Local chief information officers and IT managers would be wise to take a closer look at their service-level agreements with Cloud providers following the Amazon EC2 service outage in eastern United States over the weekend.
In-depth: Legal issues in the Cloud.
According to Longhaus research director and former Wilson HTM CIO, Scott Stewart, the outage would likely prompt questions from companies looking to adopt Cloud services.
"There will be concerns about what we are signing up for and what are our contractual rights if something goes wrong,” he told Computerworld Australia. "Each Cloud vendor provides some level of redundancy, but it doesn’t address the issue if overall access to that vendor goes down."
Amazon offers 99.95 per cent up-time in its standard agreement with clients for each of five global regions, but does not guarantee up-time for each ‘availability zone’; multiple redundant data centres used in each region.
However, he agreed with some views that the client was ultimately responsible for taking any kind of risk, like Cloud adoption.
"Cloud is the way we’re going, but it brings back to CIOs and IT managers that they have to manage that risk and remember that like electricity blackouts, the internet will go down," he said.
"We get blackouts with our electricity, it doesn’t mean we don’t love electricity, we just get annoyed with the provider and shop it around to someone else."
The outage, which affected popular Web services including Reddit, Foursquare and Quora, also had caused outages to Australian companies LabSlice and pizza franchise Crust Gourmet Group.
The latter’s chief executive officer, Michael Logos, said that the company had difficulties with its online ordering from Friday through to Sunday over the weekend, as a result of the Amazon outage.
In-store point of sale ordering systems were not affected.
“As soon as we detected the issue, we placed a message on the site letting people know what was happening, apologising for the inconvenience and encouraging them to call our stores to place orders in the interim," he said in an email.
The company had disaster recovery procedures to prevent full impact on customers during the outage but Logos admitted that the intermittent nature of the failures made it difficult to enact these procedures.
“We’re currently implementing a fail-safe system that will ensure that if such an event such was to occur again, we could flick a switch to move to a new system, without negatively affecting our customers."
Logos would not say if the company was reviewing its Cloud service-level agreement with Amazon. However, he said he was content with Amazon’s response to the issue.
Longhaus’ Stewart predicted that companies would increasingly move to applications that could run across several Cloud vendors, mitigating against the risk of future outages.
However, he cautioned this would be some years away yet.
"Being tied to one platform is going to prove problematic, so we will go with any network," he said. "In years to come we may have continious Cloud services supporting our applications using whichever network has availability."
Has your business experienced an outage due to Amazon Cloud? Let us know below!
Follow Hamish Barwick on Twitter: @HamishBarwick
Follow CIO Australia on Twitter: @CIO_Australia
Join the CIO Australia group on LinkedIn. The group is open to CIOs, IT Directors, COOs, CTOs and senior IT managers.