Amazon S3 Outage: A Gentle Reminder
February 18th, 2008On the morning of February 15, 2008, Amazon’s S3 service suffered an outage. It was resolved in a matter of hours, but affected a notable number of services and businesses.
While it was the first such outage for Amazon S3, it should serve as a gentle reminder that no service is infallible, even when operated by a giant such as Amazon. When choosing a data storage provider, a savvy consumer will keep in mind that any service offering “100% uptime” is full of malarkey. There is no such thing a 100% uptime. You may never know that any outage occurred with your service, but that doesn’t mean it never happened. With this in mind, it is prudent to be selective about the data you choose to store with your selected provider, and to practice redundancy. At the very least, redundancy equals hard-copy back-ups.
Following up on the subject of back-ups, there are a number of data backup services that utilize Amazon S3 for their storage. Other storage providers are utilized as well, but Amazon S3 is probably the most widely known. This, if anything, should serve as a good example of why you should not rely solely on remote back-ups for your data - especially if data loss may have serious ramifications. That’s not to say that remote back-up isn’t useful, or that it should be avoided. It is a useful service, and makes a great deal of sense for a variety of businesses. However, it should not be your sole form of back-up. Again, redundancy and hard-copy back-ups are key for true peace of mind.
As with most things in life, it is good to have a Plan B, or at least a disaster recovery plan. This applies to businesses, organizations, and private individuals. What kind of plan is right for you? Only you can decide that, based on your specific circumstances. With that said, every good plan starts with regular data back-up - preferably hard-copy.
More links regarding the outage: