在当今的快节奏开发环境中，确保前端服务的高效和可靠更新至关重要。本文将介绍如何通过AWS服务（如ECS和S3）以及Jenkins进行参数化构建，实现前端服务的自动化部署和更新控制。我们将深入了解使用Python脚本管理部署流程，以及如何在Jenkins中配置参数化构建，使得服务的部署更加灵活和可配置。
部署脚本
Python脚本管理部署流程
首先，让我们来看一下负责编排部署流程的Python脚本。该脚本使用Boto3库与AWS服务进行交互，并利用Requests库发起HTTP请求。以下是关键功能的概述：
#!/usr/bin/python3
import os
import requests
import boto3
import json
def get_secret(secret_name, region_name):
"""
Retrieves the specified secret from AWS Secrets Manager.
Args:
secret_name (str): The name of the secret.
region_name (str): The AWS region where the secret is stored.
Returns:
dict: The decoded JSON representation of the secret.
"""
session = boto3.session.Session()
client = session.client(service_name='secretsmanager', region_name=region_name)
get_secret_value_response = client.get_secret_value(SecretId=secret_name)
return json.loads(get_secret_value_response['SecretString'])
def send_dingding(service_name, token):
"""
Sends a message to DingTalk notifying the successful restart of a service.
Args:
service_name (str): Name of the service being restarted.
token (str): DingTalk access token.
"""
headers = {'Content-Type': 'application/json'}
message = f"<font face='黑体' color='#00EC00'>Restart {service_name} Success</font>\n"
data = {"msgtype": "markdown", "markdown": {"title": f"Restart {service_name} Success", "text": message}}
response = requests.post(f"https://oapi.dingtalk.com/robot/send?access_token={token}", headers=headers, json=data)
print(response.text)
def update_ecs_service(service_name, cluster_name='frontend-pro'):
"""
Updates the ECS service with a new deployment.
Args:
service_name (str): Name of the ECS service.
cluster_name (str): Name of the ECS cluster.
"""
boto3.client('ecs').update_service(cluster=cluster_name, service=service_name, taskDefinition=service_name,
forceNewDeployment=True)
print("****************************更新容器************************************")
def upload_to_s3(service_name):
"""
Uploads code to S3 bucket.
Args:
service_name (str): Name of the service.
"""
s3_client = boto3.client('s3')
print("****************************上传代码************************************")
s3_client.sync(f"s3://deploy-github-fe/{service_name}/", f"s3://govee-static/prod/{service_name}/",
acl='public-read')
def clear_url_cache():
"""
Clears the URL cache by sending an HTTP POST request.
"""
response = requests.post("https://apph5-api.my.com/h5/v1/commom/cache/reset")
print(response.text)
def deploy_services(service_names):
"""
Deploys services based on their names.
Args:
service_names (list): List of service names to be deployed.
"""
for service_name in service_names:
if service_name == "pro-trade-fe":
upload_to_s3(service_name)
update_ecs_service(service_name)
send_dingding(service_name, os.environ['DINGDING_TOKEN'])
elif service_name in ["pro-auth-fe", "prod-h5-game-fe"]:
upload_to_s3(service_name)
clear_url_cache()
send_dingding(service_name, os.environ['DINGDING_TOKEN'])
elif service_name == "prod-life-web-fe":
update_ecs_service(service_name)
send_dingding(service_name, os.environ['DINGDING_TOKEN'])
elif service_name in ["prod-app-h5-fe", "prod-app-hd-fe"]:
upload_to_s3(service_name)
update_ecs_service(service_name)
clear_url_cache()
send_dingding(service_name, os.environ['DINGDING_TOKEN'])
elif service_name == "prod-minger-os-web-fe":
update_ecs_service(service_name)
send_dingding(service_name, os.environ['DINGDING_TOKEN'])
else:
upload_to_s3(service_name)
update_ecs_service(service_name)
clear_url_cache()
send_dingding(service_name, os.environ['DINGDING_TOKEN'])
if __name__ == "__main__":
# Get DingDing token from AWS Secrets Manager
secrets = get_secret('base', 'us-east-1')
token = secrets.get('BUILD1_TOKEN', '') # Retrieve the token, default to empty string if not found
# Get service names from environment variable
service_names = os.environ['SERVICE_NAMES'].split(',')
# Execute deployment
deploy_services(service_names)关键功能解析：
get_secret：从AWS Secrets Manager检索指定的秘密。send_dingding：向DingTalk发送有关服务重新启动的消息。update_ecs_service：使用新的部署更新ECS服务。upload_to_s3：将代码上传到S3桶。clear_url_cache：通过发送HTTP POST请求清除URL缓存。deploy_services：部署服务的主要函数。
Shell脚本管理部署流程
#!/bin/bash
# Function to get secret from AWS Secrets Manager
get_secret() {
secret_name=$1
region_name=$2
aws secretsmanager get-secret-value --secret-id $secret_name --region $region_name --query 'SecretString' --output text
}
# Function to send DingTalk message
send_dingding() {
service_name=$1
token=$2
message="<font face='黑体' color='#00EC00'>Restart $service_name Success</font>"
curl -s -X POST "https://oapi.dingtalk.com/robot/send?access_token=$token" -H "Content-Type: application/json" \
-d "{\"msgtype\": \"markdown\", \"markdown\": {\"title\": \"Restart $service_name Success\", \"text\": \"$message\"}}"
}
# Function to update ECS service
update_ecs_service() {
service_name=$1
cluster_name='frontend-pro'
aws ecs update-service --cluster $cluster_name --service $service_name --task-definition $service_name --force-new-deployment
echo "****************************更新容器************************************"
}
# Function to upload to S3
upload_to_s3() {
service_name=$1
aws s3 sync "s3://deploy-github-fe/$service_name/" "s3://govee-static/prod/$service_name/" --acl public-read
echo "****************************上传代码************************************"
}
# Function to clear URL cache
clear_url_cache() {
curl -s -X POST "https://apph5-api.my.com/h5/v1/commom/cache/reset"
echo "****************************清理URL缓存************************************"
}
# Function to deploy services
deploy_services() {
service_names=$@
for service_name in $service_names; do
case $service_name in
"pro-trade-fe")
upload_to_s3 $service_name
update_ecs_service $service_name
send_dingding $service_name $DINGDING_TOKEN
;;
"pro-auth-fe"|"prod-h5-game-fe")
upload_to_s3 $service_name
clear_url_cache
send_dingding $service_name $DINGDING_TOKEN
;;
"prod-life-web-fe")
update_ecs_service $service_name
send_dingding $service_name $DINGDING_TOKEN
;;
"prod-app-h5-fe"|"prod-app-hd-fe")
upload_to_s3 $service_name
update_ecs_service $service_name
clear_url_cache
send_dingding $service_name $DINGDING_TOKEN
;;
"prod-minger-os-web-fe")
update_ecs_service $service_name
send_dingding $service_name $DINGDING_TOKEN
;;
*)
upload_to_s3 $service_name
update_ecs_service $service_name
clear_url_cache
send_dingding $service_name $DINGDING_TOKEN
;;
esac
done
}
# Main execution
secrets=$(get_secret 'base' 'us-east-1')
export DINGDING_TOKEN=$(echo $secrets | jq -r .BUILD1_TOKEN)
# Get service names from environment variable
service_names=${SERVICE_NAMES//,/ }
# Execute deployment
deploy_services $service_namesJenkins中的参数化构建
为了使部署流程更加灵活和可配置，我们将该脚本集成到Jenkins中，并使用Extended Choice Parameter插件进行参数化构建。这使用户能够在Jenkins构建期间选择要部署的服务。
Jenkins配置
- Extended Choice Parameter:
- Name: SERVICE_NAMES
- Parameter Type: Check Boxes
- Number of Visible Items: 10
- Delimiter:,
- Quote Value: prod-app-hd-fe, prod-app-h5-fe, prod-govee-life-web-fe, prod-community-web-fe, prod-app-platform-fe, pro-trade-fe, pro-auth-fe, prod-h5-game-fe, pro-app-h5-fe, prod-promotion-app-h5-fe, prod-desktop-web-fe, prod-app-mall-fe, prod-minger-os-web-fe
结论
通过自动化部署流水线，我们不仅节省了时间，还减少了人为错误的风险。通过Python脚本和Jenkins的集成，只需几次点击即可轻松部署前端服务到AWS。根据您的特定用例进行定制，享受简化的部署流程。
在您的博文中，您可以进一步优化、整理和详细说明脚本的生产控制、开发最佳实践等方面的内容，以使读者更好地理解和应用这一自动化流程。祝愉快编写博文！