Lambda function for subscribe slurmctld logs from Slurm on AWS ParallelCluster and submit Job messages to Slack

pcluster-slurm-slack.py

import json
import logging
import base64
import gzip
import urllib.request

LOG_STREAM = 'ip-10-0-1-xxx.i-08f68c623be216d6e.slurmctld'
SLACK_WEB_HOOK = 'https://hooks.slack.com/services/XXXX'

logger = logging.getLogger()
logger.setLevel(logging.INFO)


def lambda_handler(event, context):
    # TODO implement
    decoded_data = base64.b64decode(event['awslogs']['data'])
    json_data = json.loads(gzip.decompress(decoded_data))
    logger.info("EVENT: " + json.dumps(json_data))
    
    if json_data['logStream'] == LOG_STREAM:
        for i in range(len(json_data['logEvents'])):
            message = json_data['logEvents'][i]['message']
            print(message)
            send_data = {
                "text": message
            }
            send_text = json.dumps(send_data)
            request = urllib.request.Request(
                SLACK_WEB_HOOK, 
                data=send_text.encode('utf-8'), 
                method="POST"
            )
            with urllib.request.urlopen(request) as response:
                response_body = response.read().decode('utf-8')

    return {
        'statusCode': 200,
        'body': json.dumps('Hello from Lambda!')
    }