Lambda S3触发器报No such file or directory错误，需修改哪部分代码？

核心报错原因 #

你遇到的No such file or directory报错是因为代码仅拼接了本地临时存储路径，没有实际将S3上的图片下载到Lambda本地的/tmp目录，直接读取不存在的路径就会触发该错误。除此之外代码还有2处逻辑/依赖问题需要同步修正：

需修改的内容 #

• 1. 补充S3文件下载逻辑：获取到bucket和key后，调用download_file接口将S3文件写入本地/tmp路径
• 2. 修正代码缩进：文件读取、模型推理的逻辑必须放到for循环内部，否则仅能处理最后一条事件记录，甚至会出现变量未定义问题
• 3. 补全依赖导入：顶部新增import numpy as np，否则调用np.argmax会报模块不存在错误

修正后完整代码 #

import urllib.parse
import boto3
import uuid
import os
import io
import json
import csv
import numpy as np

# 优先读取环境变量的终端节点名，不存在则用默认值
ENDPOINT_NAME = os.environ.get('ENDPOINT_NAME', 'T4')
runtime= boto3.client('runtime.sagemaker')
s3_client = boto3.client('s3')

def handler(event, context):
    try:
        for record in event['Records']:
            bucket = record['s3']['bucket']['name']
            key = record['s3']['object']['key']
            download_path = '/tmp/{}'.format(key)
            # 新增S3文件下载逻辑
            s3_client.download_file(bucket, key, download_path)
            
            with open(download_path, 'rb') as f:
                payload = f.read()
                payload = bytearray(payload)
            
            response = runtime.invoke_endpoint(
                EndpointName=ENDPOINT_NAME,
                ContentType='application/x-image',
                Body=payload
            )
                                   
            result = json.loads(response['Body'].read().decode())
            print(result)
            index = np.argmax(result)
            image_categories = ['glass', 'plastic bag', 'food', 'metal', 'styrofoam', 'paper','plastic', 'clothing']
            print("Result: label - " + image_categories[index] + ", probability - " + str(result[index]))
            return image_categories[index]
    except Exception as e:
        print(e)

内容的提问来源于stack exchange，提问作者최희재