0

We're running a flask application and we do all our heavy processing with celery. We use a redis instance from amazon to be our message broker. We just had a fail, causing much pain and bleeding, so we're looking into fail-over strategies.

The first project that appeared to us was Celery sentinel. https://github.com/dealertrack/celery-redis-sentinel Would this be something that would give us a fail-over capability?

We've been doing some tests, and it seems not to be working as anticipated.

Benjamin Rudder
  • 1

1 Answers1

-1

In your case maybe moving the celery backend to RabbitMQ would be better, as RabbitMQ is a lot more persistent with its data

Trondh
  • 3,221
  • 1
  • 25
  • 34