Amazon’s S3 was down on Tuesday morning, and this created a partial chaos in the Internet world, though it didn’t completely break the Internet. AWS experienced a four hour outage, and this caused problems for the thousands of websites that depend on AWS for their storage and cloud computing needs.
Many major companies like Netflix, Pinterest, Spotify, Buzzfeed, Airbnb and Slack use AWS to store and retrieve their data at any time. An outage meant that these companies could not access the data they want, though the financial and operational loss for these companies is not known at this time. While this outage did not shut down websites completely, it affected certain functions like file sharing, collaboration and image uploads.
Again, not all its clients were affected, but still a substantial number experienced difficulties when a major chunk of S3 went offline on Tuesday. In fact, Amazon itself was affected by this outage, as it was not able to update its health dashboard for two hours, obviously because the data is stored on AWS.
AWS was not available for comments or updates on this issue immediately, but a message on their website read that they are working to fix the errors, and until then S3 customer applications will continue to have high error rates.
Reports on Twitter pointed to the poor response given by AWS for such an outage. The dashboard of AWS did not display any real-time events or updates, despite the company acknowledging that there was an error.
Outages and the resulting impact on businesses is one of the worst fears of any organization, and this incident brought to light the data reliability issues that still plague the cloud industry, despite all the advancements made in these areas.
Thankfully though, these incidents are not frequent, but when they come from large service providers, they become big news. The last outage from AWS that happened in September 2015 and lasted five hours.
While it’s not right to judge the industry as a whole and shun it completely, what is needed is a more cautious approach by everyone involved. Clients who depend on AWS should consider using a multicloud strategy or something like Nimble Cloud Volumes, that combines the power of Azure and AWS in the same platform. This way, even if there is a problem with one service, data reliability and business operations will not be affected in any way.
For AWS and other cloud providers, this outage may be a wake-up call, to increase the reliability and availability of their systems. This is definitely not something you would expect from the world’s largest cloud provider that has an annual revenue of more than $10 billion each year.
It is estimated that more than one million companies, from established giants like GE to startups like Snap, and many government agencies depend on AWS for their data storage and availability.
It is hoped that such incidents don’t happen in the future, and cloud providers do whatever is necessary to increase the rate of data availability.