Cloud Hosting

Tuesdays widespread internet outage carries many lessons for us all, from the need to back-up cloud servers to an awareness that without the internet, we are in big trouble.

Hundreds of thousands websites crashed Tuesday from about 12:30 p.m. to around 4:30 p.m. The outage was a result of problems atone of Amazons server farms, located in a Virginia warehouse. It affected companies large and small, from the likes of Yahoo and Apple to the North of Boston Media Group. It affected how people work, shop or look at pet-trick videos. It affected Huffington Post, Imgur, Business Insider and many other mainstream media sites.

In short, it affected a lot of people.

Amazon Web Services, which is distinct from the companys retail business, hosts a large portion of what is known as the cloud, that vast repository of shared computer memory. Tangibly Amazons cloud is series of data centers enormous, climate controlled warehouses that keep the flow of data moving all over the world.

Amazon Web Services is far from the only company that leases shared serve space in the cloud, but it is a major player. By one account, it carries and stores data for about 30 percent of the companies on the internet. Thats a lot of businesses. Its been around since 2006 and, with the exception of a few blips along the way, has remained reliable.

Amazon said Tuesdays outage was caused by high error rates in one part of its simple storage service. That may be Greek to most, but what it means is that a major portion of a public utility that people rely on as much as electricity and water was offline. In 1990, not even Al Gore could have known how important the internet much less cloud computing would become to how the world operates.

Amazons outage showed how dependent we have all become on reliable internet and shared computing services. Here at the North of Boston Media Group, for example, access to essential production software was shut off for nearly half the day. The experience was a lesson to all that when these systems go down, we and everyone else need a back-up plan.

One solution for businesses is to build their own, on-site data centers. But these are costly, particularly compared to the relatively inexpensive services offered by Amazon.

With this wake-up call come questions beyond the obvious ones about how this happened, which Amazon so far hasnt publicly detailed. One question is this: What are the ramifications of allowing companies such as Amazon or Google to carry such large pieces of our shared computing and internet architecture?

Its kind of like the financial crisis of 2008 when huge financial companies began failing and needed government bailouts because they were too big to fail. Is Amazon now too big to fail? What, if anything, can be done about it?

It seems unlikely that President Donald Trumps administration will want to start telling internet businesses what to do, but it might be time for some scrutiny on the issue.

Public utilities are heavily regulated, from electric companies to water distribution systems. Isnt the internet a public utility? Wouldnt it be wise for an independent, third-party to take a look at whats happening?

This case doesnt appear to involve a malicious attack, but the problem of hacking is real and has dealt severe blows to the internet in the past. With everyone so concerned about security these days, it makes sense to take a look at the behemoth that Amazon Web Services has become.

Continue reading here:
Our view: Lessons learned from Amazon cloud outage - The Salem News