Work out lifecycle policy savings instantly with clear inputs, formula shown and shareable results.
Lifecycle policies work because most data is written once and read rarely afterwards, so the steady-state distribution across tiers depends on growth rate and transition ages. With modest growth, the overwhelming majority of a bucket ends up in the coldest tier, which is where the saving comes from — often 70 percent or more of the standard-storage bill.
Tier distribution
hot TB ~ monthly new data x transition days / 30.44; remainder distributes to IA and archive by their transition ages; cost = sum of TB x per-tier rate
Yes, per object. For buckets with hundreds of millions of small objects, transition request charges can exceed the storage saving, which argues for filtering by object size.
Thirty days to infrequent access is the usual minimum, and 90 to 180 days to archive. Match them to real access patterns from access logs rather than guessing.