PresSTORE is a client-server backup and archive package. The server component has been available on Unix and Mac platforms for a while now, but although Windows clients have long been supported, there’s now a Windows version of the server to add to the collection.

Although we can support most platforms, it seemed sensible to run our test version up on a Windows Server 2003 machine in order to see the new bit. The installer is a simple wizard, which asks you whether you want a server or client installation; we chose “Server” for the WS2003 machine and then ran up the client on a Windows XP desktop. If you’re a Linux user, SuSE 9.1+ and Red Hat Enterprise 3+ are supported on both the client and the server; for MacOS you have the choice of Mac OS 10.3+ (PowerPC) or 10.4+ (Intel); and for Solaris you’ll need at least version 8 (SPARC) or 10 (x86).

When the server installer has finished, the only admin section you have available is the licence entry screen. Once you have a valid licence, the other options are made available and you can administer the system. By default the management GUI listens for browser connections on port 8000 (you can change this during installation if you wish), so you just point the browser at http://localhost:8000/ and log in using the appropriate user ID (the package uses the OS’s user database, so in our case we logged in with the WS2003 “administrator” credentials.

To add clients to the system, you simply select "New Client" and enter the description, IP address and communications port for the client you want to add. If you have a lot of similar remote devices to add, you can make duplicates and then simply change the bits that need to be altered (eg. the last octet of the IP address) in the copies. The GUI doesn’t really follow any recognisable standard (that is, it’ll look unfamiliar to someone who’s only been immersed in Windows) but it’s easy enough to use and although we didn’t really like it for a start, it grew on us once we got used to it.

Once you’ve set up the client devices you want the system to work with, you’ll also need to define the backup devices you have available. The system supports a pretty extravagant range of tape devices and arrays (there’s specific support for devices from the likes of Storagetek, ADIC, HP, Overland and a pile of others) but if you wish, you can choose to back up to a disk volume (or, for that matter, a “virtual jukebox” built from a collection of disk volumes) instead. The system saves data to “pools” of storage devices, which means that when you start to get close to capacity in a “pool”, you can simply add more devices to the pool to expand the capacity.

There are three basic types of functionality to the system: backup, archive and file synchronisation. The three are licensed separately, but it’s all in one application – the licence keys simply turn the various features on – and you can buy them either separately or as a bundle (called the “Professional” package).

A "backup" is your usual concept of a scheduled copy of data from client devices to storage media. So you tell it which clients to back up, then give it a schedule of when to perform its full and incremental backups. Interestingly, you can configure the system to do its full backups to a different media pool from its incrementals (there’s also the concept of a “synthetic” backup, which is used to build a “full” backup from a previous fileset without querying the client devices themselves). Along with the day-by-day schedule, the backup configuration lets you additionally define when to do monthly full backups. Interestingly, you can restrict the number of hours a backup is allowed to run (so it doesn’t trample over into the working day), and backup plans can be chained together so that instead of having to guess the start times of the various backup jobs, you can say: “Start the Accounts backup as soon as the R&D one has finished”.

An "archive" is a bit like a backup, but is much more a one-off activity. Instead of writing data to a series of media objects on a schedule, in such a way that you can restore them after a crash or inadvertent deletion, you’d use an archive to say: “OK, I want to farm off this collection of data from these clients to this piece of media”. Archives aren’t all that dissimilar to backups, in that you define clients to read from and media to write to, but in this case you drop the sources you want to archive into a "basket" and then tell it to start the archive process either at a given time or via the “Start now” button.

The final approach you can take is file synchronisation. As the name suggests, you’ll define a series of source and destination directories (each which can, of course, be on any device on the network accessible to PresStore) and tell it to make copies of files from each source to each destination on a schedule, and with a file type filter if you wish. By default it’ll just overwrite the replica if the source file has changed, but you can easily configure it to keep a specified number of previous versions should it detect that a file has changed (the old copies are stored in a folder called "__VERSIONS" by default).

When we first ran up PresStore, we didn’t really like it very much. It wasn’t immediately intuitive to use, the manual wasn’t all that brilliant, and the GUI clearly hasn’t had a great deal of attention from a GUI design specialist. (Let’s face it, the icons used to represent filesystems look like they’ve been lifted direct from AppleShare, and it’s not overly encouraging when you read options like "Setup und manage storage devices" that didn’t quite make it through the German-to-English translation process).

The more we used the package, though, the more we thought: "Hey, this is actually rather good". The GUI, though non-intuitive at first, actually turned out to be quite usable (the schedule definition screens are quite cleverly done, for instance), it’s far easier to connect clients than has been the case with many client-server packages we’ve reviewed over the years, and you can’t knock the range of backup hardware support either. Yes, some of the words have been a bit mangled in translation. No, it’s not the flashiest package to look at. And no, you won’t find an add-on that does open file support for your wild-and-wacky database package. But if your wish-list says: "Quick setup, multi-platform, supports my auto-loader and is unlikely to break ‘cos it’s German", it does all that without breaking into a sweat.


Although many of us would "automatically" go out and buy one of the well-known products from the likes of Symantec or EMC, PresStore is well worth a look as well, particularly if you have a combination of Unix/Linux/Mac OS/Windows in your organisation.