fix(deploy): clear stale celery pidfile + configurable listing goto timeout

entrypoint.sh

@@ -78,4 +78,8 @@ start_proxy_bridge_if_needed() {
 
 start_proxy_bridge_if_needed "$@"
 
+# Clean stale celery pidfiles so `docker compose restart worker/beat` works.
+# `restart` keeps container fs, so a previous PID file blocks new celery start.
+rm -f /tmp/celery-*.pid 2>/dev/null || true
+
 exec "$@"

iaai_scraper/browser/listing.py

@@ -135,11 +135,12 @@ class ListingCollector:
 
 	def open_cars_listing(self, page: Page, *, url_override: str | None = None) -> None:
 		url = url_override or self.settings.listing.cars_url
-		logger.info("Opening cars listing page: %s", url)
+		goto_timeout = max(60_000, int(getattr(self.settings, "default_timeout_ms", 60_000)))
+		logger.info("Opening cars listing page: %s (timeout=%dms)", url, goto_timeout)
 		last_err = None
 		for attempt in range(3):
 			try:
-				page.goto(url, wait_until="commit", timeout=60_000)
+				page.goto(url, wait_until="commit", timeout=goto_timeout)
 				last_err = None
 				break
 			except Exception as e: