Amazon has completed its recovery efforts from the extended crash of its popular cloud-computing services, but there will be lasting damage, both to customer data and to Amazon's reputation.
Amazon has reported on the AWS Service Health Dashboard that it has succeeded in recovering nearly all of the stuck volumes of data in its US-East Region facility, but 0.07 percent of those volumes will not be fully recoverable. The company has not specified how many customers will be impacted, but those whose data can't be retrieved will be frustrated, to say the least. It may also prove further cause for AWS users to reassess their use of the service.
Notably, that 0.07 percent of irrecoverable data represents just a portion of losses that Amazon's customers suffered from the outage. A venture-backed startup called Springpad, for example, saw the Web-based portion of its service go down for hours due to the AWS outage. Springpad offers a free service through which users can store and access all types of information via Web or mobile devices, from appointments and events to product codes and notes, while receiving targeted offers from businesses, such as a discount at a nearby restaurant.
The company's CEO, Jeff Janer, said that AWS stores all of Springpad users' critical data; when the AWS services went down, users were effectively cut off from their daily planners. While users were left frustrated, Springpad itself missed out on signing up new users for nearly two days, not to mention losing lead-driven revenue from its existing ones.
The severity of the outage caught Springpad off guard for a couple of reasons: First, Janer said AWS had been very reliable; previous outages had been few and far between and hadn't lasted longer than 30 minutes. Second, he said that he didn't anticipate that Amazon would have a single point of failure, "even in one zone that is sitting on multiple clusters."
During the outage, Janer said the company did have the option of transferring older snapshots of user data to another AWS facility, but it would have been a time-consuming endeavor. The company opted to wait and see when Amazon would be able to restore service.
The ordeal has not soured Janer from the cloud; in fact, he attributes the availability of a service such as AWS for Springpad's success: "We're a fast-growing venture-backed company. [The cloud] gives us the ability to scale in real time, to add servers and add capacity. It's been key to our business."
Springpad is, however, is reassessing alternatives to its current relationship with Amazon. Janer said the company could double down and pay extra to have its service hosted simultaneously in Amazon's East Coast and West Coast facilities. It could invest in its own IT gear and collocate in a third-party data center. Another option: It could sign on with an alternative provider like Rackspace. First and foremost, though, Janer said Springpad wants answers from Amazon as to what caused the outage -- which is still being investigated -- and what it will do to prevent it from happening down the road.
This story, "Some data irrecoverable after Amazon Web Services crash," was originally published at InfoWorld.com. Get the first word on what the important tech news really means with the InfoWorld Tech Watch blog. For the latest developments in business technology news, follow InfoWorld.com on Twitter.