Django. Split sending errors to a separate sentry project (2 DSN)

[ERAalex]

Django project, currently using the standard error sending to a single DSN:

def init_sentry():  
    sentry_sdk.init(  
        dsn='DSN for MAIN project',  
        integrations=[DjangoIntegration(), PyMongoIntegration()],  
        ignore_errors=[DoesNotExist]  
  )

it is necessary to allocate CELERY errors (only for CeleryIntegration ) to a separate sentry project (as I understand use a separate DSN)

is it possible to create a separate sentry_sdk If sentry_sdk is already running? Something like addition:

celery_sentry_client = sentry_sdk.Client(  
    dsn='DSN for SECOND project',  
    integrations=[  
        CeleryIntegration(),  
    ],  
  
)

I searched through various forums, but only outdated information, thanks

[antonpirker]

As your Django application server and the Celery workers are different processes this is possible.

You can load your DSN from the environment and when you start your Gunicorn (or Uvicorn, Daphne or whatever server you run you Django app in) you can start it with SENTRY_DSN=dsn1 gunicorn ... and when you start your celery workers you do it like this SENTRY_DSN=dsn2 celery -A app worker ....

Then everything that fails in Django will be sent to dsn1 and everything that fails in your Celery workers will be sent to dsn2.