Hello I'm trying to set up retries for some assets...
# ask-community
Hello I'm trying to set up retries for some assets. I currently have it set to retry after 30 minutes by raising a RetryRequested from within the asset definition. I'm also using concurrency limits to limit the number of concurrent runs in an attempt to limit resources. The issue I'm running into is if I have multiple jobs that need to be retried, they'll take up all of the concurrency slots even though they're just waiting. Is there any way to schedule so that it can be executed after 30 minutes without taking up one of the concurrency slots?
Hi Shriddy, Not an expert in this part of our API so I may need to reach out to someone else, but are you aware of the
? https://docs.dagster.io/deployment/run-coordinator
hey! thanks for the reply. yeah i've seen the queued run coordinator, which is what i'm using. the issue is that when the run is retried it does not add a new run to the queue but instead is retried within the current run. i think it would be better to re-queue the run instead, and i'm not sure how to accomplish that