DEV Community

Cover image for Execute a lambda every X minutes
Maxime Guilbert
Maxime Guilbert

Posted on • Edited on

Execute a lambda every X minutes

Having a treatment that is executed regularly can be really important for a lot of project to do a lot of little tasks. So we will see today how we can do it in AWS!


Global vision

Image description

As you can see, it's a really simple architecture to be able to do it. In fact, we only need EventBridge to do it.

AWS EventBridge

Quick overview about AWS EventBridge.

AWS EventBridge is an AWS service which facilitate event driven architectures. It allows with all its features to plan execution of tasks regularly, or if every time some specific actions happen.


Set up

Lambda

About the lambda code, we will use the Python "Hello World" template. This example is enough for our need in this example.

import logging
import os
import boto3

logger = logging.getLogger()

logger.setLevel(logging.INFO)

def lambda_handler(event, context):
    logger.info(f'Hello World!')
    logger.info(f'## ENVIRONMENT VARIABLES: {os.environ}')
    logger.info(f'## EVENT: {event}')

    return {
        'statusCode': 200,
    }
Enter fullscreen mode Exit fullscreen mode

Its Terraform file is simple too, and is like every other terraform file to deploy a lambda

data "archive_file" "lambda_zip" {
    type = "zip"
    source_file = "${path.module}/src/hello_world.py"
    output_file_mode = "0666"
    output_path = "${path.module}/bin/hello_world.py.zip"
}

resource "aws_lambda_function" "processing_lambda" {
    filename = data.archive_file.lambda_zip.output_path
    function_name = "hello_world"
    handler = "hello_world.lambda_handler"
    source_code_hash = data.archive_file.lambda_zip.output_base64sha256
    role = aws_iam_role.processing_lambda_role.arn
    runtime = "python3.9"
    timeout = 900
    memory_size = 10240
    tags = {
         "service": "cron_lambda"
    }
}

resource "aws_iam_role" "processing_lambda_role" {
    name = hello_world_role
    path = "/service-role/"
    permissions_boundary = // To define correctly!
    assume_role_policy = data.aws_iam_policy_document.assume-role-policy_document.json

    inline_policy {
        name = "test_policy"
        policy = data.aws_iam_policy_document.test_policy_document.json
    }
}

data "aws_iam_policy_document" "assume-role-policy_document" {
    statement {
        actions = ["sts:AssumeRole"]

        principals {
            type = "Service"
            identifiers = ["lambda.amazonaws.com"]
        }
    }
}

data "aws_iam_policy_document" "test_policy_document" {
    statement {
        actions = [
            "logs:CreateLogGroup",
            "logs:CreateLogStream",
            "logs:PutLogEvents",
            "logs:AssociateKmsKey"
        ]

        resources = ["*"]
    }
}
Enter fullscreen mode Exit fullscreen mode

Warning : permissions_boundary field is not setup in this example, and it's the only thing you need to setup to make it work.

Now that we have everything we need to deploy our lambda, we can go to the next step : write the terraform file for the EventBridge stuff!

EventBridge

Its declaration is in 3 blocks :

  • aws_cloudwatch_event_rule : Which will define what are the conditions to trigger an event (here it's our scheduler)
  • aws_cloudwatch_event_target : Which will define what will be execute when the event is triggered
  • aws_lambda_permission : Which are some permissions to allow EventBridge to call a lambda function.

The Terraform file looks like this :

resource "aws_cloudwatch_event_rule" "scheduler" {
    name = "scheduler_lambda_cron"
    description = "Schedule for Lambda Function"
    schedule_expression = "cron(0/10 * ? * MON-FRI *)"
}

resource "aws_cloudwatch_event_target" "schedule_lambda" {
    rule = aws_cloudwatch_event_rule.scheduler.name
    target_id = "processing_lambda"
    arn = aws_lambda_function.processing_lambda.arn
}

resource "aws_lambda_permission" "allow_events_bridge_to_run_lambda" {
    statement_id = "AllowExecutionFromCloudWatch"
    action = "lambda:InvokeFunction"
    function_name = aws_lambda_function.processing_lambda.function_name
    principal = "events.amazonaws.com"
}
Enter fullscreen mode Exit fullscreen mode

In this part, there's only one thing to think about, and it's the schedule_expression field. It's here that you will define when your lambda will be executed.

In this example, the lambda function will be called every 10 minutes during weekdays.

If you want to have more informations about crons and how to update it, please check this documentation from AWS.


Execution

Now that you have everything, you can deploy all the stuff and enjoy more automated treatments!


I hope it will help you! 🍺


You want to support me?

Buy Me A Coffee

Top comments (0)