Also, Linode's SLA may be 99.9%, but I've experienced with two businesses, using Linode over the course of four years, about 45 minutes total of downtime (on any given server, not our entire infrastructure). I'd much sooner take years of past experience showing incredibly good uptime over a piece of paper meant to limit liability.
I believe Linode is an excellent choice for small-medium sized online companies.
This is especially true given that more often than not, within the first 5 times you use it, you'll experience some bug in your HA setup that leads to some complicated, slow-to-diagnose-and-fix failure.
Wow. That sentence makes 99.9% uptime sound really shitty.
Might as well stop counting sysadmin screwups and hardware failures, at that point. 100% uptime! except when we're down.
The current balance of power is tipped, right now, so hard in favor of the attackers that if anyone wants to take /you/ out, well, it's pretty difficult (read: expensive) to stop them, and if you are some $20/month customer, your upstream is probably going to just finish the job and cut you off... but if the guy next to you gets attacked and you are collateral damage? You should count that against your service provider. There are usually things we can do (as service providers) to limit collateral damage, even in the cases where we are unable to protect the target. If nothing else, providers who tend to tolerate sites that get attacked often, well, they suffer collateral damage from attacks, generally speaking, more often than sites that are less tolerant. It's sad, really 'cause sometimes the targets are not doing anything really wrong, but you've gotta protect your network.
There is some advantage of scale here, too... The DoS that is hardest to fight is the pipe-filling attack. If the attacker can fill your upstream port(s), then there isn't really much you can do, besides blocking the target at your upstream, (I mean, assuming the source is random, as it often is.) The larger your upstreams are, the more you can absorb before you have a choice between cutting off the target customer and having all your customers cut off.
Except maybe if those outages hit them at the worst possible moments of the month (just before a major announcement or something).
Now, managers would froth at the mouth hearing 45 mins/month is acceptable. But those managers are idiots.
I haven't seen any major site that didn't have those (and worse) outages.
5-6 hours a year? Heck tons of sites had those. AWS had them (and thus Heroku had them), GAE had them, Twitter had them, Tumblr (well they had one-two orders of magnitude more), etc.
45 mins a month represent lots of lost revenue to you? Really? If you're not Google or something, then what exactly do you do?
Shouldn't that be like 0.1% of your revenue? And that IF you sell online and IF we assume customers aren't gonna come back. Heck, you lose more because of random fluctuations in buyer preferences than that.
Things I've experienced:
- being rebooted without notice
- poor performance (IO and / or CPU)
- provider being DDOSed
- provider moving to a new DC and telling people not to give out IPs (???) due to being DDOSed
- hosts running old virtualisation software which doesn't support recent versions of Linux, so I'm can't use an OS newer than Ubuntu 10.04
- provider being banned by PayPal then suspending my account being I couldn't pay then
- provider going out of business
I liken it to poor quality manufactured goods. Take the Android tablets you can get for less than $100. If you can afford it, you would never buy one just to save money, because it is much less hassle to spent another $100 and get something that works well.
It just isn't worth it, especially not for a business. I would love to see someone who can make a proper fault tolerant system that works on these type of machines though :)
I have a VPS I found on lowend box. $5/month for a 512MB OpenVZ. It's a pretty good deal, and it does what I need it to do; I have full control over it, I can give other people access to it, I can run any kind of website on it as long as I don't care too much about performance or uptime.
But there's no way I would run any service that was bringing in money on it. It basically has downtime every single month. Usually only for a few minutes at a time, and they tend to scrape by just barely exceeding 99.90%, but still. There also was one time when they emailed me saying there had been a power loss or some other reason that the rack my server was on went down, and I actually ended up having to go into the console to start my server back up and get everything working.
I've also found their support ... not particularly supportive.
Sure, Linode is 4x as expensive for the "same thing" (though to get Xen from my provider is only about 3x cheaper than Linode), but the amount of total money is negligible. $15/month. Maybe if I needed more resources it would be a difference of $100/month, but if I need a 4GB VPS for a service that's meant to make me money, that should still hopefully be trivial. For a service provider with much better uptime and support, that's a no-brainer.