and it appears that two out of their 3 availability zones have to go down in order for anyone to collect on sla. This effectively makes the sla meaningless afaict.
But then, I bet a lot of people use ec2 like an ordinary VPS provider, which is to say they keep their image up all the time. (this is why I see them as a competitor, even though I rent by the month, not the hour, meaning you can do things with ec2 that you can't do with me. I am working on better provisioning/billing scripts, but I've been talking about it for a while, and it hasn't happened, while I have to say the ec2 system is pretty slick)
* scheduled maintenance and emergency maintenance and upgrades * failure of access circuits to the Joyent Network * inability to obtain raw materials, supplies, or power used in or equipment needed for provision of this SLA
So, power outages aren't their problem. If the server goes down and they perform emergency maintenance, but can't get the equipment they need for 48 hours that isn't their problem. If the network goes down, that isn't their problem.
While Amazon is being a little cheap with their SLA, at least there are circumstances where it could be triggered. That's why I always respected Slicehost's SLA:
"Do you offer an SLA?
Not for Slices and here’s why: most hosting SLA agreements are just plain silly. They promise things like 99.9% uptime, but downtime excludes: scheduled maintenance, network outages, hardware failures and software trouble. Well what exactly is left to cause downtime? Here’s our SLA: we’ll do our best to keep your machines running smoothly for as long as possible and get them up ASAP should something go wrong."
And in over a year I've never had any downtime that wasn't me rebooting the box.
/I have no affiliation with Slicehost beyond being a customer.
This does leave the door open for a more reliable cloud provider who takes care of most of the HA for you.