A few days ago it was the RIM network that suddenly went down, cutting people off from their emails and other BlackBerry goodness (which some saw as a good thing rather than a catastrophe) — and this morning it was Amazon’s S3 network that suddenly went offline. The network provides cheap remote storage for dozens of Web startups, including Twitter, as well as some larger companies. What users of those services wound up with for several hours was a host of 404 and other errors.
This is something I was just talking about on TVO’s The Agenda program (there’s more info here, if you’re interested), and I wrote another post about it after reading danah boyd’s account of someone who was suddenly cut off by Google. Ironically, in that post I mentioned how I was in the process of backing up my photos to Amazon’s S3 using JungleDisk. I also wrote a column for the Globe about the RIM outage (which is here) and the implications for cloud computing.
As a friend pointed out to me on Twitter this morning — when I described the S3 outage as “another caution flag in the cloud computing race” — these kinds of outages happen all the time with hosting providers (including my blog host) and other remote storage companies, not to mention Internet cables being severed under the ocean and so on. But every time it happens it’s another reminder about the need to maintain multiple redundant backups for the data you care about.