So, wouldn’t it be cool if your organization didn’t just have a few centralized file servers and backup solutions? What if you could have racks of servers distributed throughout your entire organization that could act as redundant storage for your file servers and backup servers? And wouldn’t it be awesome if you could have all this without increasing your current hardware budget? Or even decrease it? Impossible you say? Well, maybe not.
You probably already have several distributed servers in your organization. But, instead of calling them servers, you probably call them desktops. But, what’s the difference between a server and a desktop? What’s the difference between a client and service? Can’t we start to blend the lines between these?
It’d be like implementing Grid computing into your organization. Or similar to RAID, which used to mean redundant array of inexpensive disk. Why not a redundant network of inexpensive storage servers? Most modern desktops have an under-utilized CPU, network connection, and hard drive. Why not make them useful?
It’d be like P2P in the organization, but it’d have lots of extra features like permissions, levels of redundancy, dynamic network topography mapping, distribution of data, search features, fault tolerance, version history, etc.
It’s like taking RAID, ZFS, a distributed file system, drive mirroring, backup software, data integrity, and a search engine and packaging it all into a desktop client that acts as a P2P server. It’d ideally be integrated into common directory services like LDAP, Active Directory, and Open Directory.
I doubt this is an original idea. In fact I seem to remember talking about this general concept (probably involving ZFS) with a coworker a couple months back.
What do you think? Could it work?
Does anything like this exist already?
What other features would be useful?