Why did Minecraft, Instagram go down during Azure, AWS outage. A recap

The glitch was linked to Azure’s global content delivery network and its Front Door services, due to which users were unable to access Microsoft 365, Minecraft, Outlook and other services for hours.

Microsoft Azure outage

The Microsoft Azure outage that hit on Wednesday significantly impacted the Eastern United States, where difficulties were said to have started at approximately 4 p.m. UTC (9:30 pm). In a statement, Microsoft said that the DNS problems caused the troubles, later adding that things have started to recover.

At one point, over 18,000 users reported issues in accessing Azure, as per the outage reporting website Downdetector. Microsoft also came forward to acknowledge the problem, stating that the recent configuration change triggered the outage.

On October 20, Amazon Web Services faced a major outage that affected platforms including Snapchat, United Airlines, McDonald’s, Venmo, Fortnite, Disney+ and others.

A quick recap of the AWS outage that hit on October 20

Also read: What caused Microsoft Azure outage? Tech firm says ‘We suspect that…’

What had happened

The root cause of the outage was a Domain Name System (DNS) error at AWS’s Northern Virginia data centre. DNS acts like a map for websites to translate their website name into IP addresses. But when the DNS system was affected, apps could not access their data.

What did Amazon say

Amazon.com Inc. had said that the issue appears to be related to the DNS resolution of the Dynamo DB API endpoint. They further said that the company has “applied initial mitigations” and is “seeing significant signs of recovery,” adding that most requests should now be succeeding.”

Which apps were affected

The applications that got affected from the outage included Venmo, Snapchat, Perplexity, Coinbase, HBO and HBO Max, Canva, Hulu and more. In a statement, Perplexity said that it is “down right now,” citing the AWS issue.

How many outage complaints were reported

The outage tracking platform Downdetector recorded over 11 million user reports of AWS issues, while Ookla said more than 4 million individuals faced the issue.

Also read: Microsoft issues big update as Azure outage brings down nearly half the internet

Recovery took hours

The recovery process from Amazon took hours. The mitigation process began following the identification of the problem at around 07:11 GMT (12:41 pm IST), and by 10:30 GMT (4:00 pm IST), AWS said that the DNS had been “fully mitigated”, though the clearing of backlog from the outage took several more hours. This was completed by evening when the company announced all services had “returned to normal operations.”