Who pays when cloud services go wrong, creating a service disruption? It happened April 20 for users of Amazon Web Services when its Northern Virginia data center suffered a service outage, knocking thousands of websites off the internet for 30 hours.
What are the remedies for these buyers? Ben Trowbridge, CEO of Alsbridge, has negotiated cloud computing agreements with Amazon for Alsbridge clients. Trowbridge says Amazon's standard service level agreement (SLA) on uptime is 99.95 percent for the year for computing. That means a client can expect to be down for six hours, 43 minutes and 12 seconds every 365 days.
Since 30 hours is clearly over the SLA, the typical agreement calls for Amazon to pay the buyer a service credit equal to 10 percent of its monthly bill. Calculating the cost, if the client was paying $20,000 a month, then Amazon would credit the account $2,000.
Will that cover the lost revenue? While service providers generally do not offer an SLA credit covering direct damages, Trowbridge suggests some clients "should ask for increased penalties from their cloud providers. It's a negotiation point."
