Darkraise
74aa0cf067
On a network share the database copy and cache pre-warm run for minutes, and the listener only opened after all of it, so anyone who loaded the page during that window got a connection error. Open the port immediately after registering the static frontend, and hold every route that needs the database behind a gate that answers 503 with a machine-readable marker and the current startup phase. The SPA shell needs nothing the gate blocks, so the frontend renders a "server is starting" screen on the login page's own background, polls /health, and reloads once the server reports ready. Move the database copy out of the entrypoint: it does the same freshness check the application does, and doing it before node starts kept the port closed for the longest part of startup. Distinguish a dead server from a slow one by requiring the backend's own marker, since a proxy in front of a stopped backend still answers.
137 lines
5.3 KiB
Bash
137 lines
5.3 KiB
Bash
#!/bin/sh
|
||
set -e
|
||
|
||
# =============================================================================
|
||
# Docker Entrypoint for Flashpoint Web Backend
|
||
# =============================================================================
|
||
# This script handles:
|
||
# 1. Dynamic UID/GID setup (via PUID/PGID environment variables)
|
||
# 2. Ownership of the mounted data and log directories
|
||
# 3. Application startup with dropped privileges
|
||
#
|
||
# The local database copy is deliberately NOT done here: it runs inside the
|
||
# application so the HTTP port is already open while it happens, which on a
|
||
# network share is the difference between a "server is starting" page and a
|
||
# connection error.
|
||
# =============================================================================
|
||
|
||
echo "🚀 Flashpoint Web Backend - Starting..."
|
||
|
||
# =============================================================================
|
||
# Dynamic UID/GID Setup
|
||
# =============================================================================
|
||
# Supports PUID/PGID environment variables to match host user permissions
|
||
# This allows bind-mounted volumes to have correct ownership
|
||
|
||
PUID=${PUID:-1000}
|
||
PGID=${PGID:-1000}
|
||
APP_USER="node"
|
||
APP_GROUP="node"
|
||
|
||
setup_user() {
|
||
local current_uid
|
||
local current_gid
|
||
|
||
current_uid=$(id -u $APP_USER 2>/dev/null || echo "")
|
||
current_gid=$(id -g $APP_USER 2>/dev/null || echo "")
|
||
|
||
# Check if we need to modify UID/GID
|
||
if [ "$current_uid" != "$PUID" ] || [ "$current_gid" != "$PGID" ]; then
|
||
echo "🔧 Adjusting user UID/GID..."
|
||
echo " PUID: $PUID (was: $current_uid)"
|
||
echo " PGID: $PGID (was: $current_gid)"
|
||
|
||
# Modify group GID if different
|
||
if [ "$current_gid" != "$PGID" ]; then
|
||
# Check if target GID is already in use by another group
|
||
existing_group=$(getent group "$PGID" 2>/dev/null | cut -d: -f1 || true)
|
||
if [ -n "$existing_group" ] && [ "$existing_group" != "$APP_GROUP" ]; then
|
||
groupdel "$existing_group" 2>/dev/null || true
|
||
fi
|
||
groupmod -g "$PGID" $APP_GROUP 2>/dev/null || true
|
||
fi
|
||
|
||
# Modify user UID if different
|
||
if [ "$current_uid" != "$PUID" ]; then
|
||
# Check if target UID is already in use by another user
|
||
existing_user=$(getent passwd "$PUID" 2>/dev/null | cut -d: -f1 || true)
|
||
if [ -n "$existing_user" ] && [ "$existing_user" != "$APP_USER" ]; then
|
||
userdel "$existing_user" 2>/dev/null || true
|
||
fi
|
||
usermod -u "$PUID" $APP_USER 2>/dev/null || true
|
||
fi
|
||
|
||
# Image-layer files stay owned by the build UID, which the app user no
|
||
# longer is. Ruffle installs and updates write here, so without this
|
||
# they fail with EACCES on any non-default PUID. Skipped when the UID
|
||
# is unchanged: chown -R copies every file up out of the image layer.
|
||
echo "🔧 Ensuring correct ownership of served frontend files..."
|
||
chown -R $APP_USER:$APP_GROUP /app/frontend/dist 2>/dev/null || true
|
||
else
|
||
echo "✅ User UID/GID OK (UID=$PUID, GID=$PGID)"
|
||
fi
|
||
|
||
# Always fix ownership of mounted directories (volumes override container permissions)
|
||
echo "🔧 Ensuring correct ownership of data directories..."
|
||
chown -R $APP_USER:$APP_GROUP /app/data /app/logs 2>/dev/null || true
|
||
}
|
||
|
||
# Only setup user if running as root
|
||
if [ "$(id -u)" = "0" ]; then
|
||
setup_user
|
||
fi
|
||
|
||
# =============================================================================
|
||
# Database Copy Configuration
|
||
# =============================================================================
|
||
|
||
ENABLE_LOCAL_DB_COPY="${ENABLE_LOCAL_DB_COPY:-false}"
|
||
FLASHPOINT_PATH="/data/flashpoint"
|
||
SOURCE_DB_PATH="${FLASHPOINT_PATH}/Data/flashpoint.sqlite"
|
||
LOCAL_DB_PATH="/app/data/flashpoint.sqlite"
|
||
|
||
# =============================================================================
|
||
# Main Entrypoint Logic
|
||
# =============================================================================
|
||
|
||
# Check if source database exists
|
||
if [ ! -f "$SOURCE_DB_PATH" ]; then
|
||
echo "❌ ERROR: Source database not found at: $SOURCE_DB_PATH"
|
||
echo " Make sure the Flashpoint data volume is mounted correctly"
|
||
exit 1
|
||
fi
|
||
|
||
echo "✅ Source database found: $SOURCE_DB_PATH"
|
||
|
||
# The copy itself is left to the application: it performs the same freshness
|
||
# check, and doing it here keeps the port closed for the whole copy, which is
|
||
# minutes on a network share — exactly when a user is most likely to open the
|
||
# page and get a connection error instead of the "server is starting" screen.
|
||
if [ "$ENABLE_LOCAL_DB_COPY" = "true" ]; then
|
||
echo "🔄 Local database copy is ENABLED (performed by the application at startup)"
|
||
else
|
||
echo "ℹ️ Local database copy is DISABLED"
|
||
echo " Database will be accessed directly from: $SOURCE_DB_PATH"
|
||
fi
|
||
|
||
# Display configuration summary
|
||
echo ""
|
||
echo "📋 Configuration Summary:"
|
||
echo " Flashpoint data: ${FLASHPOINT_PATH}"
|
||
echo " Local DB copy: ${ENABLE_LOCAL_DB_COPY}"
|
||
echo " Running as UID: ${PUID}, GID: ${PGID}"
|
||
if [ "$ENABLE_LOCAL_DB_COPY" = "true" ]; then
|
||
echo " Local DB path: ${LOCAL_DB_PATH}"
|
||
fi
|
||
echo ""
|
||
|
||
# Start the application
|
||
echo "🎮 Starting Flashpoint Web Backend..."
|
||
|
||
# If running as root, drop privileges using su-exec
|
||
if [ "$(id -u)" = "0" ]; then
|
||
exec su-exec $APP_USER node dist/server.js
|
||
else
|
||
exec node dist/server.js
|
||
fi
|