基于Python通过Logic Apps运行长耗时Azure Functions遇阻求助

我尝试解决这个问题已有一段时间，始终没找到可行方案。需求是通过Azure Functions在Logic Apps中运行一项长耗时任务：用Python代码对比并转换两个包含近2万行数据（后续还会新增）的Excel文件。本地运行Python代码耗时约8分钟，转为Azure Functions后本地运行需11分钟，这显然会触发Logic Apps的超时限制。

我采用了HTTP webhook方案，思路是创建一个接收请求的函数，收到请求后立即返回202 Accepted响应，同时启动线程执行长耗时任务，任务完成后调用传入的callbackUri。实现代码如下：

import logging
import azure.functions as func
import time
import requests
import threading


app = func.FunctionApp(http_auth_level=func.AuthLevel.ANONYMOUS)


@app.function_name("http_trigger1")
@app.route("http_trigger1")
def http_trigger1(req: func.HttpRequest) -> func.HttpResponse:
    logging.info('Webhook request from Logic Apps received.')

    callback_url = req.params.get('callback_url')
    data = req.params.get('data')
    if not callback_url:
        try:
            req_body = req.get_json()
        except ValueError:
            pass
        else:
            callback_url = req_body.get('callback_url')
            data = req_body.get('data')
            
    if not callback_url:
        return func.HttpResponse(
            "Enter a valid callback URL",
            status_code=400
        )
        
    try:
        # 立即返回202 Accepted
        threading.Thread(target=process_and_callback, args=(callback_url, data)).start()
        
        return func.HttpResponse(
            "Accepted for processing",
            status_code=202
        )
    
    except Exception as e:
        logging.error(f'An error occurred: {str(e)}')
        return func.HttpResponse(
            "Error occurred while invoking callback",
            status_code=500
        )

def process_and_callback(callback_url: str, data: str) -> None:
    try:
        # 模拟长耗时任务
        time.sleep(720)
        
        callback_data = {
            "Subject": data 
        }
        
        # 发起回调请求
        response = requests.post(callback_url, json=callback_data)
        response.raise_for_status() 

        logging.info(f'Callback successful with status code: {response.status_code}')
    
    except Exception as e:
        logging.error(f'An error occurred while invoking callback: {str(e)}')

用12分钟睡眠模拟长耗时任务时，该方案可正常绕过Logic Apps的超时限制，但替换为实际Excel处理代码后出现两个问题：

• Logic Apps间歇性返回RequestEntityTooLarge错误（请求体包含callbackUri及两个Base64编码的Excel文件）；
• 重新提交后任务陷入运行状态超1.5小时（本地仅需11分钟）。

测试小文件时方案可行，仅大文件出现问题。以下是实际长耗时任务的核心代码片段：

# 读取Excel文件'file1'
wb_file1 = pd.read_excel(file1)
column_A = [f"{wb_file1.iloc[i, 7]},{wb_file1.iloc[i, 20]},{wb_file1.iloc[i, 21]}" for i in range(1, len(wb_file1))]

# 读取Excel文件'file2'
wb_file2 = pd.read_excel(file2, sheet_name=sheet_name)

column_a_out = {}
duplicates = []
for i, a in enumerate(column_A):
    row_number =  wb_file2.index[(wb_file2.iloc[:, 7].astype(str) + ',' +
                            wb_file2.iloc[:, 20].astype(str) + ',' +
                            wb_file2.iloc[:, 21].astype(str)) == a].tolist()
    if not len(row_number) > 1:
        row = int(row_number[0]) + 2
        column_a = a
        column_f = f"{wb_file1.iloc[i+1, 5]}"
        column_ad = f"{wb_file1.iloc[i+1, 29]}"
        column_af = f"{wb_file1.iloc[i+1, 31]}"
        
        column_a_file1[i+1] = [row, column_a, column_f, column_ad, column_af]
    else:
        duplicates.append(a)

# 加载工作簿到openpyxl进行操作
wb_file2 = openpyxl.load_workbook(file2)
sheet_file2 = wb_file2[sheet_name]   

for items in column_a_out:
    row_num, a, f, ad, af = column_a_out[items]
    sheet_file2.cell(row=row_num, column=6, value="" if f == 'nan' else f)
    sheet_file2.cell(row=row_num, column=30, value="" if ad == 'nan' else ad)
    sheet_file2.cell(row=row_num, column=32, value="" if af == 'nan' else af)

希望有人能帮我解决该问题，如需更多信息可随时提供。

内容的提问来源于stack exchange，提问作者No Name