feat: support max_retries in jina requests (#6585)

This commit is contained in:
xielong
2024-07-25 13:10:39 +08:00
committed by GitHub
parent 55c2b61921
commit ebcc07e3e9
8 changed files with 129 additions and 20 deletions

View File

@@ -1,13 +1,16 @@
"""
Proxy requests to avoid SSRF
"""
import logging
import os
import time
import httpx
SSRF_PROXY_ALL_URL = os.getenv('SSRF_PROXY_ALL_URL', '')
SSRF_PROXY_HTTP_URL = os.getenv('SSRF_PROXY_HTTP_URL', '')
SSRF_PROXY_HTTPS_URL = os.getenv('SSRF_PROXY_HTTPS_URL', '')
SSRF_DEFAULT_MAX_RETRIES = int(os.getenv('SSRF_DEFAULT_MAX_RETRIES', '3'))
proxies = {
'http://': SSRF_PROXY_HTTP_URL,
@@ -15,34 +18,55 @@ proxies = {
} if SSRF_PROXY_HTTP_URL and SSRF_PROXY_HTTPS_URL else None
def make_request(method, url, **kwargs):
if SSRF_PROXY_ALL_URL:
return httpx.request(method=method, url=url, proxy=SSRF_PROXY_ALL_URL, **kwargs)
elif proxies:
return httpx.request(method=method, url=url, proxies=proxies, **kwargs)
else:
return httpx.request(method=method, url=url, **kwargs)
BACKOFF_FACTOR = 0.5
STATUS_FORCELIST = [429, 500, 502, 503, 504]
def get(url, **kwargs):
return make_request('GET', url, **kwargs)
def make_request(method, url, max_retries=SSRF_DEFAULT_MAX_RETRIES, **kwargs):
retries = 0
while retries <= max_retries:
try:
if SSRF_PROXY_ALL_URL:
response = httpx.request(method=method, url=url, proxy=SSRF_PROXY_ALL_URL, **kwargs)
elif proxies:
response = httpx.request(method=method, url=url, proxies=proxies, **kwargs)
else:
response = httpx.request(method=method, url=url, **kwargs)
if response.status_code not in STATUS_FORCELIST:
return response
else:
logging.warning(f"Received status code {response.status_code} for URL {url} which is in the force list")
except httpx.RequestError as e:
logging.warning(f"Request to URL {url} failed on attempt {retries + 1}: {e}")
retries += 1
if retries <= max_retries:
time.sleep(BACKOFF_FACTOR * (2 ** (retries - 1)))
raise Exception(f"Reached maximum retries ({max_retries}) for URL {url}")
def post(url, **kwargs):
return make_request('POST', url, **kwargs)
def get(url, max_retries=SSRF_DEFAULT_MAX_RETRIES, **kwargs):
return make_request('GET', url, max_retries=max_retries, **kwargs)
def put(url, **kwargs):
return make_request('PUT', url, **kwargs)
def post(url, max_retries=SSRF_DEFAULT_MAX_RETRIES, **kwargs):
return make_request('POST', url, max_retries=max_retries, **kwargs)
def patch(url, **kwargs):
return make_request('PATCH', url, **kwargs)
def put(url, max_retries=SSRF_DEFAULT_MAX_RETRIES, **kwargs):
return make_request('PUT', url, max_retries=max_retries, **kwargs)
def delete(url, **kwargs):
return make_request('DELETE', url, **kwargs)
def patch(url, max_retries=SSRF_DEFAULT_MAX_RETRIES, **kwargs):
return make_request('PATCH', url, max_retries=max_retries, **kwargs)
def head(url, **kwargs):
return make_request('HEAD', url, **kwargs)
def delete(url, max_retries=SSRF_DEFAULT_MAX_RETRIES, **kwargs):
return make_request('DELETE', url, max_retries=max_retries, **kwargs)
def head(url, max_retries=SSRF_DEFAULT_MAX_RETRIES, **kwargs):
return make_request('HEAD', url, max_retries=max_retries, **kwargs)