flok writes: "Deduplicating software/appliances are the bomb currently. Everyone wants them, everyone is developing such a thing. Software finds duplicates using hashes, and using advanced algorithms those hashes are looked up from huge memory-caches. But wait, looking up data, isn't that the task of a database? This made me think and after a couple of hours I implemented MyDeDub, a deduplicating network block device which stores its data (and hashes) in a MySQL database. As it looks now, filesystems on MyDeDub stay consistent, massive space is saved and it even performs, well, somewhat. It is implemented in Java so even an old windows 98 system might suffice. You can find the software here."