I’m sure many of you who even glazed over tech headlines this week saw that Amazon.com cloud web hosting, called EC2, had some major problems in the Virginia data center.
How can this be? Hasn’t everyone said that your website can’t go down if its on a cloud environment? Wasn’t the fact that all these computer servers are working in unison going to eliminate website downtime? HA! Computers aren’t perfect, so sometimes problems happen.
Just like in a car engine, there are many variables that factor into web hosting. Not just cloud web hosting, but any kind, shared, “grid”, VPS, dedicated etc and so on. Things happen, so don’t believe any web host that guarantees 100% reliability 365 days a year, 24 hours a day.
There are 8760 hours in a year, and many web hosts guarantee 99.9% reliability – which is really good, your website will be down about 9 hours a year.
Having said all this, it sounds like I’m bashing cloud hosting. I’m not at all, I actually love the cloud, but it has its gotchas like everything else in web design & development.
UPDATE: I have a few applications on Heroku, a Ruby on Rails hosting environment that runs on Amazons’ EC2. One application is a test skeleton website to test a RoR CMS calld Refinery. You can see it, or not, here: . At this time, almost 10:30am Saturday, it’s still having problems…days after the original problems. This proves how large the problem is for Amazon (note use of the word “is” not “was”).