flask-datadog-monitor 0.0.7

Tool for auto generating DataDog Monitors from flask endpoints

Overview

Utility for auto generating DataDog terraform monitor specifications from flask endpoints.

1. Setup DataDog Terraform integration
2. Decorate flask endpoints you want to monitor
3. Run the flask-datadog cmd line util to generate terraform files for the monitors
4. Apply Terraform changes

Usage Examples

Decorate Flask endpoints you want to monitor

# Decorating with `datadog_monitors` will generate default monitors for each
# method on the route
@datadog_monitors()
@flask_app.route('/foo', methods=['GET', 'POST'])
def handle_foo():
    return ''

# Specifying a particular monitor will allow you to tweak thresholds (and
# other parameters) but will only generate that particular monitor for
# that route
@datadog_monitors(
    monitors={
        MonitorType.APM_ERROR_RATE_THRESHOLD: {
            MonitorSpec.CRITICAL_THRESHOLD: 0.8,
            MonitorSpec.ALERT_PERIOD: '10m',
        },
    },
)
@flask_app.route('/foo', methods=['GET', 'POST'])
def handle_foo():
    return ''

Generate Terraform for monitors with cmd line utility

flask-datadog gen-terraform --env prod --service my_service path/to/flask_app:app output_dir

Running This Project Locally

Info on running the test service locally against a DataDog test account.

Assumptions

• You have a DataDog test account created

Steps to standup local test env

1. Set DATADOG_API_KEY in env vars
2. Set DATADOG_APP_KEY in env vars
3. make sync-datadog: this applies monitoring configs to datadog acnt
4. make local-up: this builds app and runs them locally with the dd-agent containers

Once the local env is up, an automatic script in the fake-client service should ping the test-app service periodically to generate fake traffic.

The test-app service and fake-client service both have mount points so you can edit the code and have changes reflected directly without rebuilding containers.

Unit / Integration Testing

1. Ensure venv is created. From root dir: python -m venv venv
2. Install testing dependencies: ./venv/bin/pip install .[testing] (setup.cfg has testing extras_require)
3. make test

Reference

Flask API interface

• Flask App Object API: https://tedboy.github.io/flask/interface_api.application_object.html

TODO

• Allow specification of just the monitors you want to default to
• "trigger when metric is <above/below> the threshold <in total | on average | etc.> ...
• test with different versions of Flask
• make terraform file generator into a cmd line executable