Fix modular mode deployment issues (#1045)

- Fix Postgres version check failing with restricted DB users (use $POSTGRES_DB instead of hardcoded 'postgres')
- Fix DVR recording broken in modular mode (respect DISPATCHARR_PORT instead of hardcoding 9191)
- Remove flushdb() from wait_for_redis.py to prevent Redis data loss on container restart
- Add DISPATCHARR_PORT to celery environment in docker-compose.yml
- Add depends_on health conditions for proper service startup ordering
- Add extra_hosts for host.docker.internal resolution on Linux
- Harden celery entrypoint with timeouts for JWT wait (120s) and migration wait (300s)
- Replace fragile showmigrations grep with migrate --check
- Add unit tests for DVR port resolution and flushdb removal regression

docker/docker-compose.yml

@@ -15,8 +15,12 @@ services:
     volumes:
       - ./data:/data
     depends_on:
-      - db
-      - redis
+      db:
+        condition: service_healthy
+      redis:
+        condition: service_healthy
+    extra_hosts:
+      - "host.docker.internal:host-gateway"
 
     # --- Environment Configuration ---
     environment:
@@ -83,9 +87,12 @@ services:
     container_name: dispatcharr_celery
     restart: unless-stopped
     depends_on:
-      - db
-      - redis
-      - web
+      db:
+        condition: service_healthy
+      redis:
+        condition: service_healthy
+      web:
+        condition: service_started
     volumes:
       - ./data:/data
     extra_hosts:
@@ -97,6 +104,10 @@ services:
       # Deployment Mode
       - DISPATCHARR_ENV=modular
 
+      # Internal Service Communication
+      # Must match the web service port for DVR recording and internal API calls
+      - DISPATCHARR_PORT=9191
+
       # PostgreSQL Connection
       - POSTGRES_HOST=db
       - POSTGRES_PORT=5432

docker/entrypoint.celery.sh

@@ -9,9 +9,19 @@ echo_with_timestamp() {
   echo "$(date '+%Y-%m-%d %H:%M:%S') - $1"
 }
 
-# Wait for Django secret key
+# Wait for Django secret key (generated by the web container on startup)
+JWT_TIMEOUT=120
+JWT_WAITED=0
 echo 'Waiting for Django secret key...'
-while [ ! -f /data/jwt ]; do sleep 1; done
+while [ ! -f /data/jwt ]; do
+    if [ $JWT_WAITED -ge $JWT_TIMEOUT ]; then
+        echo "❌ ERROR: Timed out waiting for /data/jwt after ${JWT_TIMEOUT}s."
+        echo "   Is the web container running? Does it have the /data volume mounted?"
+        exit 1
+    fi
+    sleep 1
+    JWT_WAITED=$((JWT_WAITED + 1))
+done
 export DJANGO_SECRET_KEY="$(tr -d '\r\n' < /data/jwt)"
 
 # --- NumPy version switching for legacy hardware ---
@@ -26,11 +36,21 @@ if [ "$USE_LEGACY_NUMPY" = "true" ]; then
     fi
 fi
 
-# Wait for migrations to complete (check that NO unapplied migrations remain)
+# Wait for migrations to complete
+# Uses 'migrate --check' which exits 0 only when all migrations are applied,
+# and exits 1 on unapplied migrations OR connection errors (safe either way)
+MIG_TIMEOUT=300
+MIG_WAITED=0
 echo 'Waiting for migrations to complete...'
-until ! python manage.py showmigrations 2>&1 | grep -q '\[ \]'; do
+until python manage.py migrate --check >/dev/null 2>&1; do
+    if [ $MIG_WAITED -ge $MIG_TIMEOUT ]; then
+        echo "❌ ERROR: Timed out waiting for migrations after ${MIG_TIMEOUT}s."
+        echo "   Check web container logs for migration errors."
+        exit 1
+    fi
     echo_with_timestamp 'Migrations not ready yet, waiting...'
     sleep 2
+    MIG_WAITED=$((MIG_WAITED + 2))
 done
 
 # Start Celery

docker/init/02-postgres.sh

@@ -204,14 +204,19 @@ check_external_postgres_version() {
     MIN_REQUIRED_VERSION=$PG_VERSION
 
     # Query external PostgreSQL version
-    EXTERNAL_VERSION=$(PGPASSWORD="$POSTGRES_PASSWORD" psql -w -h "$POSTGRES_HOST" -p "$POSTGRES_PORT" -U "$POSTGRES_USER" -d "postgres" -tAc "SHOW server_version;" 2>/dev/null | grep -oE '^[0-9]+')
+    # Use $POSTGRES_DB — restricted users may not have access to the default 'postgres' database
+    PG_VERSION_ERR=$(mktemp)
+    EXTERNAL_VERSION=$(PGPASSWORD="$POSTGRES_PASSWORD" psql -w -h "$POSTGRES_HOST" -p "$POSTGRES_PORT" -U "$POSTGRES_USER" -d "$POSTGRES_DB" -tAc "SHOW server_version;" 2>"$PG_VERSION_ERR" | grep -oE '^[0-9]+')
 
     if [ -z "$EXTERNAL_VERSION" ]; then
         echo "❌ ERROR: Unable to determine external PostgreSQL version"
-        echo "   Could not connect to database at ${POSTGRES_HOST}:${POSTGRES_PORT}"
+        echo "   Could not connect to database '$POSTGRES_DB' at ${POSTGRES_HOST}:${POSTGRES_PORT} as user '$POSTGRES_USER'"
+        echo "   Error: $(cat "$PG_VERSION_ERR")"
         echo "   Please verify your database connection settings."
+        rm -f "$PG_VERSION_ERR"
         return 1
     fi
+    rm -f "$PG_VERSION_ERR"
 
     # Compare versions
     if [[ "$EXTERNAL_VERSION" -lt "$MIN_REQUIRED_VERSION" ]]; then