Two reboot-triggered session killers: xrdp waitforx RandR race/empty-argv
(shim wrapper; pristine = waitforx.real) and 90xbrlapi foreground hang
(hook disabled). new_cursors=false for the cursor box. Re-apply via
17-fix-rdp-postupdate.sh. Full diagnosis:
https://projects.knownelement.com/issues/611#note-4513💘 Generated with Crush
Assisted-by: Crush:glm-5.2
Full RDP clean slate requires killing session processes AND stopping
user@UID (sessions live in the system sesman cgroup - reconnect into a
survivor never restarts the user manager). 150% HiDPI + core cursors
persisted both accounts. Display-performance work (16K compressed-PDU
drops, staged 1MB send buffer, EGFX-at-24bpp retest) captured as #679.
Details: https://projects.knownelement.com/issues/679
Run outputs moved to logs/ (incl. sda partition-table backup). New from
the 15:57-16:03 parallel session: 16-grow-root-online.sh (RAN: / = 431G
online, swap removed by design — GROW-ROOT-RUNBOOK superseded) and
15-docker-to-data2.sh (staged, not run). README/NEXT/RUNBOOK/REBUILD
reflect current reality. Moot rdp-testprep timer removed from the VM.
Details: https://projects.knownelement.com/issues/615
The always-alive console session held the user's single plasmashell and
black-screened RDP. Live conf removed, staged/apply-guest.sh step now
removes it instead of writing it, REBUILD KDE policy updated. Console
shows the greeter at boot; RDP is the primary desktop; Sunshine will
get a kiosk account or XFCE when it lands.
Details: https://projects.knownelement.com/issues/611
With the server proven clean at 5160x2160 (band analysis + freerdp on
every codec), Jump still painted bars - its EGFX large-frame decode is
broken regardless of budget. drdynvc=false blocks EGFX for all clients;
clipboard/audio ride separate channels and survive. Classic codec
verified clean at 5160 via harness. Script part 3 + xrdp bounced.
Details: https://projects.knownelement.com/issues/611
Server-side bisect (session rendered pure black, colors=1) pinned it to
plasma-kwin_x11.service timing out and crash-looping: software-GL
compositing init under xorgxrdp never completes at 5160x2160. Fix:
Compositing=false for the RDP accounts (X11-only key, Wayland console
unaffected) + 300s kwin unit start timeout. 12-fix-rdp-bigres.sh now
installs both parts. freerdp harness matrix green across all four flows.
Details: https://projects.knownelement.com/issues/611
Full VMs run on the workstation (q35 + host-passthrough + KVM): Debian
cloud guest reached DHCP + ssh + running systemd. REGULAR-VMS.md holds
the recipe and the three workstation gotchas (disk perms under images/,
/boot-only kernels for virt-aa-helper, console=ttyS0 for serial).
Details: https://projects.knownelement.com/issues/615
vfio-pci owns 10de:06dd/10de:0be5 at next boot (IOMMU group 20 clean);
VM 5111 gains machine q35 + hostpci0 pcie=1, live at the human's
pre-Friday roll. Revert + verification steps in staged/gpu-passthrough.md.
Details: https://projects.knownelement.com/issues/606
Firecracker-class microVMs for dev: 13-microvm-setup.sh (idempotent
installer) + microvm CLI + microvm-net.service (per-user tap pools,
/30s, NAT). Debian 13 cloud assets in /usr/local/lib/microvm/base.
Firecracker v1.16.1 installed but its KVM cap check is refused by this
nested Westmere KVM; qemu -M microvm engine default, FC retry after #601.
Details + UAT: https://projects.knownelement.com/issues/615
Jump-from-iPad showed color bars/cut sides at the big external monitor.
Headless freerdp UAT proved the server clean at 5160x2160 on all codecs
and mid-session resize; mitigation raises the EGFX compressed-frame
budget via systemd drop-in, applied by 12-fix-rdp-bigres.sh (idempotent,
restart-guarded). Human UAT pending.
Detail: https://projects.knownelement.com/issues/611#note-3341
Verified after the 22:53 multiqueue bounce: NIC queues live at 4/4 and
2/2 (the whole point of the reboot), PVE config clean with nothing
pending, both gateway lanes healthy, all 9 Open Terminal ports bound on
the tailscale IP, PSI metrics fresh, and no removed relaunchers
respawned. Moved the day-profile flip from 05:00 to 07:00 per human
ruling (live timer re-armed + staged copy and docs synced; the gateway
z.ai peak ladder is untouched). Also staged a ready-to-run script that
drops each account's own Open Terminal key into its ~/.creds as
openwebui.creds - not run yet; the master key list stays at
~/.creds/open-terminal.env.
💘 Generated with Crush
Assisted-by: Crush:glm-5.2
[#610]
open-terminal 0.11.34 bare-metal (human-authorized host install) with a
systemd template unit: reachableceo on tailscale-only :30000 (cwd
~/projects) and all 8 other accounts on :30001-30008, each with its own
key kept out of ps in per-user TOMLs, all users added to the docker group.
Keys live in ~/.creds + /etc/ukrrs/open-terminal, never in this repo.
UAT: throwaway OpenWebUI v0.11.1 (matched to the human's Cloudron prod)
driven purely over its HTTP API against the beta gateway lane - connection
verify/config PASS, direct exec as uid 1001 PASS, and the chat round trip
PASS: the model emitted run_command, we executed it through OpenWebUI's
terminal proxy (server holds the key), and the final answer named 9/9 real
project directories; earlier negative runs prove the test catches
hallucinated output. Ops note OPEN-TERMINAL.md documents production wiring
(prod openwebui container v0.3.10 is too old and needs an upgrade first).
Also tonight: agent-stack relaunchers removed (6-, backups kept) so no
screen/crush sessions auto-start on reboot; 22:00 night-profile flip
observed live; docs synced for the earlier teardown.
💘 Generated with Crush
Assisted-by: Crush:glm-5.2
[#610]
REPORT.md gets a post-pass status banner + per-item STATUS tags + phase/QM
checklist truth; RUNBOOK-TODAY amended for the Oct window replan (Friday is
troubleshooting-only #609); all ~/optimize paths updated after the repo move
to ~/projects/ultix. Adds the 22:00 night-flip watcher log and
6-remove-agent-stacks.sh (boot-time screen/crush relauncher teardown with
backups to removed-agent-stacks/).
💘 Generated with Crush
Assisted-by: Crush:glm-5.2
[#602]
#601 moves to version october-2026-window (due Oct 31; Redmine API refuses
start_date on update, recorded in the note), #608 = cgroup_parent wiring
(the gateway starvation fix on 8 vCPU), #609 = Fri 2026-09-04
troubleshooting-only outage with the physical checklist.
💘 Generated with Crush
Assisted-by: Crush:glm-5.2
[#602]
Hardware window #601 slipped to ~Oct 2026: interim fences and daynight
values become the operating config, cpuset reservation defers with it, and
the compose cgroup_parent wiring becomes the gateway starvation fix. Adds
5-activate-multiqueue.sh — the one remaining VM-5111 reboot (human-timed)
to bring the staged net queues live. Tracking/README/crush prefs updated.
💘 Generated with Crush
Assisted-by: Crush:glm-5.2
[#602]
Everything for the ultix-streaming (VM 5111, pfv-tsys5) performance pass:
full report + host audit results, staged/gated configs, guest prep + host
one-shot + post-reboot-fix + netcheck lifecycle scripts, grow-root manual
runbook, rolling tracking HUD, questions v1, and the gateway boot-race
hardening units. Applied and verified live 2026-08-31; open work is
tracked in Redmine project 55 as #601-#607. [#602]
💘 Generated with Crush
Assisted-by: Crush:glm-5.2
mrcharles