The repo demonstrates how to manage monitors in DataDog using CloudFormation. As of 2021-03-21 only simple monitors alerting for failed lambda invocations can be created.
All infrastructure here is deployed using the SAM cli.
- Define the SAM config file:
version=0.1
[default.deploy.parameters]
profile = "personal"
stack_name = "datadog-monitor-dev"
s3_bucket = "custom-resources-dev-20200318"
s3_prefix = "datadog"
region = "eu-west-1"
capabilities = "CAPABILITY_IAM"
confirm_changeset = true
tags = "project=\"datadog-monitor-dev\" stage=\"dev\""
-
Ensure that an integration exists between your AWS account and DataDog account. Detailed instructions are available here.
-
The custom resource requres programmatic access to DataDog. As such you must save your DataDog API Key and App Key in AWS Secrets Manager as follow:
{
"DD_CLIENT_API_KEY": "API KEY",
"DD_CLIENT_APP_KEY": "APP KEY"
}
The DataDog module contains a Monitors class which uses the DataDog API to create, update or delete simple lambda monitors. The create_monitor method creates a DataDog method which alerts for any failed lambda invocations within the last hour.
from datadog import Monitor
LambdaFunctionName = "test-lambda"
DD_monitor = Monitor(api_key, app_key)
monitor_id = DD_monitor.create_monitor(
name="Test Monitor",
message=f"{LambdaFunctionName} has failed!\n@recipient@email.com",
priority=1,
functionname=LambdaFunctionName,
tags=["env:dev"]
) Harry Daniels