Stephan Berbig 27b7012ec9
Some checks failed
CI / ruff (push) Has been cancelled
CI / mypy --strict (push) Has been cancelled
CI / pytest (push) Has been cancelled
CI / bandit (push) Has been cancelled
CI / pip-audit (push) Has been cancelled
scripts: wire-multi-backend.sh — one-shot multi-backend bootstrap
Drops the multi-step paste-by-paste setup down to one command. Detects
every running ollama/ollama container on the host, classifies them
(embedded vs extras based on which Docker network they share with the
gateway), checks each one's auth config, rewrites OLLAMA_BACKENDS in
.env atomically, recreates the gateway, and probes everything.

What it does, in order:
  1. preflight: docker on PATH, .env present, gateway container running
  2. `docker ps --filter ancestor=ollama/ollama` → every running Ollama
  3. for each one:
       - find the network it shares with the gateway; attach to `proxy`
         if it's on no shared network
       - read OLLAMA_AUTH and OLLAMA_AUTH_TOKEN from `docker inspect`
       - FAIL LOUD if auth=true with empty token (with the exact fix)
       - classify: on a `*internal*` net = the gateway's embedded backend;
         otherwise an extra
  4. build the OLLAMA_BACKENDS JSON: embedded first (so it gets routing
     priority), then extras in discovery order, with per-backend tokens
     where present
  5. write .env atomically (host-side temp + rename — no in-container
     perm issues to worry about)
  6. `docker compose up -d gateway` to pick up the new env
  7. wait for /healthz, then run probe-ollama and list-backends so the
     operator sees the end state immediately

Token is redacted before echoing the resulting OLLAMA_BACKENDS line, so
the script can run safely with terminal logging on.

Idempotent: re-running produces the same OLLAMA_BACKENDS line. No
half-states possible — every error path exits before .env is touched.
2026-05-27 23:15:59 +02:00
2026-05-26 20:52:33 +02:00
2026-05-26 20:52:33 +02:00
2026-05-26 20:52:33 +02:00
2026-05-26 20:52:33 +02:00

neuronetz-gateway

A secure, multi-tenant API gateway in front of an Ollama instance. It is the hot path of the Neuronetz API: every request to the models flows through here, authenticated, rate-limited, budgeted, and audited.

The Ollama backend is never reachable from the public internet. It is bound to an internal Docker network with no published ports. All access is via this gateway, behind TLS terminated by Caddy.

Status: v0.1.0 — in development. See scope-docs/SPEC.md for the full specification and scope-docs/AGENT_PROMPT.md for the phased build plan. SPEC.md is the source of truth.

What it does

  • Auth — API keys as Bearer tokens, stored as Argon2id hashes, verified in constant time.
  • Multi-tenant — tenants own keys; limits and budgets inherit tenant → key.
  • Rate limiting — per-key and per-tenant RPM / TPM / concurrent connections.
  • Budgets — daily / monthly / total token budgets, enforced fail-closed.
  • Dual API surface — native Ollama (/api/*) and OpenAI-compatible (/v1/*), both streaming.
  • Hard-blocked mutations/api/pull, /api/push, /api/create, /api/copy, /api/delete, /api/blobs/* always return 403. Not configurable.
  • Audit log — always-on request metadata; opt-in, TTL'd prompt logging per key.

Administration (dashboards, tenant self-service) lives in a separate service, neuronetz-console; it is not part of this repository.

Architecture

Internet ──TLS──> Caddy ──HTTP──> gateway ──┬──> Postgres   (keys, budgets, audit)
                                            ├──> Redis      (key cache, rate limits)
                                            └──> Ollama     (internal network only)

Quickstart (dev)

Requires Docker + Docker Compose. The dev stack runs Postgres, Redis, and the gateway — no Caddy and no Ollama (so /readyz reports 503 until a real Ollama backend is wired in; that is expected).

git clone <repo> neuronetz-gateway && cd neuronetz-gateway
cp .env.example .env          # adjust if you like; defaults work for local dev
docker compose -f docker-compose.dev.yml up --build

The gateway runs alembic upgrade head on startup, then serves on http://localhost:8080.

curl -i http://localhost:8080/healthz   # -> 200  {"status":"ok"}
curl -i http://localhost:8080/readyz    # -> 503  (no Ollama backend in the dev stack)

Production

docker-compose.yml brings up the full stack — Caddy (TLS via Let's Encrypt for api.neuronetz.ai), the gateway, Postgres, Redis, and Ollama. The ollama service has no ports: mapping and is reachable only on the internal Docker network. See docs/DEPLOYMENT.md (added in a later phase) and ops/caddy/Caddyfile.example.

Managing tenants and keys

Use the bootstrap CLI (Typer). Keys have the form nz_<prefix><secret>; the full key is printed exactly once at creation and only its Argon2id hash is stored.

neuronetz-gateway create-tenant --name acme
neuronetz-gateway create-key   --tenant acme --name prod-server-1
neuronetz-gateway list-keys    --tenant acme
neuronetz-gateway revoke-key   --prefix nz_abc12345

Development

just dev          # run the dev stack
just test         # pytest + coverage
just lint         # ruff
just typecheck    # mypy --strict
just migrate      # alembic upgrade head

Tooling: Python 3.12, uv, FastAPI + uvicorn, SQLAlchemy 2.0 (async) + asyncpg, Redis, httpx, structlog, Pydantic. Lint/type/security gates: ruff, mypy --strict, bandit, pip-audit.

License

Apache 2.0 — see LICENSE. Owner: Stephan Berbig / Neuronetz.

Description
AI API
Readme Apache-2.0 290 KiB
Languages
Python 86.2%
HTML 8.1%
Shell 4.4%
Dockerfile 0.9%
Just 0.4%