Make the reverse proxy a required, pluggable part of every deployment

gunicorn previously only ever bound to a Unix socket, which assumes the
reverse proxy is nginx running on this same host — it can't be reached by
a proxy on a different machine at all (e.g. Nginx Proxy Manager on its own
box), which is this project's actual deployment target. Switch gunicorn to
listening on a plain TCP address instead, controlled by a new
EVECTION_GUNICORN_BIND setting in .env.

install.sh now asks whether the proxy is local or remote and picks a
sensible bind default either way (127.0.0.1:8000 vs 0.0.0.0:8000, with a
reminder to firewall the port in the remote case), and no longer tries to
automate nginx/certbot itself — docs/install.md's new "Reverse proxy"
section covers both shapes instead, since which proxy and where it runs is
the operator's choice, not something this script should assume or set up.
upgrade.sh's health check now curls the TCP bind instead of the old socket
path.

Also add WhiteNoise so gunicorn serves /static/ itself: the old approach
(nginx aliasing the staticfiles directory on disk) silently breaks the
moment the proxy isn't on this host and can't see this filesystem.
nginx.conf.example's own /static/ alias becomes an optional same-host perf
optimization rather than a requirement.

Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
This commit is contained in:
ksmithandClaude Sonnet 5 committed 2026-10-31 09:01:46 -06:00
1 parent 62702e13e3
commit 21c33e8188
7 files changed
+157 -91

No files matched your search

+21 -3
View File
@@ -3,12 +3,30 @@ from .base import env
DEBUG = False
# nginx terminates TLS and proxies to gunicorn over a Unix socket (see
# docs/install.md) — this tells Django's request.is_secure() to trust the
# header nginx sets, and the redirect/cookie/HSTS settings below assume it.
# A reverse proxy (nginx, Caddy, Traefik, Nginx Proxy Manager on another
# host, ...) is required in front of gunicorn in every deployment — see
# docs/install.md — and it's always the one terminating TLS. This tells
# Django's request.is_secure() to trust the X-Forwarded-Proto header it
# sets, and the redirect/cookie/HSTS settings below assume it's set.
SECURE_SSL_REDIRECT = env.bool("SECURE_SSL_REDIRECT", default=True)
SECURE_PROXY_SSL_HEADER = ("HTTP_X_FORWARDED_PROTO", "https")
# WhiteNoise serves /static/ directly from gunicorn instead of relying on
# the reverse proxy to alias the staticfiles directory on disk — the proxy
# may not even be on this host (see docs/install.md), so it can't be
# trusted to have filesystem access to $APP_DIR/staticfiles.
MIDDLEWARE = [
"django.middleware.security.SecurityMiddleware",
"whitenoise.middleware.WhiteNoiseMiddleware",
*MIDDLEWARE[1:], # noqa: F405
]
STORAGES = {
"default": {"BACKEND": "django.core.files.storage.FileSystemStorage"},
"staticfiles": {
"BACKEND": "whitenoise.storage.CompressedManifestStaticFilesStorage",
},
}
SESSION_COOKIE_SECURE = True
CSRF_COOKIE_SECURE = True
SESSION_COOKIE_SAMESITE = "Lax"
+6 -5
View File
@@ -9,11 +9,12 @@ Group=evection
WorkingDirectory=/opt/evection/app
EnvironmentFile=/opt/evection/app/.env
Environment=DJANGO_SETTINGS_MODULE=config.settings.prod
RuntimeDirectory=evection
ExecStart=/opt/evection/venv/bin/gunicorn \
--workers 3 \
--bind unix:/run/evection/gunicorn.sock \
config.wsgi:application
# Evection is always meant to sit behind a reverse proxy (see
# docs/install.md) — it listens on a plain TCP address, never a Unix
# socket, so that proxy can be on this host or a different one.
# EVECTION_GUNICORN_BIND (set in .env by scripts/install.sh) controls
# what address:port that is.
ExecStart=/bin/sh -c 'exec /opt/evection/venv/bin/gunicorn --workers 3 --bind "$EVECTION_GUNICORN_BIND" config.wsgi:application'
ExecReload=/bin/kill -s HUP $MAINPID
Restart=on-failure
+15 -4
View File
@@ -1,6 +1,12 @@
# Reverse proxy for Evection. Copy to /etc/nginx/sites-available/evection,
# adjust server_name and the certificate paths (certbot will fill those in
# for you if you use `certbot --nginx`), then symlink into sites-enabled.
# Reverse proxy for Evection, for the case where nginx runs on the SAME host
# as Evection (see docs/install.md's "Reverse proxy" section — if your proxy
# is on a different host/VM instead, e.g. Nginx Proxy Manager, this file
# doesn't apply to you directly, but the proxy_set_header lines below still
# show what any proxy needs to forward). Copy to
# /etc/nginx/sites-available/evection, adjust server_name, the certificate
# paths (certbot will fill those in for you if you use `certbot --nginx`),
# and the port in proxy_pass if you didn't use EVECTION_GUNICORN_BIND's
# default of 127.0.0.1:8000, then symlink into sites-enabled.
server {
listen 80;
@@ -24,6 +30,11 @@ server {
client_max_body_size 10M; # OPML uploads are small, but leave headroom
# Optional: gunicorn already serves /static/ itself via WhiteNoise (with
# its own far-future cache headers, since filenames are content-hashed),
# so this block is purely a same-host perf optimization that skips
# gunicorn for static assets — not required, and not an option at all
# once the proxy is on a different host than $APP_DIR/staticfiles.
location /static/ {
alias /opt/evection/app/staticfiles/;
expires 30d;
@@ -36,7 +47,7 @@ server {
# else below, so no special-casing is needed here.
location / {
proxy_pass http://unix:/run/evection/gunicorn.sock;
proxy_pass http://127.0.0.1:8000;
proxy_set_header Host $host;
proxy_set_header X-Real-IP $remote_addr;
proxy_set_header X-Forwarded-For $proxy_add_x_forwarded_for;
+61 -26
View File
@@ -1,9 +1,16 @@
# Installing Evection
Evection is a traditional Python/Django application: `pip install`, a WSGI server
(gunicorn), a Celery worker + beat scheduler for background feed polling, and an
nginx reverse proxy in front of it. There's no Docker image for v1 — everything
below runs directly on the host.
(gunicorn), and a Celery worker + beat scheduler for background feed polling.
There's no Docker image for v1 — everything below runs directly on the host.
**A reverse proxy in front of gunicorn is required, always** — Evection has no
TLS support of its own and assumes one is terminating HTTPS for it (see "Reverse
proxy" below). It can be nginx/Caddy/Traefik running on the same host, or
something like Nginx Proxy Manager running on an entirely different host/VM —
`scripts/install.sh` doesn't set one up for you, since which one and where it
runs is up to you; it just leaves gunicorn listening on a TCP address for that
proxy to reach.
Both SQLite and PostgreSQL are fully supported. Steps that differ are marked
**SQLite** / **Postgres**; everything else applies to both.
@@ -32,9 +39,11 @@ run at all until a tag exists.
## Quick install
Once the prerequisites below are in place, `scripts/install.sh` automates
everything from "create a system user" through "systemd units are running" —
and optionally nginx + a Let's Encrypt certificate too. It's interactive
(asks for your domain and database choice) and safe to re-run.
everything from "create a system user" through "systemd units are running",
including asking where your reverse proxy will run so it can set up gunicorn's
listen address accordingly. It's interactive (asks for your domain, database
choice, and proxy location) and safe to re-run. It does **not** set up the
proxy itself — see "Reverse proxy" below for that, once this finishes.
```bash
sudo mkdir -p /opt/evection
@@ -61,14 +70,16 @@ and 4 below first — prerequisites and what `.env` controls.
- **Postgres only**: a PostgreSQL server (12+) and its client headers if you're not
using the bundled `psycopg[binary]` wheel (most platforms are covered by the wheel,
so this is rarely needed)
- nginx (or another reverse proxy)
- A domain name pointed at the host, if you want this reachable outside your LAN
- A reverse proxy, already running or that you're about to set up — on this
host or a different one (see "Reverse proxy" below)
- A domain name pointed at wherever that proxy terminates TLS, if you want
this reachable outside your LAN
Debian/Ubuntu example for the non-Python pieces (swap `python3.13-venv` for
whatever `python3 --version` reports on your system):
```bash
sudo apt install redis-server nginx python3.13-venv
sudo apt install redis-server python3.13-venv
# Postgres only:
sudo apt install postgresql
```
@@ -129,11 +140,15 @@ At minimum, set the following in `.env`:
sudo -u postgres createdb evection -O evection
```
- `REDIS_URL` — `redis://localhost:6379/0` is fine for a single-host install
- `EVECTION_GUNICORN_BIND` — the `address:port` gunicorn listens on; see
"Reverse proxy" below for how to pick this
Evection reads `config.settings.prod` when run via the commands below (see each
systemd unit's `DJANGO_SETTINGS_MODULE`). That module turns on HTTPS-only cookies,
HSTS, and a Content-Security-Policy header on top of the base settings — see
`config/settings/prod.py` if you need to adjust it for your proxy setup.
HSTS, and a Content-Security-Policy header on top of the base settings, and
serves static files itself via WhiteNoise rather than relying on the proxy to
have filesystem access to this host — see `config/settings/prod.py` if you need
to adjust any of that.
## 5. Initialize the database and static files
@@ -192,28 +207,47 @@ sudo systemctl status evection-gunicorn evection-celery-worker evection-celery-b
sudo journalctl -u evection-gunicorn -f
```
## 7. nginx
## 7. Reverse proxy
Copy `docs/deploy/nginx.conf.example` to `/etc/nginx/sites-available/evection`,
adjust `server_name` and the certificate paths, then enable it:
gunicorn only ever listens on plain HTTP, on the `EVECTION_GUNICORN_BIND`
address you set in step 4 — it never terminates TLS itself. A reverse proxy
in front of it is not optional: it's what the setup wizard, login, and every
other page actually get reached through. Two shapes this takes:
**Proxy on the same host as Evection.** Set `EVECTION_GUNICORN_BIND=127.0.0.1:8000`
(install.sh's default when you say the proxy is local) so gunicorn isn't
reachable from outside this machine at all — only the proxy, listening on
80/443, is. `docs/deploy/nginx.conf.example` is a working nginx config for
this case; copy it into place and get a certificate the usual way:
```bash
sudo cp docs/deploy/nginx.conf.example /etc/nginx/sites-available/evection
# edit it: server_name, and the proxy_pass port if you didn't use 8000
sudo ln -s /etc/nginx/sites-available/evection /etc/nginx/sites-enabled/
sudo nginx -t && sudo systemctl reload nginx
```
Get a TLS certificate (Let's Encrypt via certbot is the easiest path):
```bash
sudo apt install certbot python3-certbot-nginx
sudo certbot --nginx -d your-domain.example
```
Evection's `prod` settings enable `SECURE_SSL_REDIRECT` and HSTS by default, so
plan to have TLS working before your first real visit — see
`config/settings/prod.py` to disable the redirect temporarily if you need to debug
over plain HTTP first.
**Proxy on a different host or VM** (Nginx Proxy Manager, a Caddy box, a
container host, etc.). Set `EVECTION_GUNICORN_BIND=0.0.0.0:8000` (or to this
host's specific IP) so the proxy can actually reach gunicorn over the
network, then **firewall that port to only accept connections from the
proxy's host** — gunicorn itself has no auth in front of it, so it must
never be reachable from the open internet directly. Point the proxy's
upstream/forward target at `this-host:8000`, and make sure it forwards
`X-Forwarded-Proto` (whatever sets up TLS passthrough/termination normally
handles this, but double-check — see the next paragraph for why it matters).
Whichever proxy you use, `docs/deploy/nginx.conf.example`'s `location /`
block shows the three headers that need forwarding either way.
Either way, the proxy must set `X-Forwarded-Proto: https` on requests it
forwards once it's terminating TLS — Evection's `prod` settings
(`SECURE_PROXY_SSL_HEADER` in `config/settings/prod.py`) trust that header to
know the original request was HTTPS, which is what keeps `SECURE_SSL_REDIRECT`
from causing a redirect loop. Plan to have TLS working end-to-end before your
first real visit; see `config/settings/prod.py` to disable the redirect
temporarily if you need to debug over plain HTTP first.
## 8. Visit the site
@@ -257,9 +291,10 @@ Equivalent by hand:
```bash
cd /opt/evection/app
sudo -u evection -H git fetch --tags --force
LATEST_TAG="$(sudo -u evection -H git tag --sort=-creatordate | head -n1)"
sudo -u evection -H git checkout --detach "$LATEST_TAG"
git fetch --tags --force
LATEST_TAG="$(git tag --sort=-creatordate | head -n1)"
git checkout --detach "$LATEST_TAG"
chown -R evection:evection /opt/evection/app
sudo -u evection -H /opt/evection/venv/bin/pip install .
sudo -u evection -H env DJANGO_SETTINGS_MODULE=config.settings.prod \
/opt/evection/venv/bin/python manage.py migrate
+1
View File
@@ -22,6 +22,7 @@ dependencies = [
"psycopg[binary]>=3.1,<4",
"gunicorn>=21,<24",
"django-csp>=4.0,<5",
"whitenoise>=6.6,<7",
]
[project.optional-dependencies]
+38 -51
View File
@@ -1,6 +1,11 @@
#!/usr/bin/env bash
# Automates docs/install.md's manual steps: system user, virtualenv, .env,
# migrate/collectstatic, systemd units, and (optionally) nginx + TLS.
# migrate/collectstatic, systemd units. Does NOT set up a reverse proxy —
# Evection always requires one in front of gunicorn (see docs/install.md),
# but which one (local nginx/Caddy/Traefik, or something like Nginx Proxy
# Manager on a different host entirely) is up to you, so this script just
# leaves gunicorn listening on EVECTION_GUNICORN_BIND for you to point it
# at.
#
# Run as root, from inside the checkout you want to serve:
# sudo ./scripts/install.sh
@@ -11,7 +16,7 @@
# Non-interactive use (e.g. scripted provisioning) — set these before
# running, and the script won't prompt:
# EVECTION_ALLOWED_HOSTS, EVECTION_DATABASE_URL, EVECTION_REDIS_URL,
# EVECTION_USER (defaults to "evection")
# EVECTION_GUNICORN_BIND, EVECTION_USER (defaults to "evection")
set -euo pipefail
@@ -58,12 +63,10 @@ if ! "$PYTHON_BIN" -c 'import ensurepip' >/dev/null 2>&1; then
exit 1
fi
for bin in redis-cli nginx; do
if ! command -v "$bin" >/dev/null 2>&1; then
echo "Warning: '$bin' not found on PATH. Install it before going live" \
"(see docs/install.md's prerequisites section: apt install redis-server nginx)." >&2
fi
done
if ! command -v redis-cli >/dev/null 2>&1; then
echo "Warning: 'redis-cli' not found on PATH. Install it before going live" \
"(see docs/install.md's prerequisites section: apt install redis-server)." >&2
fi
# --- 2. System user -----------------------------------------------------
@@ -138,12 +141,31 @@ if [ ! -f "$ENV_FILE" ]; then
fi
DATABASE_URL="${DATABASE_URL:-sqlite:///$APP_DIR/db.sqlite3}"
GUNICORN_BIND="${EVECTION_GUNICORN_BIND:-}"
if [ -z "$GUNICORN_BIND" ] && [ -t 0 ]; then
echo
echo "Evection must run behind a reverse proxy — gunicorn needs an" \
"address:port that proxy can reach."
read -rp "Is the reverse proxy on this same machine, or a different one? [same/different] [same]: " PROXY_LOCATION
if [[ "${PROXY_LOCATION:-}" =~ ^[Dd] ]]; then
read -rp "Address:port for gunicorn to listen on (e.g. 0.0.0.0:8000, or this host's own IP:8000) [0.0.0.0:8000]: " GUNICORN_BIND
GUNICORN_BIND="${GUNICORN_BIND:-0.0.0.0:8000}"
echo " Remember to firewall this port to only accept connections" \
"from the reverse proxy's host."
else
read -rp "Address:port for gunicorn to listen on [127.0.0.1:8000]: " GUNICORN_BIND
GUNICORN_BIND="${GUNICORN_BIND:-127.0.0.1:8000}"
fi
fi
GUNICORN_BIND="${GUNICORN_BIND:-127.0.0.1:8000}"
cat > "$ENV_FILE" <<EOF
SECRET_KEY=$SECRET_KEY
DEBUG=False
ALLOWED_HOSTS=$ALLOWED_HOSTS
DATABASE_URL=$DATABASE_URL
REDIS_URL=${EVECTION_REDIS_URL:-redis://localhost:6379/0}
EVECTION_GUNICORN_BIND=$GUNICORN_BIND
EVECTION_SIGNUP_ENABLED=False
EVECTION_MIN_FETCH_INTERVAL_MINUTES=15
EVECTION_PWA_CACHE_VERSION=v1
@@ -154,6 +176,11 @@ else
echo "==> $ENV_FILE already exists — leaving it alone"
fi
# Needed below for the final "gunicorn is listening on..." message even
# when $ENV_FILE already existed and the block above was skipped.
GUNICORN_BIND="$(grep -m1 '^EVECTION_GUNICORN_BIND=' "$ENV_FILE" | cut -d= -f2-)"
GUNICORN_BIND="${GUNICORN_BIND:-127.0.0.1:8000}"
# --- 5. Migrate + static files --------------------------------------------
echo "==> Running migrations"
@@ -181,49 +208,9 @@ echo "==> Services installed. Checking status..."
systemctl --no-pager status evection-gunicorn evection-celery-worker evection-celery-beat \
|| echo "One or more services didn't start cleanly — see 'journalctl -u evection-gunicorn' etc." >&2
# --- 7. nginx (optional) ---------------------------------------------------
SET_UP_NGINX="${EVECTION_SETUP_NGINX:-}"
if [ -z "$SET_UP_NGINX" ] && [ -t 0 ]; then
echo
read -rp "Set up nginx and an HTTPS certificate now? [y/N]: " SET_UP_NGINX
fi
if [[ "${SET_UP_NGINX:-}" =~ ^[Yy]$ ]]; then
if ! command -v nginx >/dev/null 2>&1; then
echo "nginx isn't installed — install it first (apt install nginx), then" \
"run docs/install.md's nginx section manually." >&2
else
DOMAIN="${EVECTION_DOMAIN:-}"
if [ -z "$DOMAIN" ] && [ -t 0 ]; then
read -rp "Domain name for nginx/TLS (must match a hostname in ALLOWED_HOSTS): " DOMAIN
fi
if [ -z "$DOMAIN" ]; then
echo "No domain given — skipping nginx setup. Run docs/install.md's nginx section manually." >&2
else
sed -e "s/your-domain.example/$DOMAIN/g" \
-e "s#/opt/evection#$EVECTION_HOME#g" \
"$APP_DIR/docs/deploy/nginx.conf.example" > "/etc/nginx/sites-available/evection"
ln -sf /etc/nginx/sites-available/evection /etc/nginx/sites-enabled/evection
if nginx -t; then
systemctl reload nginx
echo "==> nginx configured for $DOMAIN (HTTP only so far)"
if command -v certbot >/dev/null 2>&1; then
certbot --nginx -d "$DOMAIN" || echo "certbot failed — run it manually: certbot --nginx -d $DOMAIN" >&2
else
echo "certbot isn't installed. Install it and run:" \
"apt install certbot python3-certbot-nginx && certbot --nginx -d $DOMAIN" >&2
fi
else
echo "nginx config test failed — not enabling it. Check /etc/nginx/sites-available/evection." >&2
fi
fi
fi
else
echo "==> Skipping nginx — see docs/install.md step 7 to set it up yourself."
fi
echo
echo "==> Done."
echo " Visit the site (once nginx/TLS is in place) — the setup wizard creates your admin account."
echo " gunicorn is listening on $GUNICORN_BIND — point your reverse proxy at it"
echo " (see docs/install.md's \"Reverse proxy\" section) before visiting the site."
echo " Once that's in place, the setup wizard creates your admin account on first visit."
echo " To upgrade later: sudo ./scripts/upgrade.sh"
+15 -2
View File
@@ -26,7 +26,6 @@ APP_DIR="$(cd "$(dirname "${BASH_SOURCE[0]}")/.." && pwd)"
EVECTION_HOME="$(dirname "$APP_DIR")"
VENV_DIR="$EVECTION_HOME/venv"
EVECTION_USER="${EVECTION_USER:-evection}"
SOCKET="/run/evection/gunicorn.sock"
# $APP_DIR is owned by $EVECTION_USER, but the git steps below run as
# root — without this, git refuses to touch a repo it doesn't own
@@ -99,7 +98,21 @@ fi
echo "==> Verifying"
sleep 2
if curl -sf --unix-socket "$SOCKET" http://localhost/health/ >/dev/null; then
# gunicorn always listens on a plain TCP address (EVECTION_GUNICORN_BIND in
# .env, set by scripts/install.sh) rather than a Unix socket, since the
# reverse proxy in front of it may be on a different host entirely — see
# docs/install.md. 0.0.0.0 isn't a valid address to connect *to*, so check
# against localhost in that case; this health check only needs to prove
# gunicorn itself is answering on this host, not reach it the way the
# proxy does.
HEALTH_BIND="$(grep -m1 '^EVECTION_GUNICORN_BIND=' "$APP_DIR/.env" | cut -d= -f2-)"
HEALTH_BIND="${HEALTH_BIND:-127.0.0.1:8000}"
HEALTH_HOST="${HEALTH_BIND%:*}"
HEALTH_PORT="${HEALTH_BIND##*:}"
if [ "$HEALTH_HOST" = "0.0.0.0" ]; then
HEALTH_HOST="127.0.0.1"
fi
if curl -sf "http://$HEALTH_HOST:$HEALTH_PORT/health/" >/dev/null; then
echo "Health check OK"
else
echo "Health check FAILED — check: journalctl -u evection-gunicorn -n 50" >&2