Currently our cache controls all or nothing, TTL settings set across the board for each of the separate cache 'buckets': experiments per context, ff per context, and segment lists.
We would like to be able to extend the TTL for these separately, so that we can potentially keep experiments and segments alive longer while keeping feature-flag cache life pretty short and responsive.
This would create 3 different env vars to configure these settings per type, and retire the old setting.
Currently our cache controls all or nothing, TTL settings set across the board for each of the separate cache 'buckets': experiments per context, ff per context, and segment lists.
We would like to be able to extend the TTL for these separately, so that we can potentially keep experiments and segments alive longer while keeping feature-flag cache life pretty short and responsive.
This would create 3 different env vars to configure these settings per type, and retire the old setting.