If you don’t care about backups, you can stop reading here. And good luck with that approach!
I have, along with many others I’m sure, long been wondering what to do about backups. Naturally I want a secure backup with as little hassle as possible.
Having photography as a hobby means copious gigabytes of data to back up, a task that is really laborious to do by burning DVDs. Also, when you get to the point of having dozens of backup-DVDs they take up surprisingly much space. So a second hard disk is called for, but they have a couple of problems: first, they fill up and you need to buy another one. Second, even external hard disks are typically in the same physical location as the main computer, thus on the line of fire should there be, well, a fire or some other localized disaster.
So what to do? I’ve long been pondering doing network-based backups. Most services offered for that purpose suffer from ridiculously small quotas or absurdly high prices or both. Like Welho, offering a whopping 5GB for 5eur/month. Saunalahti at least improves tenfold upon that offer, but 50GB is still far too little.
To get around that, I’ve been using my US virtual host that has a generous – unlimited is pretty generous, don’t you think? – amount of space for my backup needs. The trouble with that is that being cheap, it’s not the worlds most reliable virtual host. Honestly though I haven’t had any problems with them, but the lack of 100% trust is reason enough to not rely solely on them.
So I ended up in the unenviable position of having backups on two separate hard disks and on DVDs and on the unreliable server. However, my latest trip produced about 30GB worth of stuff to back up in under two weeks and that pushed me over the edge. I’m quitting the DVD-burning business and the second hard disk (actually a fourth but who’s counting..) that is now bursting at its seams can partly go, too. I’ll still keep the unreliable US-host as a backup backup.
Instead, my primary backup is now Amazon S3. While the service does cost money and has even been known to experience data loss in rare cases (not a pretty prospect), I consider it a good deal. Having my important data now stored in three different countries on two continents, with one service actually having an SLA, makes it highly unlikely that they will all be destroyed within the same day. Right?
So getting to the actual topic, how do we get to the 200 GB per month? Well, this is where JungleDisk comes in – a great little piece of software built on top of Amazon S3 that does automagic backups to it and lots of other cool stuff. So while I don’t yet produce 200GB worth of data each month, the initial backup process does consume quite a bit of bandwidth – and for that, the 200GB this month is just the beginning.
This is also where we get to the annoyingly slow uplink speeds. I know I’m exceptionally lucky in that regards with an uplink speed of 10Mbps – though real throughput to Amazon S3 servers over HTTPS seems to be around 4.5Mbps. That’s pretty fast but the initial process still takes many days. I can only imagine how frustrating (or rather impossible) this would be over standard DSL uplink speeds of 256Kbps, 512Kbps or even at 1Mbps. In this case, additional bandwidth does enable new services.
Business-wise, I am fully aware that at this usage rate, I am probably generating losses for my broadband operator. Fortunately that’s a SEP 
Our data storage need seems to be increasing with the increasing convenience in storing them. Good you made us aware of the Amazon thing, or else I had to buy a new hard disk