CloudWatchアラーム発報をトリガーにBacklogへ課題を自動作成するシステムを作ってみた

			
			import json
import requests
import os
import boto3
from datetime import datetime, timezone, timedelta
import logging

# --- Standard Configuration ---
# Set up logging to standard output
logger = logging.getLogger()
logger.setLevel(logging.INFO)

# Set up timezone for JST
JST = timezone(timedelta(hours=+9), 'JST')

# --- Backlog API Constants ---
# {your_space_id}の部分はご自身のBacklogスペースIDに書き換えてください
BACKLOG_API_URL = "https://{your_space_id}.backlog.jp/api/v2" 
# Priority: 3は検証したプロジェクトでは「中」に相当
DEFAULT_PRIORITY_ID = 3 
# Status: 1は検証したプロジェクト「未対応」に相当
STATUS_ID_OPEN = 1
# HTTP request timeout in seconds
REQUEST_TIMEOUT = 30

# --- AWS Clients (initialized once per container) ---
ssm = boto3.client('ssm')

def get_parameter(name):
    """
    Retrieves a parameter from AWS SSM Parameter Store.
    Trims leading/trailing whitespace.
    """
    try:
        # 環境変数でプレフィックスを定義することで、開発/本番環境の切り替えを容易にしている
        prefix = os.environ.get('PARAMETER_PREFIX', '/backlog/prod')
        param_path = f"{prefix}/{name}"
        response = ssm.get_parameter(Name=param_path, WithDecryption=True)
        value = response['Parameter']['Value'].strip()
        logger.info(f"Successfully retrieved parameter: {name}")
        return value
    except Exception as e:
        logger.critical(f"CRITICAL: Could not retrieve required parameter '{name}'. Error: {e}")
        raise

# --- Configuration Loading (runs once on Lambda cold start) ---
try:
    API_KEY = get_parameter('backlog-api-key')
    PROJECT_ID = get_parameter('project-id')
    ISSUE_TYPE_ID = get_parameter('issue-type-id')
    ASSIGNEE_ID = get_parameter('assignee-id')
    CATEGORY_ID = get_parameter('category-id')
    STATUS_ID_CLOSED = get_parameter('status-id-closed')

    notify_users_str = get_parameter('notify-user-ids')
    NOTIFY_USER_IDS = [user_id.strip() for user_id in notify_users_str.split(',') if user_id.strip()]

except Exception:
    API_KEY = None
    logger.critical("Failed to load one or more required parameters during initialization.")

def lambda_handler(event, context):
    """
    Main handler for the Lambda function.
    Parses an SNS message from a CloudWatch alarm.
    Creates a Backlog issue on ALARM, closes it on OK.
    """
    if not API_KEY:
        logger.error("Configuration is missing or failed to load. Aborting execution.")
        return {'statusCode': 500, 'body': json.dumps('Configuration error during initialization.')}

    try:
        sns_message = json.loads(event['Records'][0]['Sns']['Message'])
        alarm_name = sns_message.get('AlarmName', 'Unknown Alarm')
        logger.info(f"Processing alarm: {alarm_name}")

        state = sns_message.get('NewStateValue')
        if state == 'ALARM':
            create_backlog_issue(sns_message)
        elif state == 'OK':
            close_backlog_issue(sns_message)
        else:
            logger.warning(f"Received unknown alarm state: {state}")

        return {'statusCode': 200, 'body': json.dumps('Process completed successfully.')}

    except (json.JSONDecodeError, IndexError, KeyError) as e:
        logger.error(f"Failed to parse SNS message event. Error: {str(e)}. Event: {event}")
        return {'statusCode': 400, 'body': json.dumps('Bad request: Invalid event structure.')}
    except Exception as e:
        logger.error(f"An unexpected error occurred in lambda_handler: {str(e)}", exc_info=True)
        return {'statusCode': 500, 'body': json.dumps(f'Internal Server Error: {str(e)}')}

def close_backlog_issue(message):
    """
    Finds the corresponding open issue for a recovered alarm and closes it.
    """
    alarm_name = message.get('AlarmName', 'N/A')
    logger.info(f"Attempting to close issue for recovered alarm: {alarm_name}")

    try:
        search_params = {
            'apiKey': API_KEY,
            'projectId[]': [PROJECT_ID],
            'statusId[]': [STATUS_ID_OPEN],
            'keyword': alarm_name,
            'sort': 'updated',
            'order': 'desc',
            'count': 1
        }
        response = requests.get(f"{BACKLOG_API_URL}/issues", params=search_params, timeout=REQUEST_TIMEOUT)
        response.raise_for_status()

        issues = response.json()
        if not issues:
            logger.warning(f"No open issue found for alarm '{alarm_name}'. Nothing to close.")
            return

        target_issue = issues[0]
        issue_key = target_issue.get('issueKey')
        logger.info(f"Found open issue to close: {issue_key}")

    except requests.exceptions.RequestException as e:
        logger.error(f"Failed to search for Backlog issue. Status: {e.response.status_code if e.response else 'N/A'}, Response: {e.response.text if e.response else 'N/A'}")
        raise

    try:
        jst_timestamp_for_desc = datetime.now(JST).strftime('%Y-%-m-%-d %H:%M:%S %Z')
        update_data = {
            'statusId': STATUS_ID_CLOSED,
            'comment': f"アラートが復旧しました。\n発生時刻 (JST): {jst_timestamp_for_desc}"
        }
        update_params = {'apiKey': API_KEY}

        response = requests.patch(
            f"{BACKLOG_API_URL}/issues/{issue_key}",
            params=update_params,
            data=update_data,
            timeout=REQUEST_TIMEOUT
        )
        response.raise_for_status()
        logger.info(f"Successfully closed issue {issue_key}")

    except requests.exceptions.RequestException as e:
        logger.error(f"Failed to update Backlog issue {issue_key}. Status: {e.response.status_code}, Response: {e.response.text}")
        raise

def create_backlog_issue(message):
    """
    Constructs and sends a request to create a new issue in Backlog.
    """
    alarm_name = message.get('AlarmName', 'N/A')
    utc_timestamp_str = message.get('StateChangeTime', '')

    try:
        utc_dt = datetime.fromisoformat(utc_timestamp_str.replace('Z', '+00:00'))
        jst_dt = utc_dt.astimezone(JST)
        date_str_for_title = jst_dt.strftime('%y%m%d%H%M')
        jst_timestamp_for_desc = jst_dt.strftime('%Y-%-m-%d %H:%M:%S %Z')
    except (ValueError, TypeError):
        logger.warning(f"Could not parse timestamp: {utc_timestamp_str}. Using current time.")
        jst_dt = datetime.now(JST)
        date_str_for_title = jst_dt.strftime('%y%m%d%H%M')
        jst_timestamp_for_desc = jst_dt.strftime('%Y-%m-%d %H:%M:%S %Z')

    title = f"【アラート発生】{date_str_for_title}-{alarm_name}"
    description = f"""お世話になっております。
CloudWatchアラートの発生を検知しました。

## ■ アラート内容
{message.get('NewStateReason', '詳細不明')}

## ■ 詳細情報
- **アラーム名:** {alarm_name}
- **発生時刻 (JST):** {jst_timestamp_for_desc}
- **リージョン:** {message.get('Region', 'N/A')}
- **アカウントID:** {message.get('AWSAccountId', 'N/A')}
"""

    today_jst = datetime.now(JST).strftime('%Y-%m-%d')
    params = {'apiKey': API_KEY}
    data = {
        'projectId': PROJECT_ID,
        'summary': title,
        'description': description,
        'issueTypeId': ISSUE_TYPE_ID,
        'priorityId': DEFAULT_PRIORITY_ID,
        'assigneeId': ASSIGNEE_ID,
        'categoryId[]': [CATEGORY_ID],
        'startDate': today_jst,
        'dueDate': today_jst,
        'notifiedUserId[]': NOTIFY_USER_IDS
    }

    try:
        logger.info(f"Sending request to create issue for alarm: {alarm_name}")
        response = requests.post(
            f"{BACKLOG_API_URL}/issues",
            params=params,
            data=data,
            timeout=REQUEST_TIMEOUT
        )
        response.raise_for_status()
        issue_data = response.json()
        issue_key = issue_data.get('issueKey', issue_data.get('id', 'N/A'))
        logger.info(f"Successfully created Backlog issue: {issue_key}")
    except requests.exceptions.HTTPError as e:
        logger.error(f"Failed to create Backlog issue. Status: {e.response.status_code}, Response Body: {e.response.text}")
        raise
    except requests.exceptions.RequestException as e:
        logger.error(f"HTTP request to Backlog failed: {str(e)}")
        raise

		

			
			
{
  "Records": [
    {
      "EventSource": "aws:sns",
      "EventVersion": "1.0",
      "EventSubscriptionArn": "arn:aws:sns:ap-northeast-1:XXXXXXXXXXXX:test-topic",
      "Sns": {
        "Type": "Notification",
        "MessageId": "XXXXXXXX-XXXX-XXXX-XXXX-XXXXXXXXXXXX",
        "TopicArn": "arn:aws:sns:ap-northeast-1:XXXXXXXXXXXX:test-topic",
        "Subject": "ALARM: test-alarm",
        "Message": "{\"AlarmName\":\"test-alarm\",\"AlarmDescription\":\"Test alarm description\",\"AWSAccountId\":\"XXXXXXXXXXXX\",\"NewStateValue\":\"ALARM\",\"NewStateReason\":\"Threshold Crossed: 1 datapoint [5.0 (14/09/25 08:00:00)] was greater than the threshold (1.0).\",\"StateChangeTime\":\"2025-09-14T08:00:00.000Z\",\"Region\":\"ap-northeast-1\",\"OldStateValue\":\"OK\",\"Trigger\":{\"MetricName\":\"CPUUtilization\",\"Namespace\":\"AWS/EC2\",\"StatisticType\":\"Statistic\",\"Statistic\":\"AVERAGE\",\"Unit\":null,\"Dimensions\":[{\"name\":\"InstanceId\",\"value\":\"i-XXXXXXXXXXXXXXXXX\"}],\"Period\":300,\"EvaluationPeriods\":1,\"ComparisonOperator\":\"GreaterThanThreshold\",\"Threshold\":1.0}}",
        "Timestamp": "2025-09-14T08:00:00.000Z",
        "SignatureVersion": "1",
        "Signature": "XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX",
        "SigningCertUrl": "https://example.com/cert",
        "UnsubscribeUrl": "https://example.com/unsubscribe"
      }
    }
  ]
}


		

			
			
{
    "Version": "2012-10-17",
    "Statement": [
        {
            "Sid": "ParameterStoreAccess",
            "Effect": "Allow",
            "Action": [
                "ssm:GetParameter",
                "ssm:GetParameters",
                "ssm:DescribeParameters"
            ],
            "Resource": [
                "arn:aws:ssm:ap-northeast-1:123456789012:parameter/backlog/prod/*"
            ]
        },
        {
            "Sid": "KMSDecryptForSSM",
            "Effect": "Allow",
            "Action": "kms:Decrypt",
            "Resource": "arn:aws:kms:ap-northeast-1:123456789012:key/*"
        }
    ]
}