A budget review asks whether the primary backup target is two years or five years from full. The engineer quotes today's percent-used and the raw capacity numbers from this morning's dashboard. What monitoring output actually answers a lifecycle question like this?

A refresh question is a question about time, and only a time series can answer it. The useful artifact derives growth from months of physical usage history, expresses it against usable capacity - raw less raid, system overhead, and reduction assumptions - and marks the projected full month so procurement and installation lead times can be compared against it. The same trend serves the capacity threshold alerts and the budget packet, which is why it belongs in the standing monitoring configuration rather than in a one-off analysis. Today's snapshot cannot distinguish two systems growing at wildly different rates that happen to share a percentage: 61 percent used means two years to one estate and fifteen to the other, and a documented projection method is the discipline a budget needs, not speculation. The dedup ratio alone is an efficiency statement, not a forecast - it can collapse next quarter when someone onboards image-based backups, so multiplying today's ratio by future data assumes the future resembles the past in precisely the variable most likely to move. Per-directory accounting answers who, not when: summing details is chargeback, and a sum has no time axis, so it cannot produce a date. Exam caveat: subtract replication effects and rebuild the projection whenever a new workload changes the growth regime. Operational check: produce the one-page trend with the projected exhaustion month marked, and compare that month against the budget cycle's ordering lead time.