New York start-ups love them some Amazon Web Services, the cloud hosting platform that remotely stores data and makes it easy to add new servers quickly in the event of an increase in usage. But the giant stumbled this morning with an outage at its Eastern data center which took down sites across the web including Reddit, Foursquare and Quora, and while many sites are back up, some continue to experience downtime; locally, web app maker Aviary has been down all day.
Start-ups are naturally disgruntled. Betabeat reported earlier that Jack Groetzinger, co-founder of General Assembly-housed start-up SeatGeek, says he’s considering moving to Rackspace, which provides similar services to about the same number of customers as Amazon. “In the tech world people know this is Amazon’s fault, but the average person visiting our site is going to think we are the ones with the tech problem,” he said.
Amazon hasn’t said anything to its customers yet about refunds, discounts or other compensation for the downtime, which cost a lot of companies time, money and reputation. But some compensation will be in order, considering the company promises its customers 99.95 percent uptime.
Aviary’s CTO Israel Derdik was frustrated but sympathetic. “Communication from the Amazon Web Services has been kind of sporadic, they’ve posted regular updates but have been vague as to ETAs on fixes. ‘A couple of hours’ is hardly useful,” he told Betabeat in an email. “I’ve been on the other side of this scenario so I know what it feels like to be under the gun to get things up and running again. They have all hands on deck working on it and I’m sure they’re not slacking.”
Interestingly, he also said the outage could cause some customers to spend more money with Amazon Web Services. “What I will say is that this incident really makes people evaluate their redundancy and fail-over policies on EC2. If anything, AWS will start generating more revenue as folks create and maintain new redundant servers in different regions,” Mr. Derdik said.
Has Aviary considered moving to Rackspace?
It’s not that easy, Mr. Derdik explained. “We’re pretty heavily invested in AWS’s platform so we will probably add a few redundant servers on the West coast as a failsafe,” he said, but he acknowledged Amazon’s competitors are worth a look. “I’m also going to reevaluate Rackspace and see if they can provide us the redundacy we need. That would require learning the quirks of a new platform so I’m not loving that idea. Microsoft’s Azure might work for us as well since our web servers run Windows,” he said.
Betabeat also contacted Thrillist, which experienced issues including a delay for their crucial daily newsletter. “They are talking to our tech guys. They are working fast I’m sure,” Thrillist CEO Ben Lerer told Betabeat this morning.
We asked how pissed they were, on a scale of one to 10. “It’s annoying but it’s a 3,” he said. “This hasn’t happened before and it’s not like I’m the only one who’s down.”
Foursquare, which experienced intermittent issues today, declined to say how pissed they were and referred us to their status blog, which reads: “Our usually-amazing datacenter hosts, Amazon EC2, are having a few hiccups this morning, which affected us and a bunch of other services that use them. Everything looks to be getting back to normal now. We’ll update this when we have the all clear. Thanks for your patience.”
The most recent updates from Amazon say its Northern Virginia Elastic Compute Cloud and relational database are experiencing “instance connectivity, latency and error rates,” and its Elastic Beanstalk is experiencing increased error rates.
Sounds like a big opportunity for Rackspace, but so far evangelist Robert Scoble has remained silent.