Amazon’s EC2 Service Suffers Outage
gigaom.com
gigaom.com
How is this different from any other data center? Note too that on EC2, your instances will be spread randomly throughout the facility - you'd actually be better off having this happen on EC2 vs. in a traditional DC. Furthermore, this was a localized incident inside of a single availability zone - the other four (now five, actually) AZs were completely unaffected.
The author clearly does not know what he is talking about.
I'm glad one commenter on the article had the sense to call him out. Too bad he was followed by the usual set of trolls.
It should be noted that the 365 Main datacenter that hosts (hosted?) Craigslist, Six Apart, Technorati, and a number of high profile web companies had a total failure a while back (http://radar.oreilly.com/archives/2007/07/365-main-datace.ht...). Similarly, Rackspace (whose business is not having to worry about your servers) had 3 outages in 2 days (http://gawker.com/tech/followup/rackspace-outage-was-third-i...).
So, Amazon's 3 hour outage that affected a minority of their customers doesn't seem that outrageous. You're totally right. The author was probably writing a bit of link-bait to get read. No one reads "Amazon EC2 has minor outage that compares favorably against their competitors and affected a minority of their customers." That's just business-as-usual talk. However, if you question the entire viability of cloud computing because of a 3 hour outage, well, that deserves reading.
Once power was back, however, so were our instances, which was a pleasant surprise.
If this is the case, I'd interpret this outage as a testament to how good AWS is!
The report also states it was a problem isolated to a group of racks, a failing power distribution unit which I take it was on the "wrong side" of the UPS. A lightning storm affected something on the "inside" of the UPS, but nothing else in the datacenter? I'm curious how that would even happen (and shouldn't a lightning rod be in use?).
Also, it seems like Google's per-motherboard UPS system would have been the setup to have in order to avoid this problem in the first place.
Long answer: there's no way to know. The names that Amazon gives a specific location aren't something you can rely on. So, your us-east-1a might be a different availability zone from my us-east-1a. I'm guessing Amazon did this to avoid everyone "going with the default" and launching instances in us-east-1a and having way more instances wanted to be launched there than their other availability zones. But it does mean that there is no way of identifying by name what the availability zone was that went down since the only names we have identify different places depending on our account.