plan for October hardware slip: presume 8 vCPU/48G, multiqueue bounce script

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]

crush.md

@@ -14,6 +14,11 @@
   ALLOWED (old curl ban retired). NO installs on the host (apt-get/pip/etc).
 - Missing CLI tool? docker pull it, standing permission — run via
   `docker run --rm <pinned image>`. Work in containers.
+- PUSH AS YOU GO (ruling 2026-08-31): commit+push work as it lands; the
+  human monitors gitea/redmine/discourse activity, not this host.
+- pfv-tsys5 (ruling 2026-08-31): crush may reboot OTHER VMs there anytime
+  (qm); VM 5111 reboots are human-triggered (crush lives here); host never
+  reboots casually; hardware window ~Oct 2026 — presume 8 vCPU / 48G.
 
 ## Output format (user preference, 2026-08-31)
 - The crush TUI sidebar truncates long chat output. NEVER hand the user a long