Can you afford to backup your site with Amazon S3?
Since we launched, a lot of people have approached us saying that they could replicate what we do with Automatic Website Backups through some scripts and Amazon’s S3 service.
There are a number of critical issues you could point at as to why you’d choose us over S3, but perhaps the most critical is proven uptime. Since we began internally monitoring our services a few months ago, we have seen no downtime of production servers.
We built our infrastructure from the ground up to be scalable to an infinite number of nodes in any data center anywhere in the world, and we stand by it vigorously. Similarly, we built our infrastructure around the goal of storing lots and lots of data (in the form of .tar.gzs) and distributing lots and lots of data (as new backups are uploaded and downloaded).
Beginning earlier this afternoon, Amazon S3 went down, and brought with it not only the data behind many startups (as S3 is very commonly used to store user uploaded images and videos for various web services), but the location and core system behind many people’s website and personal computer backups. Amazon S3 also went down back in mid-February.
Of course, the odds that during this exact time you’ll desparately need a backup taken or need to be able to access backups are fairly low, but we don’t take such an important necessity lightly.
What if you did need to download a backup, or if you were depending on the fact that a backup was supposed to be taken?
While some may see a price benefit to working with S3, for many things (backups included), downtime isn’t an option — regardless of who your provider is.
On a side note, we’re very careful about when we make changes to our infrastructure, regardless of the cause (optimizations, new features, etc.). We have an internal window from approximately 25 after the hour to 55 after the hour to make any changes necessary, so we don’t interrupt any ongoing backups.
