Why did Amazon’s AWS crash and can it recur?

0


that of Amazon (AMZN) – Get the Amazon.com, Inc. report The cloud computing service provider said on Friday it was developing a fix for the issue that caused a widespread outage on December 7 and expects the changes to be rolled out within the next two weeks.

The Seattle-based tech giant said it was confident that patching its network systems would fix the issues it was having and that there should be no recurrence of problems.

Amazon Web Services provided a summary on Friday of the Dec. 7 service outage that occurred in its Northern Virginia area, US-East-1, which hampered some of the biggest names in tech and affected thousands. users for about 11 hours that day.

Among the tech companies that reported issues throughout the day was Meta Platforms (FB) – Get the Class A report from Meta Platforms Inc. Facebook and Instagram, Zoom (ZM) – Get the Class A report from Zoom Video Communications, Inc., Reddit, Salesforce (CRM) – Get salesforce.com, inc. Reporting and Snapchat.

Plus, Amazon Alexa, Ring Doorbells, Walt Disney (SAY) – Get the Walt Disney Company report The Disney + streaming service and even Amazon’s in-house apps, including its delivery services, have taken a hit.

Amazon said that automated activity to scale the capacity of one of the AWS services hosted in its main AWS network caused a surge in connection activity that overwhelmed networking devices between its internal network. and the main AWS network, causing communication delays between them. networks. The malfunction resulted in persistent congestion and performance issues on devices connecting the two networks, according to the summary.

The company claimed that it immediately deactivated the scaling activities that caused the problem and that it will not resume activities until it completes the fix. It said it has also deployed an additional network configuration that protects potentially affected network devices, even in the face of a similar congestion event.

Network congestion issues have also affected the AWS Service Health dashboard, which the company uses to communicate service disruptions to customers. AWS has announced plans to release a new version of its Service Health Dashboard early next year to improve communications with customers.

Downdetector said nearly 30,000 problem incidents were reported at the height of the problem. Some Downdetector users have reported issues with the company’s ancillary services, such as Ring doorbells and Alexa home assistants.

Facebook was experiencing a less dramatic increase in problems with as many as 2,000 reports recorded by Downdetector. Facebook’s problems could be linked to the AWS outage.

Amazon shares on Friday closed 1.1% higher at $ 3,444.24. The stock fell to $ 3,440.90 after hours.


Share.

About Author

Comments are closed.