Setting Up CloudWatch Alarms with Boto3 to Monitor EC2 Instance Performance

monitor_ec2_instance.py

import boto3

# Specify the region where the EC2 instances reside
region = 'us-east-1'

# Specify the tag key and value to filter the EC2 instances
tag_key = 'env'
tag_value = 'prod'


#Specify sns topic
sns_arn = "arn:aws:sns:us-east-1:335055665325:EC2_Alert"

# Specify the alarm thresholds for CPU and Memory utilization
cpu_threshold = 50.0
memory_threshold = 60.0

# Create a CloudWatch client
cloudwatch = boto3.client('cloudwatch', region_name=region)

# Get the list of EC2 instances with the specified tag
ec2_client = boto3.client('ec2', region_name=region)
response = ec2_client.describe_instances(
    Filters=[
        {
            'Name': 'tag:' + tag_key,
            'Values': [tag_value]
        },
        {
            'Name': 'instance-state-name',
            'Values': ['running']
        }
    ]
)


ec2 = boto3.client('ec2', region_name='us-east-1')
instance_ids = [instance['InstanceId'] for instance in ec2.describe_instances()['Reservations'][0]['Instances']]


for reservation in response['Reservations']:
    for instance in reservation['Instances']:
        instance_ids.append(instance['InstanceId'])

# Create CloudWatch alarms for CPU and Memory utilization for each EC2 instance
for instance_id in instance_ids:
    # Create CPU utilization alarm
    cloudwatch.put_metric_alarm(
        AlarmName=f'CPUUtilizationAlarm-{instance_id}-sreuniversity',
        ComparisonOperator='GreaterThanOrEqualToThreshold',
        EvaluationPeriods=3,
        Threshold=cpu_threshold,
        Period=60,
        Namespace='AWS/EC2',
        MetricName='CPUUtilization',
        Statistic='Average',
        Dimensions=[
            {
                'Name': 'InstanceId',
                'Value': instance_id
            },
        ],
        AlarmActions=[
            sns_arn # Specify your SNS topic ARN here to receive notifications
        ]
    )

    # Create Memory utilization alarm
    cloudwatch.put_metric_alarm(
        AlarmName=f'MemoryUtilizationAlarm-{instance_id}-sreuniversity',
        ComparisonOperator='GreaterThanOrEqualToThreshold',
        EvaluationPeriods=3,
        Threshold=memory_threshold,
        Period=60,
        Namespace='CWAgent',
        MetricName='MemoryUtilization',
        Statistic='Average',
        Dimensions=[
            {
                'Name': 'InstanceId',
                'Value': instance_id
            },
        ],
        AlarmActions=[
            sns_arn  # Specify your SNS topic ARN here to receive notifications
        ]
    )