Fifteen years of finished work on this morning's storage
It is the filing cabinet in the corner, sitting in the best office in the building. Nobody put it there on purpose. The project simply finished and the folder went quiet.
Almost every practice we look after keeps completed work on exactly the same storage as the model somebody has open right now: the fastest tier, the one with the shortest backup window, the one replicated offsite every night. Ten or fifteen years of it, often three quarters of everything the practice owns.
It happens because there is no moment when a project stops. There is a final certificate, an invoice, a party and the folder just stops being touched. Nothing in the process says "move this".
What it actually costs
- Storage you are paying premium rates for. Whether that is a SAN, a NAS or cloud, the fast tier costs several times the slow one and the archive does not need the speed
- Backup windows that keep growing. Most backup runs re-scan everything. Data that has not changed since 2019 is scanned nightly, forever
- Longer restores when it matters. After an incident you are restoring the archive alongside the live projects, so the thing the studio is waiting for arrives later than it needed to
- Search that returns the wrong decade. Everybody has opened a 2017 detail thinking it was current. Once is amusing
You cannot simply delete it
This is the part that stops most practices, reasonably. Completed work has to be kept: your appointment documents, your professional indemnity cover and any statutory duties on the building set the period and none of that is an IT decision. Building safety obligations have made the case for keeping more, for longer, not less.
The answer is not deletion. It is putting the archive somewhere that suits what it is: retrievable rather than instant, cheap rather than fast and still there in year twelve when somebody rings about a roof.
A morning's work, roughly
- Make "closed" an event with a date. One field, one person who sets it. Without this nothing else is repeatable
- Agree what moves. Usually everything from projects closed more than twelve or twenty-four months ago, kept whole rather than filleted
- Move it to a slower tier. Cloud archive storage, a separate volume, or a second NAS. Minutes to retrieve is fine. Days is not
- Keep an index that survives the move. A simple list of project, client, dates and where it now lives, held somewhere other than the archive itself
- Restore one project from it. Before you trust it and again once a year. An archive nobody has ever read back is a hope, not a backup
What the live system does next
The live server notices immediately. Backup windows shorten, often dramatically, because the nightly job stops carrying a decade of dead weight. Restores get faster for the same reason. Searches return this year's work. Your next storage purchase gets sized for what the practice is doing rather than for everything it has ever done.
None of this is clever. It is just a decision nobody has been asked to make, because the cost arrives quietly, spread across a backup window and a renewal quote, rather than as a problem with a name.
How much of your primary storage is work you finished before 2022? If nobody knows, that is worth an afternoon on its own. We handle cloud migration for AEC firms, backup and continuity as part of managed IT support in London and the everyday IT support architecture practices run on. If the archive has never been moved, let us take a look at it.
