Backup Planning

How to Estimate Backup Drive Space From a Week of Changes

Estimate room for the first backup and for retained changes, then compare the estimate with a week of actual backup growth.

A 250 GiB first backup tells you where the drive starts. Its future size depends on how quickly the protected files change and how many old versions you want to keep. Use a week of ordinary work to estimate both sides of that decision.

Measure in one unit throughout the estimate. This example uses GiB. A drive advertised in decimal GB will show a lower number if your measuring tool displays GiB. Use the free capacity your destination actually reports after formatting.

Observe a representative week

Record initial protected data size and each day's destination space used after the backup finishes. Use a week that includes the kind of work you actually do. Mark unusual events such as a virtual machine update or an imported video library. Record source-side changes separately if the backup disk starts pruning during the week. The destination's net growth after pruning can conceal substantial new writes.

Input in a small example Value
Initial protected data 250 GiB
Working range for additional retained data per day 3 to 9 GiB
Intended planning window 28 days
Base estimate 334 to 502 GiB
With an illustrative 25 percent spare margin About 420 to 630 GiB

The base calculation is initial size + daily change × retained days. It is a budget range, and the sample does not predict actual snapshot growth. Compression, file sharing across snapshots, retention and exclusions can change the total. The 25 percent margin gives the sample budget room for a larger source or a busier month. Adjust it if your recorded week already includes the largest imports you expect.

Treat large rewrites separately

Write down examples such as a 75 GiB disk image that changes repeatedly. Measure how much the destination grows after one routine change to that file. Extrapolate from the observed destination change and how many of those versions you intend to keep. Apple identifies large, frequently updated files as a reason backups may take longer, even though Time Machine normally backs up changes after the first run.

Apple's Time Machine setup guidance recommends a backup device with at least twice the Mac's storage capacity and explains its hourly, daily and weekly schedule. It removes the oldest backups when the drive fills. Your estimate helps you decide whether that starting point is enough for the history you want. In tools such as restic, the set of retained snapshots and the operation that reclaims unreferenced storage are separate concerns.

Separate the first copy from the history you keep

Before measuring, list the volumes and folders you expect the backup to include. Note the files you know change frequently, such as a data library or a virtual disk image. Record how much destination space is used after the first successful run. If the destination already contains older snapshots or other directories, account for that separately. A number read from the source may not match what the backup keeps after its own compression or exclusions. Use the observed destination figure for the plan where you can.

Measure again after a normal day and a day with more changes. A project that imports hundreds of new photos can have a higher burst than an ordinary week of documents. Write the imports and large rewrites into the observation log with the date and retention rules. If the backup rotated old files while you were testing, compare the oldest and newest recoverable dates as well. A nearly flat usage line could represent pruning at about the same rate as new data is written.

In the example range, 250 GiB plus 3 GiB times 28 days is 334 GiB. At nine GiB every day, it reaches 502 GiB. Multiplying both ends by 1.25 gives about 418 and 628 GiB before rounding. Use the low and high values as planning cases. If you need a minimum of four weeks on heavier days, compare the larger scenario with the destination's measured usable space and what the backup tool will do when it fills. A label on a drive box does not tell you the actual number in your tool's units.

Give one costly file its own scenario

Suppose one virtual disk changes four times a week. Measure the separate growth after each change if you can. If the backup grows by a few GiB when the source rewrites a much larger file, its behavior is better described by that observed growth than by the file's full size. If the destination grows by nearly the entire source file each time, retaining every version can require much more space. Neither result transfers automatically to another backup tool or configuration.

Choose an appropriate level of restore history for that file based on the work you need to recover. If you decide that one category gets a different schedule, make sure there is still a supported and tested recovery path for it. Avoid excluding valuable work just to make the capacity graph look stable. For an external working volume, use the coverage test to check the intended backup before building a capacity estimate around it.

Reconcile the estimate with the oldest restore point

After operating the backup through the planning window, open the history and find a sample as old as the period you intended. Run a disposable file restore for one important folder. If only newer versions remain, compare the actual retention, free space and the pattern of changed files. If the oldest date is available but the sample is absent, check whether that source was in scope on that day.

Extend the measurement after a change in work, retention or hardware. Save the resulting range with a note about which tool, volumes and period you measured. This helps you decide whether a new destination is needed or whether one large category needs a different approved plan. Keep application procedures and example arithmetic distinct so the next person knows which values are forecasts and which are observable results.

Allow for the weeks you do not measure. If your work creates a large archive every quarter, add that event to a separate estimate and check where it will be backed up. A four-week trial that missed the archive cannot tell you how the drive will handle its arrival. Use the first real post-archive backup to update the estimate.

Sources