powerflex-monitoring 1.8.0

Tools to assist in monitoring a Python service.

PowerFlex Python Monitoring library (powerflex-monitoring)

Tools to assist in monitoring a Python service.

Installation

You can install from PyPi directly:

pip install powerflex-monitoring

Demo

See ./demo.py for a usage example. Also see the type hints and the docstrings.

To run the demo:

1. Install docker and docker-compose
2. Run make up
3. Run docker container ls to see the open ports and try hitting the monitoring URLs listed below

$ docker logs powerflex_monitoring_main
[INFO] Added route at port 8000: /monitoring/health/
[INFO] Added route at port 8000: /monitoring/ready/
[INFO] Added route at port 8000: /monitoring/metrics/
[INFO] 🚀 Starting monitoring server at port 8000
[INFO] Connected to NATS_ADDRESS nats:4222
[INFO] Configuration for NATS health check
[INFO] Service has been marked as ready. Cause: Service is initialized

Feature: monitoring server

from powerflex_monitoring.monitoring import MonitoringServer

• Readiness and healthiness endpoints are at /monitoring/ready/ and /monitoring/health/. These are useful in environments like Kubernetes: https://kubernetes.io/docs/tasks/configure-pod-container/configure-liveness-readiness-startup-probes/
• Prometheus metrics are at /monitoring/metrics/ . Use the prometheus_client library to expose more metrics.

Feature: NATS health checker

from powerflex_monitoring.nats_health_check import NatsHealthChecker

This class checks if NATS is healthy and marks the service as unhealthy if NATS is ever down for a certain period of time.

Feature: Redis health checker

This feature checks if Redis service is healthy and marks the service as unhealthy if Redis is ever down for a certain period of time. How to use it?

1. Import the RedisHealthChecker

from powerflex_monitoring.redis_health_check import RedisHealthChecker

2. Instantiate a RedisHealtChecher object with the Redis connection that you want to ckeck and the monitoring server to which you will inform the health status

redis_health_checker = RedisHealthChecker(redis_conn, monitoring_server)

3. Run the async function periodically_check_redis_health like this:

await redis_health_checker.periodically_check_redis_health()

Feature: wait for service to be ready

This is useful when starting your integration tests. If you use the readiness check, then you can delay an action until your service is ready.

python -m powerflex_monitoring.wait_for_service_ready \
  --ready-check-url http://localhost:8000/monitoring/ready/

Make sure this library is installed in your virtualenv or globally, otherwise you won't be able to run this script.

See our Makefile for an example of using this script to wait before running integration tests.

Test results

This library is unit tested with 100% coverage in Python versions 3.8 to 3.11 and in pypy 3.9.

Developing

1. Install the dependencies with make setup. This will use pipenv to manage a virtualenv.
2. Run the commands in the Makefile to test your code. Run pipenv shell before running the commands or run pipenv run make ... to run the Makefile commands with pipenv.

• make commitready
• make test-unit or make test-unit-all-python-versions
• make format-fix
• make lint
• make type-check-strict

1. Please keep the unit test coverage at 100%.

Releasing to PyPi.org

1. Make sure all code checks have passed with make commitready.
2. Make sure you commit all code you wish to release with git commit.
3. Set the version in ./src/powerflex_monitoring/VERSION Please attempt to follow semantic versioning.
4. Run make bump-version to commit the change to the VERSION file.
5. Run make release to upload the package to pypi.org and to push a new git tag