controller 0.328.0, every box). Only homepage's read-only Docker socket is allowed. A refusal keeps the app running and mails you. | done | — | — | — |
|
||||
| D3 | R-985 | „Zero-knowledge": change the wording, or remove the operator's root reach to the key? | **Change the wording now**; revisit after the first sale | Remove the reach: a design change to break-glass | the promise stays stronger than the product | the website text |
|
||||
| D4 | R-982 | Root SSH with a password from the LAN? | **No — key only**; the password stays for the console | Keep it: anyone on the household network can try it | the host keeps accepting root password logins from the LAN | — |
|
||||
| D5 | R-993 (R-991) | Approve the minimum machine (`audits/sizing-2026-10-13/SIZING.md` §5: x86-64, 4 cores, 8 GB, 256 GB SSD)? | **Approve** | Change it: one edit | no hardware page; a buyer with a Raspberry Pi can still download the installer | the hardware page |
|
||||
| D6 | R-992 | The scratch box cannot read the private drill catalog: give it a read-only login, or make that copy readable? | **A read-only login** | Readable copy: the drill catalog becomes public | every new-app test needs hand-staging | new-app tests |
|
||||
| D7 | R-990, R-989 | Rule yes / no / later on the forum ideas (`audits/forum-research-2026-10-12/FELHOM.md` §5) and the selfh.st ideas (`audits/selfhst-survey-2026-10-12/FELHOM.md` §6) | — | — | nothing changes; the research ages | — |
|
||||
| D8 | R-206 | Install DooPlex's Docker settings from git (restarts Docker), and prune the build cache weekly: yes or no? | **Yes, in a quiet window** | — | a rebuild of DooPlex still has the settings in git | — |
|
||||
| D13 | R-1140 | Keep the hub's access logs 90 days (today about 13 hours)? | **A — a small hub-only access log on the hub's own disk** (≈ 15 MB for 90 days) | B: longer logs for every container on DooPlex (≈ 1.3 GB compressed) | the next „was the key used?" question has about 13 hours of logs | — |
|
||||
| D10 | R-1125 | **DECIDED 2026-10-17: A** — the 23 memory requests now match real use (catalog `1af6ea7`); the install refusal and page say what to do (controller 0.327.0). An 8 GB box no longer installs the whole household set. | done | — | — | — |
|
||||
| D11 | R-1126 | **DECIDED 2026-10-17: A** — FileBrowser is published only after its default password is checked; it never answers on the default (controller 0.327.1). The tunnel still runs before the claim, by design (`01` §5). | done | — | — | — |
|
||||
| D12 | R-1129 | **CLOSED 2026-10-17: not a defect** — CC misread the CI log; the failure mail always worked. A forced test mail went to admin@felhom.eu at 10:48:54 UTC (CC saw it arrive). Your key change did no harm. | done | — | — | — |
|
||||
| D9 | R-1062 | The N100's USB drive is also its whole-box backup drive. Split it in two anyway? | **No — keep one drive**; the lived-in box keeps its nightly whole-box backup | Split: the box loses its local whole-box backup (or add a second drive first) | it stays one drive; multi-drive tests stay on Tester 1 | — |
|
||||
|
||||
## Waiting on you — do
|
||||
- **Vouch agent 0.159.1** in the hub (Configs → Day-0 artifacts: 0.159.1, sha256 `de8f6ad9…aff9`, bundle `2af3d27d…7cd6`). The three test hosts already run it. If you do not: a new box installs 0.158.1.
|
||||
- **Nézd meg a demo-hp-n a Calibre-Web „Adatok és mentés" fülét és a Fájlkezelőt** (`felhom.enkisfelhom.hu`, controller 0.321.0, minden dobozon kiadva). A mentés egy részben, egyszerű szavakkal; a Fájlkezelő nem kér frissítést és nem ajánl „Install app"-ot.
|
||||
- **Look at `roms/roms/` on the N100's drive in the file manager (R-1099).** RomM now scans `roms/<platform>/`. A game uploaded through RomM before today sits one folder deeper: if there are platform folders in `roms/roms/`, move them up into `roms/` and press „Scan” in RomM. If not: those games show under a platform called „roms”.
|
||||
- **Nézd meg a Calibre-Web oldalát a demo-hp-n** (`felhom.enkisfelhom.hu`, controller 0.319.0, minden dobozon kiadva). A fejléc, a fülsor és minden rész most külön panel (az „A” stílus). A belépési adatok egy panelben vannak, egymás mellett. A Calibre-Web jelszaváról a doboz nem tud, mert az alkalmazás előbb települt, mint hogy a sablonban jelszó lett volna — ezt most egy mondat mondja ki. Ha nem tetszik, írd meg, mit változtassak. Ha nem nézed meg: semmi nem romlik el.
|
||||
|
||||
Reference in new issue
Block a user