Zanie Blue
Zanie Blue
@blakete first switch to using a hosted server (i.e. `prefect server start` or Prefect Cloud) and if you're already doing that switch to Postgres as a backing database.
@blakete for the Prefect Server (previously known as Orion) database, yes.
@cleung1996 I don't work at Prefect anymore, but I don't think there's automated migration. cc @serinamarie
@secrettoad I'm not sure I follow why task caching would not solve your problem?
@secrettoad Thanks for the additional details! Could you use idempotency keys to avoid creating the extra flow run in the first place?
This example now runs without error. #6689 can be used to track a nicer interface for spawning concurrent subflow runs.
Addressed by #8887 and #8702
This appears to be resolved in the latest release — it returns immediately :)
In the short term, we could try to patch ``` flow_run_context.task_runner = await stack.enter_async_context( flow.task_runner.start() ) ``` to create a copy of the task runner before starting it?
I don't want to introduce a new concept (i.e. a builder/factory type) to users in order to resolve this issue. We can add a new concept behind the scenes (i.e....