Files
infrastructure/llm-server/llm/ansible/README.md
T
spektr d53b2400c9 LLM-Server: llm/ansible für Check-Modus und ansible-core 2.21 fit gemacht
Nach Live-Check-Modus gegen ghostnet-ki (ok=23, changed=4, failed=0):
- Freiraumprüfung auf existierendes Elternelement (df schlägt im Check-Modus
  auf dem noch nicht angelegten Store fehl)
- Rechte-Durchsetzung nur über vorhandene Artefakte (file-Modul im Check-Modus)
- assert ohne loop (2.21 lehnt loop auf assert ab); statische Meldungen,
  Modell-Anwesenheit im Check-Modus nur informell
- find file_type=any für Character-Devices (renderD*)
- README: Aufruf mit --private-key nach Workstation-Konvention

Preview prognostiziert korrekt: 2 Modellverzeichnisse, 2 SOURCE.yml,
Netz-Quadlet, 2 Container-Quadlets; gepinntes ROCm-Image bereits im Store.

Generated by Mistral Vibe.
Co-Authored-By: Mistral Vibe <vibe@mistral.ai>
2026-11-01 17:31:45 +01:00

62 lines
2.6 KiB
Markdown
Raw Blame History

This file contains ambiguous Unicode characters
This file contains Unicode characters that might be confused with other characters. If you think that this is intentional, you can safely ignore this warning. Use the Escape button to reveal them.
# LLM serving stack – Ansible control project
Executes the LLM serving stack for exactly one environment on
`ghostnet-ki` (GNAI-008 section 7.4). Host baseline, Podman runtime,
workload accounts and AMD GPU access are managed by
[`../../host/ansible`](../../host/ansible/README.md); this project never
changes host state.
## Layout
- `inventories/<environment>/` – one inventory per KI environment
(currently `testing`); deployment uses the `ansible-deploy` account with
`become: true`.
- `playbooks/ai.yml` – applies `model_store` then `inference_engine`.
- `roles/model_store/` – shared, pinned model weights under
`/var/lib/ghostnet-ai/shared/models` (GNAI-010 sections 3 and 4).
- `roles/inference_engine/` – rootless Quadlets, network, image provisioning
and the exclusive profile switch for llama.cpp (GNAI-010 sections 5 and 6).
## Control node setup
Use the repository-wide control venv built from
[`../../../shared/ansible/requirements-control.txt`](../../../shared/ansible/requirements-control.txt)
(`ansible-core==2.21.4`), for example:
```bash
python3 -m venv .venv
.venv/bin/pip install -r shared/ansible/requirements-control.txt
cd llm-server/llm/ansible
../../../.venv/bin/ansible-playbook -i inventories/testing playbooks/ai.yml --syntax-check
```
## Usage (Testing)
All commands use the operator deploy key
`~/.ssh/id_ed25519_ghostnet_ai_ansible_deploy` (see AGENTS.md; no ssh-agent
required):
```bash
# Preview only – no changes, no downloads, no service start:
../../../.venv/bin/ansible-playbook -i inventories/testing playbooks/ai.yml \
--check --diff --private-key ~/.ssh/id_ed25519_ghostnet_ai_ansible_deploy
# Provision model weights only (tag models):
../../../.venv/bin/ansible-playbook -i inventories/testing playbooks/ai.yml \
--tags models --private-key ~/.ssh/id_ed25519_ghostnet_ai_ansible_deploy
# Render quadlets and start/verify the selected profile (tag service):
../../../.venv/bin/ansible-playbook -i inventories/testing playbooks/ai.yml \
--tags service --private-key ~/.ssh/id_ed25519_ghostnet_ai_ansible_deploy
```
`inference_engine_active_profile` selects which rendered quadlet is started;
all other profiles stay rendered but stopped. Check mode downloads no model
and starts no service. A second identical run must report `changed=0`
(GNAI-010 section 10.4).
The full technical acceptance (64k/128k start, VRAM distribution, prefill
ladder, restart behaviour) is an operator-run protocol documented in
[GNAI-010](../../../docs/platform/llm-server/ai/GNAI-010-TECH_Llama-Serving.md);
this role automates deployment and basic health only.