orchestrator-migration: Phase A started — cc-ci-orchestrator VM created (2GB/2vCPU)

Operator go-ahead (Pi is OOM-thrashing/slow). Created the dedicated loops VM
cc-ci-orchestrator (2GB RAM / 2 vCPU / 30GB, incus-base-vm NixOS) on b1 via the
Incus API, mirroring the known-good cc-nix-test spec; started it — cloud-init is
running nixos-rebuild boot + reboot + tailnet join. Status flipped DRAFT->IN
PROGRESS with the remaining Phase-A items noted (add cc-ci-root key via incus
exec, confirm tailnet+ssh, write the reproducible TF project).

Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
This commit is contained in:
2026-05-30 15:53:06 +01:00
parent 71a4a1fea4
commit 742a08b677

View File

@ -10,7 +10,13 @@ relocating this orchestrator session there too.
(claude CLI, proxy, loop supervisor) as systemd services that come back on boot — turning a reboot
into a non-event. It also consolidates the orchestrator next to the infra it manages.
**Status:** DRAFT — awaiting operator go-ahead before any infra creation / cutover.
**Status:** IN PROGRESS (operator go-ahead 2026-05-30 — the Pi is OOM-thrashing/slow).
**Phase A under way:** VM `cc-ci-orchestrator` (**2 GB / 2 vCPU / 30 GB**, `incus-base-vm`) CREATED via
the Incus API + started 2026-05-30; cloud-init → `nixos-rebuild boot` → reboot → tailnet join in
progress. Remaining Phase-A items: (i) add the orchestrator's `cc-ci-root` pubkey via `incus exec`
(create-time cloud-init only authorized the 2 TF-default keys), (ii) confirm tailnet + ssh, (iii)
write the reproducible Terraform project `projects/cc-ci-orchestrator/` for the record (created via
API this time — note the drift). Then Phase B (the `cc-ci-orchestrator` NixOS-config git repo).
---