Your files on site, and a copy somewhere else.
We install and manage Synology NAS units for offices, professional firms and healthcare providers in Rome: shared folders for teams, snapshots, backups following the 3-2-1 rule with an off-site copy, monitoring and restore tests.
A backup nobody has tested isn't a backup.
An external disk plugged into a colleague's PC, a copy made by hand whenever someone remembers, a NAS reachable from the internet with the admin user. As long as all goes well, nobody notices. Then a disk fails, ransomware encrypts the shared folders or someone deletes the wrong folder, and it turns out the last good copy is months old, or was encrypted along with everything else.
From sizing to a restore test.
- 01
Assessment and sizing
How much data you have, how fast it grows, who uses it and how long it must be kept. Some healthcare data, for instance, has to be kept for 10 years: space and copies are designed from there.
- 02
Tested disks and RAID
We choose the Synology model and RAID level based on capacity and fault tolerance. Every disk passes a SMART test before it joins the volume: we have had a brand-new disk, at zero power-on hours, fail it. It was kept out of production and went back to the supplier under RMA with the test report.
- 03
Folders, users and access
Shared folders for teams and departments, permissions per user and group, connection to company accounts if there is a domain. Everyone sees what they need, not everything. Off-site access goes through a VPN or Synology's services, without exposing the NAS to the internet.
- 04
Snapshots and off-site copy
Scheduled Btrfs snapshots, to go back to an earlier version of your files after a deletion or a ransomware attack, and an encrypted off-site copy with Hyper Backup to AWS S3 or Synology C2. That's the 3-2-1 rule: three copies, on two media, one outside the building.
- 05
Monitoring and restore tests
Email alerts for disks, space, failed backups and unusual logins. And regular restore tests, of a file, a folder or a whole volume, to know the copy can really be read.
Copies you can restore.
The configured NAS, plus documentation of how the copies are made, where they are and how to recover them.
Where losing data isn't an option.
Healthcare providers and clinics
One NAS per site, off-site copies and long retention for the data that needs it.
Professional firms
Accountants, lawyers, engineers: archives of files and projects shared between colleagues.
Offices and SMEs
Department folders instead of files scattered across PCs, USB sticks and email attachments.
Companies with several sites
The same configuration at every site and one place to check the copies.
Design, video and architecture studios
Large files opened over the local network, with earlier versions close at hand.
The NAS is the easy part. The copies are what count.
We look after the NAS and backups at every site of a healthcare group, with part of the data to be kept for 10 years. The weak point is rarely the NAS itself: it's unverified disks, permissions that are too broad, copies nobody has tested. That's why we test disks before using them, always keep an off-site copy and check that the copies can be read. It's the same method we have applied since 2011 to the more than 110 servers we manage.
Before you save everything.
Isn't RAID already a backup?
No. RAID protects against a disk failing, but if a file is deleted or encrypted by ransomware, RAID copies the damage across every disk. You need snapshots and at least one copy on a different medium, outside the building.
Why a copy in the cloud?
Because a fire, theft or flood hits the NAS and the backup disks at the same time if they're in the same room. An encrypted copy on AWS S3 or Synology C2 stays available even when your site isn't.
Can we access the NAS from home?
Yes, through a VPN or Synology's remote access services, without opening any NAS ports to the internet. An exposed NAS is one of ransomware's favourite targets.
How do you know the backups work?
The NAS emails us if a copy fails, if a disk shows signs of failure or if space is running out. And at regular intervals we run a real restore test, not just a check that the job shows as completed.