Google Cloud Run部署Python Flask API遇实例启动超时问题求助

我在Google Cloud Run中部署了一个Flask应用作为API服务。当Cloud Run启动新实例处理API请求时，该Flask应用无法正常响应请求，但从Cloud Run控制台日志来看，应用似乎已成功启动。

Cloud Run控制台日志片段 #

2024-04-03 19:30:57.734 PDT  [2024-04-04 02:30:57 +0000] [1] [INFO] Starting gunicorn 21.2.0
2024-04-03 19:30:57.735 PDT  [2024-04-04 02:30:57 +0000] [1] [INFO] Listening at: http://0.0.0.0:8080 (1)
2024-04-03 19:30:57.735 PDT  [2024-04-04 02:30:57 +0000] [1] [INFO] Using worker: gthread
2024-04-03 19:30:57.757 PDT  [2024-04-04 02:30:57 +0000] [2] [INFO] Booting worker with pid: 2
2024-04-03 19:30:58.672 PDT  APP Loading: 0.8971238136291504 sec
2024-04-03 19:31:04.226 PDT  [nltk_data] Downloading package punkt to /home/nltk_data...
2024-04-03 19:31:04.516 PDT  [nltk_data]   Unzipping tokenizers/punkt.zip.
2024-04-03 19:31:05.829 PDT  [nltk_data] Downloading package stopwords to /home/nltk_data...
2024-04-03 19:31:05.908 PDT  [nltk_data]   Unzipping corpora/stopwords.zip.
2024-04-03 19:31:08.050 PDT  STARTUP HTTP probe succeeded after 3 attempts for container "backend-1" on path "/".
2024-04-03 19:31:54.527 PDT  GET  504  713 B  60 s  Chrome 123 https://backend-uw.a.run.app/reports?user=12345 

main.py代码片段 #

import time
__init_time = time.time()

import json
import logging
import os
import shutil
from threading import Thread
import uuid

from flask import Flask
from flask import request
from flask import abort
from flask import jsonify
from flask import Response
from flask_cors import CORS
from lazy_import import lazy_class

# 延迟加载耗时较长的模块，以降低后端应用加载延迟
GptController = lazy_class(
    'controls.gpt_controller.GptController')
DocumentsController = lazy_class(
    'controls.documents_controller.DocumentsController')
FirestoreController = lazy_class(
    'controls.firestore_controller.FirestoreController')
UserController = lazy_class(
    'controls.user_controller.UserController')
Utilities = lazy_class(
    'utils.utilities.Utilities')

logging.getLogger('requests').setLevel(logging.ERROR)
logging.getLogger('httpx').setLevel(logging.ERROR)

TEMP_FOLDER = 'temp'

app = Flask(__name__)
cors = CORS(app)
app.config['CORS_HEADERS'] = 'Content-Type'

# 初始化应用时创建所需文件夹
os.makedirs(
    os.path.join(app.instance_path, TEMP_FOLDER),
    exist_ok=True)

@app.route("/", methods=["GET"])
def home():
    """Cloud Run启动健康检查的入口路由"""
    __time = time.time()
    while (time.time()-__time) < 10:
        try:
            if FirestoreController().firestore_client:
                return "<h1>App Alive</h1>"
        except Exception as _:
            pass
        time.sleep(0.5)
    abort(500, 'Server not ready')

@app.route("/reports", methods=["GET"])
def report_operations():
    """
    处理'report'相关API操作的入口函数
    """
    return { "status": "success" }


print(f'APP Loading: {time.time() - __init_time} sec')

if __name__ == "__main__":
    app.run(
        debug=True,
        host="0.0.0.0",
        port=int(os.environ.get("PORT", 8080))
    )

Dockerfile代码 #

FROM python:3.9

# 允许日志信息立即出现在Knative日志中
ENV PYTHONUNBUFFERED True

# 将本地代码复制到容器镜像中
ENV APP_HOME /app
WORKDIR $APP_HOME
COPY . ./

# 安装Python依赖
RUN python -m venv xplor-env
ENV PATH="xplor-env/bin:$PATH"
RUN pip install --upgrade pip
RUN pip install gunicorn
RUN pip install -r requirements.txt

CMD exec gunicorn --bind :$PORT --workers 1 --threads 8 --timeout 0 main:app

已采取的优化措施 #

• 使用懒加载将应用加载延迟降至约1秒
• 在应用的/路由添加了HTTP启动健康检查探针

内容的提问来源于stack exchange，提问作者splashmonkey