Rule evaluation caching

[markphelps]

Describe the solution you'd like

Would like to support evaluation caching for rule based evaluation. IE https://flipt.dev/concepts/#evaluation

This would behave similarly to what was added in #6, but would likely be more complex since there are multiple entities that need to be cached (segments, distributions, rules, variants, etc) and invalidated.

One approach is to just cache the evaluation result and not the actual individual actors (noted above), but would have to come up with a scheme to invalidate this cached result if any of the above changed.

[stale]

This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. Thank you for your contributions.

[badboyd]

Can we open it again?

[badboyd]

Can we just cache the rules and distributions, and do the evaluation? Then let the cache be purged after an expiration time.

[markphelps]

Yep! Ive actually started working on this this week

[markphelps]

Tracking in https://github.com/markphelps/flipt/tree/cache-redo, I hope to have this land early this coming week

[markphelps]

@badboyd this should be fixed as of v0.12

The docs on the caching feature are here: https://flipt.io/docs/configuration/#caching

Lemme know how it works for you!