如何通过Python编程暂停MongoDB Atlas集群？两种尝试均失败

需要实现MongoDB Atlas集群的自动暂停功能，已知Atlas CLI提供atlas clusters pause [options]命令，但希望通过Python完成该操作（PyMongo不支持此管理类操作），尝试以下两种方法均未成功：

尝试过的方法 #

方法一：调用Atlas CLI命令 #

import os
os.system("atlas clusters pause Cluster0 --projectId 662b391615a47202d49f2fa0 --output json")

方法二：调用MongoDB Atlas API #

import requests
import json

# Atlas API configuration
base_url = "https://cloud.mongodb.com/api/atlas/v1.0"
project_id = "662b391615a47202d49f2fa0"
cluster_name = "Cluster0"

# Your Atlas API credentials
public_key = "YOUR_PUBLIC_KEY"
#private_key = "YOUR_PRIVATE_KEY"

# Endpoint for pausing a cluster
endpoint = f"{base_url}/groups/{project_id}/clusters/{cluster_name}"

# Request headers
headers = {
    "Content-Type": "application/json",
    "Accept": "application/json"
}

# Request payload
payload = {
    "paused": True
}

# Send the PATCH request
response = requests.patch(
    endpoint,
    auth=(public_key),
    headers=headers,
    data=json.dumps(payload)
)

# Check the response
if response.status_code == 200:
    print(f"Cluster {cluster_name} paused successfully.")
else:
    print(f"Failed to pause cluster. Status code: {response.status_code}")
    print(response.text)

问题分析与修正方案 #

方法一问题修正 #

os.system无法捕获详细错误信息，且依赖CLI的环境配置（需确保已通过atlas login完成认证），改用subprocess更便于调试和控制：

import subprocess

try:
    # 执行CLI命令，捕获输出和错误
    result = subprocess.run(
        ["atlas", "clusters", "pause", "Cluster0", "--projectId", "662b391615a47202d49f2fa0", "--output", "json"],
        capture_output=True,
        text=True,
        check=True
    )
    print("集群暂停成功:", result.stdout)
except subprocess.CalledProcessError as e:
    print("集群暂停失败:", e.stderr)

若未提前登录CLI，可直接在命令中传入凭证：

result = subprocess.run(
    ["atlas", "clusters", "pause", "Cluster0", "--projectId", "662b391615a47202d49f2fa0", "--publicKey", "你的公钥", "--privateKey", "你的私钥", "--output", "json"],
    capture_output=True,
    text=True,
    check=True
)

方法二问题修正 #

原代码存在两个关键错误：

1. API认证需同时传入公钥和私钥，auth参数应为元组(public_key, private_key)
2. Atlas API修改集群状态为异步操作，响应状态码通常是202 Accepted而非200

修正后的代码：

import requests

base_url = "https://cloud.mongodb.com/api/atlas/v1.0"
project_id = "662b391615a47202d49f2fa0"
cluster_name = "Cluster0"

# 补全私钥
public_key = "YOUR_PUBLIC_KEY"
private_key = "YOUR_PRIVATE_KEY"

endpoint = f"{base_url}/groups/{project_id}/clusters/{cluster_name}"

headers = {
    "Content-Type": "application/json",
    "Accept": "application/json"
}

payload = {
    "paused": True
}

response = requests.patch(
    endpoint,
    auth=(public_key, private_key),  # 传入完整密钥对
    headers=headers,
    json=payload  # 直接用json参数自动序列化，无需手动dumps
)

if response.status_code in (200, 202):
    print(f"Cluster {cluster_name} 暂停请求已受理，正在处理中")
else:
    print(f"暂停失败，状态码: {response.status_code}")
    print(response.text)

注意：需确保你的API密钥已被授予Cluster Manager角色权限，否则会返回权限不足的错误。

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