Files
felhom.eu/documentation/audits/night-2026-09-25/PROGRESS.md
T
2026-09-25 04:54:52 +02:00

31 lines
3.1 KiB
Markdown

# PROGRESS — night 2026-09-25 (resume here)
Baselines verified 21:24 CEST: controller 1495ca43fb86 (v0.270.0), agent 4fe2f81a322b (v0.133.0), felhom.eu 75ff26408a3c, catalog c8025093d23c. Trees clean.
Capacity: phase0-capacity.txt. demo-felhom pool 3.13 %, root 31 %. demo-hp pool 59.01 % data / 2.65 % meta, root (local) 84.33 %. 9202 lives on nvme-scratch (dir, 7.6 %).
- [x] Phase 0 baselines + capacity
- [x] drill repo = live main c802509, actions off (phase0-drill-reset.txt); Part C commits on top
- [x] A1 demo-felhom 0.133.0 committed 21:35:09 CEST; hub reads 0.133.0
- [x] A1 demo-hp 0.133.0 committed 21:50:02 CEST; hub reads 0.133.0
- [x] A2 restore test on, both (agent.json.pre-r672 back; -1 copy kept as agent.json.night-0925-off)
- [x] A3 demo-felhom PASS 85 s pool 3.13→5.68→3.13; demo-hp REFUSED (needs 30.3 GiB, has 22.1), rc 4
- [x] A4 retention 3→1 (agent.json.pre-a4-retention), pvesm prune keep-last=1, backup by product trigger OK (8.18 GB local + PBS), R-685 filed
- [x] B controller v0.271.0 pushed 9cf13a3, image built+pushed, 11 red-proofs red
- register: R-685 added, R-676 updated; TO CLOSE after C: R-680, R-678, R-643, R-684, R-672/673 (delivered), R-450? (slice 6 automatic part)
- [x] C night1 (done 2, undone 1, skip needs_person), night2 (done 2, failed_before), night3 kill (step put back, leg not resumed R-686), night4 power cut (resumed, done), data read back every time
- [~] C night5 switch OFF; night6 switch ON (vikunja re-tested). 9202 window before = 02:30; controller.yaml.pre-night0925 saved
- [x] F agent v0.134.0 released 0722b2c/309e368 (sha 7593bebe…); INCIDENT: first build read GET /storage → real vzdump 9201 started 22:40:21, aborted 22:45:37, no archive (F1). Signed both boxes 20:49Z; demo-felhom committed 22:52:45
- register so far: closed R-672 R-673 R-684 R-680 R-678; opened R-685 R-686 R-687; R-676 updated. Before: 341 rows / 685,148 B
- [ ] D floor + real night
- [ ] E / F / G
- [x] C night5 switch off (nothing pressed), night6 on (vikunja re-tested → done). Data read back 8/8.
- [x] D1 floor 0.271.0 (MinAgent 0.131.0) at 20:59Z; D2 both boxes arrived ~15 s; D3 switch ON (absent), W 02:30, leg 04:15; D0: demo-felhom opengist 1.13→1.15 waiting; demo-hp none
- [x] F3 agent 0.134.0 on both boxes (hub reads 0.134.0)
- [x] E bench 9401 created+destroyed (template removed); n8n 2.41.1→2.41.2 and mealie v3.27.0→v3.28.0 PROVEN on bench AND box; control failed (correct). TODO after 02:30: --write-ladder both, commit each, push, CI. Then reset drill again.
- [x] G partial: 9202 apps removed via product, window 02:30, live catalog c802509 (switch key now explicit true), drill reset to live
- [ ] 02:31 publish E; 04:10 start D watch (d_watch.py) on both boxes
- [x] E published: n8n f14a608, mealie b996218 (CI 980/981), catalog CHANGELOG/REPORT dd9c6ad (982); drill reset to dd9c6ad
- [x] D real night: demo-felhom opengist done 20 s; demo-hp 0 steps; hub has both update_leg; no whole-box backup due on either (gate never waited)
- [x] G all three layers recorded (G1..G5); report DRILL-night-2026-09-25.md; STATUS; register 338 rows / 682,053 B
- unproven.py: NOT WALKED 35 of 55 (unchanged)