The FileSaver is a new device from Adaptec which is a cross between a backup server and a NAS box. It's a 1U server with up to four internal hard disks (in a RAID5 array) which uses standard Windows filesharing protocols, but which is actually a backup server for the client computers on the network. An agent installed on each client PC automatically backs files up to the FileSaver's internal disks and users can recover files at will, without operator intervention.

The unit is based on Red Hat Linux, although you're not exposed to this at all (you don't even get a login prompt on the console). You plug it in and fire it up and once it's started you launch the manager application. This scans the network for the FileSaver device. When it finds one it launches a Web browser and points it at the FileSaver's HTML GUI.

The first thing you'll do is set the IP address of the unit to a sensible value. Once you've done this, you can restart the GUI and set up the various management parameters – the usual kind of stuff such as the IP address of an SMTP server and the username to send email alerts and warnings to. You can also view the disk usage by the various users on the system, and delete user entries, although user creation is left to the client installer. The system uses a concept Adaptec calls RAIN – Redundant Array of Independent Nodes – which is basically a failover system; multiple devices can be configured to share the data storage load and thus provide resilience in the event of a device failure. RAIN setup is pretty straightforward because, as with the initial setup, the GUI scans the network and picks up other devices automatically.

Because the backup operation is automated, the FileSaver needs a client program to be installed on each user's PC. This is initiated from the Web GUI – just click the Client Software link and follow the prompts. Part of the client installation process is to provide a user ID and password, which are used to create an appropriate user account on the FileSaver. One user can have multiple data repositories, representing multiple machines. Once the installer has finished and the client PC has rebooted, the client software performs an initial scan and full backup of the PC. This took a couple of hours on our test machine, which was a 1.4GHz Pentium Mobile laptop with about 15.5Gbytes of data; although we could tell it was running, it didn't slow the PC down unduly when we were doing other work while the backup ran. Once the initial backup has been done, the client writes changes to the unit throughout the day, instead of doing the entire backup on a daily or weekly schedule.

Restoration of files is done through the client application, and it's a straightforward step-by step approach: you choose the files and folders to restore, you choose where to put them (generally you'd place them in a temporary folder instead of just writing them back where they came from) and then you let the unit find the files and recover them. If you have a lot of files on a workstation (ours had about 100,000) it can take some time to build the tree for you to select from, but the speed of the restoration process itself is OK. The client software also allows customisation of what the system should back up and what it should leave out, as well as letting you define how many versions of a file to keep before throwing the oldest away.

Because our review unit was running a pre-release version of the operating software, there were a couple of small issues (a missing help file, for instance) but this is the kind of stuff that will almost certainly be fixed by the time the product launches to the public.


This type of device doesn't replace the traditional tape-based server backup; it does complement it nicely by providing data security for users' desktop PCs, though.