The internet appears to be recovering after a major outage knocked out thousands of websites, including Amazon, Snapchat, Reddit and several banks.

Downdetector, a website that tracks complaints about outages, showed thousands of problems on Monday morning.

The issue originates from Amazon Web Services (AWS), which describes itself as “the world’s most comprehensive” cloud service.

Internet outage – as it happened

It offers companies a virtual backbone, giving them access to servers, databases and storage without having to build their own infrastructure.

Millions of businesses are thought to use AWS and everything from Amazon’s Alexa to the Bank of Scotland to crypto company Coinbase, Duolingo, Fortnite, Halifax, HMRC, Lloyds Bank, Roblox and more were affected by this morning’s issue.

Why has a problem with one company caused so many issues?

Amazon is the cloud provider for so many companies that when it sneezes, the internet catches a cold.

Despite the name, the cloud is really just a collection of data centres – that is, a collection of computers stored in warehouses.

The problem here is apparently a DNS issue related to Amazon’s data centres in North Virginia.

DNS is the address book for the internet so this means that Amazon has a connection problem.

The data centre itself might be fine, but all those apps and websites can’t find it.

It’s like a postman trying to post a letter when the address has rubbed off.

The fact that this is happening in Amazon’s largest data centre complex makes the ripples even bigger.

For various reasons – cheap electricity, proximity to Washington DC, access to undersea cables – North Virginia has become a global centre for cloud storage, so an issue there has the capacity to affect anyone who depends on the internet.

These days, as we’re seeing, that’s almost everyone.

Read more: What’s affected by internet outage – all we know so far

In the US, there were long queues at LaGuardia airport in New York as check-in kiosks appeared to be failing and apps were down, according to The New York Times.

Some reservations weren’t showing up on airline apps, and customers complained on social media that they couldn’t drop their bags, CNBC reported.

The United Airlines’ website was among the sites that users reported issues with on Downdetector, which tracks internet outages.

What has Amazon Web Services said?

The UK government told Sky News it was “aware of an incident affecting Amazon Web Services” after HMRC was among those affected by the outage.

“Through our established incident response arrangements, we are in contact with the company, who are working to restore services as quickly as possible,” said the spokesperson.

Around three hours after the problems began, AWS issued an update saying its engineers were “seeing significant signs of recovery”.

“Most requests should now be succeeding,” it said, although it added it was working through a “backlog” of requests.

By 11.35am UK time, it said the issue was fully resolved, although their services might run slowly because of a backlog of requests built up over the morning.

‘I haven’t been able to log in’

Hundreds of Sky News readers got in touch to tell us how they had been affected by this morning’s outage.

“I deliver parcels for Amazon Flex and I haven’t been able to log in to my account all morning to be able to go to work today,” said Marie Louise.

A worried Snapchat user told us they have a “1,321 day streak, which I don’t want to lose, and I also have 14 years of Memories on there which I don’t want to lose.”

And one reader took it as an opportunity for a little treat: “MyFitnessPal is down. I guess the calories don’t count this morning, right?” wrote Callum.

What to do if you’re affected?

If you’re affected by the outage, you’re partly at the mercy of AWS to get their services back up and running, but there are things you can do in the meantime.

Hayley Andrews Smith, an automations specialist at Myriad, advised people on LinkedIn to check their business’s automated processes as “anything hosted on AWS might be affected”.

She also suggested pausing critical workflows until the systems are back up.

You should also keep your guard up for scammers.

When Vodafone’s internet services went down last week, users were immediately plagued by scammers pretending to be customer service agents.

“When companies rush to restore access, systems and staff are stretched thin — and that’s when attackers strike,” said Charlotte Wilson from cybersecurity firm CheckPoint.

“Expect a spike in fake ‘refund’ or ‘discount’ offers, phishing emails, and scam links claiming to fix the problem.”

Check email and web addresses from anyone saying they’re contacting you from AWS and do not click on any links you aren’t sure about.

Should Europe rely on US tech firms?

Coinbase, the world’s biggest bitcoin custodian and the largest US-based cryptocurrency exchange, has told customers “all funds are safe” after the outage and reassured users that their team was working on the issue.

Cori Crider, the executive director of the thinktank Future of Technology Institute, said the outage shows Europe “can no longer afford to run the majority of its critical infrastructure on US tech monopolies like Amazon Web Services”.

“The last time this happened, the Crowdstrike outage did billions of damage to Europe’s economy in a single day. How much harm will be done this time?”

Image:
Duolingo, Snapchat, Lloyds Bank and Wordle (clockwise) all showed error messages this morning

Last July, cybersecurity firm CrowdStrike accidentally brought parts of the world to a standstill when a software update introduced a bug.

The outage disrupted internet services, affecting 8.5 million Microsoft Windows devices.

Although it took just 78 minutes for the company to identify the problem and start rolling out a fix, the impact of the outage lasted far longer.

In the UK, the CrowdStrike outage left GPs unable to access systems that manage appointments or allow them to view patient records or even send prescriptions to pharmacies.

Flights were cancelled or delayed and passengers were left stranded as airline systems were knocked offline or staff were forced to handwrite boarding passes and luggage tags.