Spaces:
Sleeping
Sleeping
Download tests/load_test_throttle.py from Qurrah/replyflow-rag: direct link, hf CLI and curl.
- Browser
- Download file 4.44 kB
-
https://huggingface.co/spaces/Qurrah/replyflow-rag/resolve/main/tests/load_test_throttle.py
- Command line
-
hf download hf://spaces/Qurrah/replyflow-rag/tests/load_test_throttle.py
-
curl -L -o load_test_throttle.py https://huggingface.co/spaces/Qurrah/replyflow-rag/resolve/main/tests/load_test_throttle.py
4.44 kB
| """ | |
| Real load test against your actual Shopify store β fires enough concurrent | |
| GraphQL requests to trigger Shopify's real cost-based throttling, so you can | |
| watch your retry logic kick in live in the logs. | |
| β οΈ WARNING before running this: | |
| - This burns real API rate-limit budget on your store. Only run against | |
| a DEV/TEST store, never production, unless you know what you're doing. | |
| - Shopify's throttle bucket is shared across your whole app for that | |
| store β this WILL temporarily slow down/throttle any other requests | |
| hitting the same store while the test runs. | |
| - Start with CONCURRENT_REQUESTS = 5 and increase gradually. Going | |
| straight to 50 is how you find out the hard way what your bucket size is. | |
| Usage: | |
| export SHOP_DOMAIN="your-dev-store.myshopify.com" | |
| export SHOPIFY_ACCESS_TOKEN="shpat_xxxxx" | |
| python load_test_throttle.py | |
| """ | |
| import os | |
| import time | |
| import logging | |
| from concurrent.futures import ThreadPoolExecutor, as_completed | |
| from dotenv import load_dotenv | |
| load_dotenv() # reads .env from the current directory into os.environ | |
| logging.basicConfig(level=logging.INFO, format="%(asctime)s | %(levelname)s | %(message)s") | |
| logger = logging.getLogger(__name__) | |
| # --- ADJUST THIS IMPORT to match your project structure --- | |
| from clients.shopify.product_search_api import search_products | |
| SHOP_DOMAIN = os.environ.get("SHOPIFY_SHOP_DOMAIN") | |
| ACCESS_TOKEN = os.environ.get("SHOPIFY_ACCESS_TOKEN") | |
| CONCURRENT_REQUESTS = 15 # start low, raise gradually | |
| QUERY = "" # empty = match everything, cheapest possible query | |
| LIMIT_PER_REQUEST = 50 # higher limit = more query "cost" per call = throttles faster | |
| def fire_one(request_num: int) -> tuple[int, bool, str]: | |
| """Returns (request_num, succeeded, message).""" | |
| start = time.time() | |
| try: | |
| products = search_products(SHOP_DOMAIN, ACCESS_TOKEN, QUERY, limit=LIMIT_PER_REQUEST) | |
| elapsed = time.time() - start | |
| return request_num, True, f"OK β {len(products)} products in {elapsed:.2f}s" | |
| except Exception as e: | |
| elapsed = time.time() - start | |
| return request_num, False, f"FAILED after {elapsed:.2f}s β {type(e).__name__}: {e}" | |
| def main(): | |
| if not SHOP_DOMAIN or not ACCESS_TOKEN: | |
| raise SystemExit( | |
| "Set SHOP_DOMAIN and SHOPIFY_ACCESS_TOKEN environment variables first.\n" | |
| " export SHOP_DOMAIN='your-dev-store.myshopify.com'\n" | |
| " export SHOPIFY_ACCESS_TOKEN='shpat_xxxxx'" | |
| ) | |
| logger.info( | |
| "Firing %d concurrent requests at %s (limit=%d per request)...", | |
| CONCURRENT_REQUESTS, SHOP_DOMAIN, LIMIT_PER_REQUEST, | |
| ) | |
| logger.info("Watch for 'THROTTLED' or '429' log lines from product_search_api itself.") | |
| start = time.time() | |
| results = [] | |
| with ThreadPoolExecutor(max_workers=CONCURRENT_REQUESTS) as executor: | |
| futures = [executor.submit(fire_one, i) for i in range(CONCURRENT_REQUESTS)] | |
| for future in as_completed(futures): | |
| request_num, ok, message = future.result() | |
| status = "β " if ok else "β" | |
| logger.info("%s Request #%02d: %s", status, request_num, message) | |
| results.append(ok) | |
| total_time = time.time() - start | |
| succeeded = sum(results) | |
| failed = len(results) - succeeded | |
| print("\n" + "=" * 60) | |
| print(f"Total time: {total_time:.2f}s") | |
| print(f"Succeeded: {succeeded}/{len(results)}") | |
| print(f"Failed: {failed}/{len(results)}") | |
| print("=" * 60) | |
| if failed == 0 and total_time > 3: | |
| print( | |
| "No hard failures, but if total time is noticeably higher than a\n" | |
| "single request would take, some of these likely got throttled and\n" | |
| "silently succeeded after your retry logic kicked in β check the\n" | |
| "logs above for 'THROTTLED' or '429' lines to confirm." | |
| ) | |
| elif failed == 0: | |
| print( | |
| "Everything succeeded fast β you didn't hit the throttle threshold.\n" | |
| "Try raising CONCURRENT_REQUESTS or LIMIT_PER_REQUEST and re-run." | |
| ) | |
| else: | |
| print( | |
| "Some requests failed even after retries β you exceeded your\n" | |
| "MAX_RATE_LIMIT_RETRIES budget under sustained load. This is\n" | |
| "expected once you push hard enough; it tells you your retry\n" | |
| "ceiling under this load level." | |
| ) | |
| if __name__ == "__main__": | |
| main() |