mirror of
https://github.com/langgenius/dify.git
synced 2026-09-24 23:22:26 +08:00
perf: use global httpx client instead of per request create new one (#34311)
Co-authored-by: autofix-ci[bot] <114827586+autofix-ci[bot]@users.noreply.github.com>
This commit is contained in:
wangxiaolei
co-authored by
autofix-ci[bot]
autofix-ci[bot]
parent
19530e880a
commit
57f358a96b
@@ -2,8 +2,14 @@ import json
|
||||
|
||||
import httpx
|
||||
|
||||
from core.helper.http_client_pooling import get_pooled_http_client
|
||||
from services.auth.api_key_auth_base import ApiKeyAuthBase, AuthCredentials
|
||||
|
||||
_http_client: httpx.Client = get_pooled_http_client(
|
||||
"auth:jina_standalone",
|
||||
lambda: httpx.Client(limits=httpx.Limits(max_keepalive_connections=50, max_connections=100)),
|
||||
)
|
||||
|
||||
|
||||
class JinaAuth(ApiKeyAuthBase):
|
||||
def __init__(self, credentials: AuthCredentials):
|
||||
@@ -31,7 +37,7 @@ class JinaAuth(ApiKeyAuthBase):
|
||||
return {"Content-Type": "application/json", "Authorization": f"Bearer {self.api_key}"}
|
||||
|
||||
def _post_request(self, url, data, headers):
|
||||
return httpx.post(url, headers=headers, json=data)
|
||||
return _http_client.post(url, headers=headers, json=data)
|
||||
|
||||
def _handle_error(self, response):
|
||||
if response.status_code in {402, 409, 500}:
|
||||
|
||||
@@ -2,8 +2,14 @@ import json
|
||||
|
||||
import httpx
|
||||
|
||||
from core.helper.http_client_pooling import get_pooled_http_client
|
||||
from services.auth.api_key_auth_base import ApiKeyAuthBase, AuthCredentials
|
||||
|
||||
_http_client: httpx.Client = get_pooled_http_client(
|
||||
"auth:jina",
|
||||
lambda: httpx.Client(limits=httpx.Limits(max_keepalive_connections=50, max_connections=100)),
|
||||
)
|
||||
|
||||
|
||||
class JinaAuth(ApiKeyAuthBase):
|
||||
def __init__(self, credentials: AuthCredentials):
|
||||
@@ -31,7 +37,7 @@ class JinaAuth(ApiKeyAuthBase):
|
||||
return {"Content-Type": "application/json", "Authorization": f"Bearer {self.api_key}"}
|
||||
|
||||
def _post_request(self, url, data, headers):
|
||||
return httpx.post(url, headers=headers, json=data)
|
||||
return _http_client.post(url, headers=headers, json=data)
|
||||
|
||||
def _handle_error(self, response):
|
||||
if response.status_code in {402, 409, 500}:
|
||||
|
||||
@@ -10,6 +10,7 @@ from tenacity import retry, retry_if_exception_type, stop_before_delay, wait_fix
|
||||
from typing_extensions import TypedDict
|
||||
from werkzeug.exceptions import InternalServerError
|
||||
|
||||
from core.helper.http_client_pooling import get_pooled_http_client
|
||||
from enums.cloud_plan import CloudPlan
|
||||
from extensions.ext_database import db
|
||||
from extensions.ext_redis import redis_client
|
||||
@@ -18,6 +19,11 @@ from models import Account, TenantAccountJoin, TenantAccountRole
|
||||
|
||||
logger = logging.getLogger(__name__)
|
||||
|
||||
_http_client: httpx.Client = get_pooled_http_client(
|
||||
"billing:default",
|
||||
lambda: httpx.Client(limits=httpx.Limits(max_keepalive_connections=50, max_connections=100)),
|
||||
)
|
||||
|
||||
|
||||
class SubscriptionPlan(TypedDict):
|
||||
"""Tenant subscriptionplan information."""
|
||||
@@ -131,7 +137,7 @@ class BillingService:
|
||||
headers = {"Content-Type": "application/json", "Billing-Api-Secret-Key": cls.secret_key}
|
||||
|
||||
url = f"{cls.base_url}{endpoint}"
|
||||
response = httpx.request(method, url, json=json, params=params, headers=headers, follow_redirects=True)
|
||||
response = _http_client.request(method, url, json=json, params=params, headers=headers, follow_redirects=True)
|
||||
if method == "GET" and response.status_code != httpx.codes.OK:
|
||||
raise ValueError("Unable to retrieve billing information. Please try again later or contact support.")
|
||||
if method == "PUT":
|
||||
|
||||
@@ -9,12 +9,23 @@ import httpx
|
||||
from flask_login import current_user
|
||||
|
||||
from core.helper import encrypter
|
||||
from core.helper.http_client_pooling import get_pooled_http_client
|
||||
from core.rag.extractor.firecrawl.firecrawl_app import CrawlStatusResponse, FirecrawlApp, FirecrawlDocumentData
|
||||
from core.rag.extractor.watercrawl.provider import WaterCrawlProvider
|
||||
from extensions.ext_redis import redis_client
|
||||
from extensions.ext_storage import storage
|
||||
from services.datasource_provider_service import DatasourceProviderService
|
||||
|
||||
# Reuse pooled HTTP clients to avoid creating new connections per request and ease testing.
|
||||
_jina_http_client: httpx.Client = get_pooled_http_client(
|
||||
"website:jinareader",
|
||||
lambda: httpx.Client(limits=httpx.Limits(max_keepalive_connections=50, max_connections=100)),
|
||||
)
|
||||
_adaptive_http_client: httpx.Client = get_pooled_http_client(
|
||||
"website:adaptivecrawl",
|
||||
lambda: httpx.Client(limits=httpx.Limits(max_keepalive_connections=50, max_connections=100)),
|
||||
)
|
||||
|
||||
|
||||
@dataclass
|
||||
class CrawlOptions:
|
||||
@@ -225,7 +236,7 @@ class WebsiteService:
|
||||
@classmethod
|
||||
def _crawl_with_jinareader(cls, request: CrawlRequest, api_key: str) -> dict[str, Any]:
|
||||
if not request.options.crawl_sub_pages:
|
||||
response = httpx.get(
|
||||
response = _jina_http_client.get(
|
||||
f"https://r.jina.ai/{request.url}",
|
||||
headers={"Accept": "application/json", "Authorization": f"Bearer {api_key}"},
|
||||
)
|
||||
@@ -233,7 +244,7 @@ class WebsiteService:
|
||||
raise ValueError("Failed to crawl:")
|
||||
return {"status": "active", "data": response.json().get("data")}
|
||||
else:
|
||||
response = httpx.post(
|
||||
response = _adaptive_http_client.post(
|
||||
"https://adaptivecrawl-kir3wx7b3a-uc.a.run.app",
|
||||
json={
|
||||
"url": request.url,
|
||||
@@ -296,7 +307,7 @@ class WebsiteService:
|
||||
|
||||
@classmethod
|
||||
def _get_jinareader_status(cls, job_id: str, api_key: str) -> dict[str, Any]:
|
||||
response = httpx.post(
|
||||
response = _adaptive_http_client.post(
|
||||
"https://adaptivecrawlstatus-kir3wx7b3a-uc.a.run.app",
|
||||
headers={"Content-Type": "application/json", "Authorization": f"Bearer {api_key}"},
|
||||
json={"taskId": job_id},
|
||||
@@ -312,7 +323,7 @@ class WebsiteService:
|
||||
}
|
||||
|
||||
if crawl_status_data["status"] == "completed":
|
||||
response = httpx.post(
|
||||
response = _adaptive_http_client.post(
|
||||
"https://adaptivecrawlstatus-kir3wx7b3a-uc.a.run.app",
|
||||
headers={"Content-Type": "application/json", "Authorization": f"Bearer {api_key}"},
|
||||
json={"taskId": job_id, "urls": list(data.get("processed", {}).keys())},
|
||||
@@ -374,7 +385,7 @@ class WebsiteService:
|
||||
@classmethod
|
||||
def _get_jinareader_url_data(cls, job_id: str, url: str, api_key: str) -> dict[str, Any] | None:
|
||||
if not job_id:
|
||||
response = httpx.get(
|
||||
response = _jina_http_client.get(
|
||||
f"https://r.jina.ai/{url}",
|
||||
headers={"Accept": "application/json", "Authorization": f"Bearer {api_key}"},
|
||||
)
|
||||
@@ -383,7 +394,7 @@ class WebsiteService:
|
||||
return dict(response.json().get("data", {}))
|
||||
else:
|
||||
# Get crawl status first
|
||||
status_response = httpx.post(
|
||||
status_response = _adaptive_http_client.post(
|
||||
"https://adaptivecrawlstatus-kir3wx7b3a-uc.a.run.app",
|
||||
headers={"Content-Type": "application/json", "Authorization": f"Bearer {api_key}"},
|
||||
json={"taskId": job_id},
|
||||
@@ -393,7 +404,7 @@ class WebsiteService:
|
||||
raise ValueError("Crawl job is not completed")
|
||||
|
||||
# Get processed data
|
||||
data_response = httpx.post(
|
||||
data_response = _adaptive_http_client.post(
|
||||
"https://adaptivecrawlstatus-kir3wx7b3a-uc.a.run.app",
|
||||
headers={"Content-Type": "application/json", "Authorization": f"Bearer {api_key}"},
|
||||
json={"taskId": job_id, "urls": list(status_data.get("processed", {}).keys())},
|
||||
|
||||
Reference in New Issue
Block a user