# Deploying the Bifrost portal — `bifrost-portal.fenja.ai`

This app runs as a **Node standalone SSR server** (Astro `@astrojs/node`) behind
the **existing nginx** on the Fenja VPS, coexisting with the other Fenja site.
nginx terminates TLS and reverse-proxies the `bifrost-portal.fenja.ai` hostname
to the app on `127.0.0.1:4322`. Data is a single SQLite file plus an uploads dir.

You run every step here. Nothing in this repo touches the live server on its own.

| Thing | Value |
|---|---|
| Hostname | `bifrost-portal.fenja.ai` |
| App bind | `127.0.0.1:4322` (loopback only) |
| Code | `/opt/bifrost-portal` (git checkout, built in place) |
| Persistent data | `/var/lib/bifrost-portal/` (`bifrost.db`, `uploads/`, `backups/`) |
| Service user | `bifrost` (unprivileged) |
| systemd unit | `bifrost-portal.service` |
| Env file | `/etc/bifrost-portal.env` (chmod 600) |

Artifacts referenced below all live in this repo: `deploy/bifrost-portal.service`,
`deploy/nginx/bifrost-portal.fenja.ai.conf`, `.env.production.example`,
`scripts/deploy.sh`, `scripts/backup.sh`.

---

## 0. Pre-flight — confirm the port is free

Another Fenja app already runs on this box. Make sure nothing holds `4322`:

```bash
sudo ss -ltnp | grep ':4322' || echo "4322 is free"
```

If it's taken, pick another loopback port and change it in **both** `/etc/bifrost-portal.env`
(`PORT=`) and `deploy/nginx/bifrost-portal.fenja.ai.conf` (`proxy_pass`).

## 1. DNS

Add an A/AAAA record `bifrost-portal` → the VPS IP, same target as the existing
Fenja site. Confirm before requesting a cert:

```bash
dig +short bifrost-portal.fenja.ai
```

## 2. One-time server provisioning

Run as a sudo-capable user. Assumes Node 22, pnpm, nginx, sqlite3, certbot, git
are already present (the existing Fenja app implies most are).

```bash
# Service user (no login shell, no home spam)
sudo useradd --system --create-home --home-dir /var/lib/bifrost-portal \
     --shell /usr/sbin/nologin bifrost

# Persistent data dirs
sudo install -d -o bifrost -g bifrost /var/lib/bifrost-portal
sudo install -d -o bifrost -g bifrost /var/lib/bifrost-portal/uploads
sudo install -d -o bifrost -g bifrost /var/lib/bifrost-portal/backups

# Code dir
sudo install -d -o bifrost -g bifrost /opt/bifrost-portal
```

Clone the repo into the code dir (uses the same `git.fenja.ai` remote this repo
already points at — make sure the `bifrost` user, or the user running deploys,
has a deploy key that can read it):

```bash
sudo -u bifrost git clone ssh://git@git.fenja.ai:2222/joh/project-bifrost-platform.git \
     /opt/bifrost-portal
```

## 3. Environment file

```bash
sudo cp /opt/bifrost-portal/.env.production.example /etc/bifrost-portal.env
sudo chown bifrost:bifrost /etc/bifrost-portal.env
sudo chmod 600 /etc/bifrost-portal.env
# Generate the session secret and paste it in as BIFROST_SECRET:
openssl rand -hex 32
sudo nano /etc/bifrost-portal.env
```

Make sure `BIFROST_DB_PATH`, `BIFROST_UPLOAD_DIR`, `HOST`, `PORT`, `NODE_ENV`
match the table above. `BIFROST_SECRET` is what signs sessions and invite
tokens — rotating it later logs everyone out and invalidates pending invites.

## 4. First build + database

```bash
cd /opt/bifrost-portal
sudo -u bifrost pnpm install --frozen-lockfile   # rebuilds better-sqlite3 for this arch
sudo -u bifrost pnpm build

# Create + migrate the production DB at BIFROST_DB_PATH:
sudo -u bifrost --preserve-env=BIFROST_DB_PATH \
     bash -c 'set -a; source /etc/bifrost-portal.env; set +a; node scripts/migrate.js'

# Seed the real pilot data (one time only — skip on later deploys):
sudo -u bifrost \
     bash -c 'set -a; source /etc/bifrost-portal.env; set +a; pnpm db:seed:production'
```

> `better-sqlite3` is native and must be built **on the server** for its CPU
> arch (ARM on a Hetzner CAX11). That's why we `pnpm install` here rather than
> copying `node_modules`. `scripts/deploy.sh` does this on every deploy.

## 5. systemd service

```bash
sudo cp /opt/bifrost-portal/deploy/bifrost-portal.service /etc/systemd/system/
sudo systemctl daemon-reload
sudo systemctl enable --now bifrost-portal
sudo systemctl status bifrost-portal        # should be active (running)
curl -fsS http://127.0.0.1:4322/login >/dev/null && echo "app responding"
```

Allow the `bifrost` user to restart just this unit without a password (used by
`deploy.sh`):

```bash
echo 'bifrost ALL=(root) NOPASSWD: /usr/bin/systemctl restart bifrost-portal, /usr/bin/systemctl status bifrost-portal' \
  | sudo tee /etc/sudoers.d/bifrost-portal
sudo chmod 440 /etc/sudoers.d/bifrost-portal
```

## 6. nginx + TLS

```bash
sudo cp /opt/bifrost-portal/deploy/nginx/bifrost-portal.fenja.ai.conf \
        /etc/nginx/sites-available/
sudo ln -s /etc/nginx/sites-available/bifrost-portal.fenja.ai.conf \
           /etc/nginx/sites-enabled/
sudo nginx -t && sudo systemctl reload nginx     # :80 block live for ACME
```

Issue the certificate (certbot edits the file to wire up the cert paths):

```bash
sudo certbot --nginx -d bifrost-portal.fenja.ai
sudo nginx -t && sudo systemctl reload nginx
```

Verify, then optionally enable HSTS (uncomment the `Strict-Transport-Security`
line in the nginx conf and reload) once you're happy HTTPS is solid:

```bash
curl -fsSI https://bifrost-portal.fenja.ai/login | head -n1
```

## 7. Nightly backups

```bash
sudo -u bifrost crontab -e
# add:
15 3 * * *  /opt/bifrost-portal/scripts/backup.sh >> /var/log/bifrost-backup.log 2>&1
```

`backup.sh` writes 30-day-retained `.backup` snapshots to
`/var/lib/bifrost-portal/backups`. Per SPEC §7.3, add a second cron line to sync
that dir to the Hetzner Storage Box (rclone/rsync) for offsite copies.

---

## Ongoing deploys

After pushing to `master` on `git.fenja.ai`:

```bash
sudo -u bifrost bash -c 'cd /opt/bifrost-portal && ./scripts/deploy.sh'
```

It pulls, installs (rebuilding native deps), builds, migrates, and restarts.
The DB and uploads are untouched.

## Rollback

```bash
cd /opt/bifrost-portal
sudo -u bifrost git log --oneline -n 10        # find the good commit
sudo -u bifrost bash -c 'BRANCH=<good-sha> ./scripts/deploy.sh'  # or: git reset --hard <sha> then deploy
```

Schema migrations are forward-only — a code rollback past a migration may need a
DB restore. To restore a backup (stop the app first):

```bash
sudo systemctl stop bifrost-portal
sudo -u bifrost bash -c 'gunzip -c /var/lib/bifrost-portal/backups/bifrost-YYYYMMDD-HHMMSS.db.gz > /var/lib/bifrost-portal/bifrost.db'
sudo systemctl start bifrost-portal
```

## Troubleshooting

- **502 from nginx** — app not running or wrong port. `systemctl status bifrost-portal`,
  `journalctl -u bifrost-portal -n 50`, and re-check `PORT` matches `proxy_pass`.
- **App starts then exits** — usually a missing/invalid `/etc/bifrost-portal.env`
  or unwritable `BIFROST_DB_PATH`. Check `journalctl -u bifrost-portal`.
- **`better-sqlite3` errors on boot** — native module built for the wrong arch.
  Re-run `pnpm install --frozen-lockfile` on the server and rebuild.
- **Migrations hit the wrong DB** — confirm the env file is sourced; `migrate.js`
  honors `BIFROST_DB_PATH` (verified) and falls back to the repo-local dev db otherwise.