I don't know if they're the best at everything. I've been using Google Compute Engine extensively lately and I think it is a lot more polished feeling at times than EC2.

I am interested to know which part of the GCE is a lot more polished than EC2.

One data point: On AWS, a hardware outage takes your machine down (with notification). GCE offers live migration, which you generally don't even know has happened. See http://www.rightscale.com/blog/cloud-industry-insights/googl...

Note: GCE live migration take place only when there is a "scheduled maintenance", it is not designed for immediate hardware failure, so it is not a silver bullet and should not rely on them for your service's availability.

You should be building your AWS infrastructure around failure scenarios, and while you should be notified of a failure, your app should continue to chug along.

/DevOps

Isn't he just saying that Google does this automatically for you?

Well the TCP load balancer is pretty nice. Our EC2 load balancer (HTTP, through heroku) seemed to die all the time.