Originally published by Info Security Magazine
Big tech companies have monopolized the internet, and it’s not good for business. We know this – it’s been a focal discussion in the mainstream press for at least the last ten years. But their power and dominance in the digital world are only increasing, and with little bite back from governments, so it’s up to businesses to act for their own good.
I don’t need to tell you about Amazon’s hold over the internet, whether it’s streaming content, buying products, or connecting your business to the cloud. Keith Ingram, owner of vinyl records shop Assai Records, told the BBC that during the pandemic, “we wouldn’t have survived without Amazon.” But while Amazon has been helpful to companies during global lockdowns, the first thing you should not do as a business owner is get in a position where you rely on a single provider, or a single piece of tech, to survive. Because variety is not just “the spice of life,” it’s how economies thrive.
What This Means for the Cloud
The argument around diversifying tech – especially providers of cloud connectivity – can’t be made more apparent than through last month’s website outage, where we saw hundreds of the world’s most popular websites go down. To give a full perspective, Fastly is responsible for managing 10% of the world’s internet traffic, which includes the UK’s Covid vaccination website, the White House’s website in th e US, Amazon and Reddit’s global websites, among numerous other well known, well used and critical websites and services.
Relying on one provider causes an obvious risk; if something happens to that provider – in this case Fastly – 10% of the world’s internet goes down. When a website goes down, there is no traffic, and traffic = revenue. Gartner says that each minute a website is down, it costs a company on average $5600. For Amazon, the biggest e-commerce company on the planet, you’re looking at a figure of around $220,000 per minute. But it’s not just the monetary implications of your website going down – your business’ reputation is at stake. Not to mention vital online services that were not just disrupted but eviscerated, albeit for about an hour.
Looking at the internet and cloud computing, the more internet a single company manages, the more fragile the ecosystem becomes. Tech giants are centralizing a technology that was built to be decentralized. It was built for innovation – for different services and businesses to be available to all but for them to also be able to fail and not completely disrupt the lives of millions – or billions – of people. The more we centralize, the less reliable and safe the internet will become.
What Went Wrong
In this instance, it’s believed a new configuration was pushed out to Fastly’s entire network, causing the issue. There are two things to consider here. First, the issue pointed to human error, which can be identified and solved in minutes in a smaller company. Human error in a company where thousands of people work on the tech could take anywhere from hours and days to weeks or even months to properly isolate. They were lucky this time – they got companies back online relatively quickly – but how much money did each business lose from going offline? And what happens next time something like this happens?
The second thing to consider was that the companies whose websites went down clearly did not have a proper disaster recovery plan in place. They relied on Fastly to store their data and connect them to the cloud without considering what would happen if the entire Fastly network was disrupted.
How Can Companies Protect Themselves
Whether it’s a human error like Fastly’s bad code or an external impact like the OVH fires, a good disaster recovery plan means that no matter what happens, you can get your company back online in minutes. In simple terms, it’s a safety blanket for when disaster strikes.
While the cheapest disaster recovery plan is to share your data across multiple locations with the same vendor, outages can be avoided entirely by using two or more vendors, as traffic is routed to the secondary site in the event of a disaster. In that sense, some of these companies may well have had a disaster recovery plan in place that was not adequate for the size and importance of their cloud connectivity.
To properly protect themselves, companies must diversify the tech they use. Do not rely on a handful of big tech providers to manage your internet connectivity. Hopefully, Fastly’s outage will serve as a wake-up call to businesses to stop putting all their eggs in one basket when it comes to cloud connectivity and to put a proper disaster recovery plan in place.
