Attach SQL comments to correlate celery tasks with SQL statements.
This helps in easily correlating slow performance with async tasks and giving insights into backend database performance. In short it provides some observability into the state of your client-side applications and their impact on the database’s server-side.
You're using celery with django, and want to trace the origin of database queries.
Does it replace / overlap sqlcommenter
No. I've been using sqlcommenter in production and it lacked the ability to annotate queries that were run in non-http context. Meaning all queries that originated from async flows were untagged.
This package fills that gap.
In a simple celery setup, just passing the BaseTask when declaring a celery object will do the job.
# if not using a base class, pass directly where you declare celery config
from celery_sqlcommenter import BaseTask
app = Celery("my-awsome-application", task_cls=BaseTask)If you already have a custom BaseTask with specific functionality, replace the original celery.Task with celery_sqlcommenter.BaseTask
# old configuration
from celery import Task
class MyBaseTask(Task):
# custom methods you've implemented
def on_failure(self, exc, task_id, args, kwargs):
pass
# simply replace Task with BaseTask provided by the package
from celery_sqlcommenter import BaseTask
# notice Task changed to BaseTask
class MyBaseTask(BaseTask):
# custom methods you've implemented
def on_failure(self, exc, task_id, args, kwargs):
passInternally BaseTask inherits celery.Task so all functionality stays intact
Install the package from pypi
# vanilla
pip install celery-sqlcommenter
# poetry
poetry add celery-sqlcommenter
Please open an issue or drop a message on my socials, will be happy to help.
