Flyte enables production-grade orchestration for machine learning workflows and data processing created to accelerate local workflows to production.

Flyte

Hey all, I'm looking into  caching of outputs right now. I was wondering if we could also set a TTL/expiry date.
This is not so that flyte triggers some delete function in the blob, but rathers is aware that the cache can expire. (because of data lifecycles on the storage side).
Or will flyte automatically detect if a cached file does not exist and rerun the task instead?

Hey <@U06MPSVVAQL> , you can't set ttl per item but you can set it in a propeller config for all items. For the exact reason you stated

<@UNW4VP36V> Is it the `max-ttl-hours` setting? Because that seems to do something else: `Maximum number of hours a completed workflow should be retained. Number between 1-23 hours`
Unless checking the cache is the part of the `retaining the completed workflow`&gt;

<@U06MPSVVAQL> it should be <https://github.com/flyteorg/flyte/blob/master/flytepropeller/pkg/controller/nodes/catalog/config.go#L50-L50|this max-cache-age config option>.

Thanks <@U029U35LRDJ> I see that <https://docs.flyte.org/en/latest/deployment/configuration/generated/scheduler_config.html#max-cache-age-config-duration|scheduler> and <https://docs.flyte.org/en/latest/deployment/configuration/generated/flyteadmin_config.html#max-cache-age-config-duration|admin> also expose a `max-cache-age` config option Should I set the same value in for all of these or how they different?