According to Bloomberg, AWS, a cloud service under the Amazon flag, has resolved service failure on Monday for approximately 15 hours. The event highlighted the high dependence of the global Internet on the cloud service of a single company.
According to the world’s largest cloud calculator, AWS services had “normalized” by 6 p.m. Monday, New York time.
AWS accounts for about one third of the global cloud computing market. According to data from Downdetector of the monitoring platform, the network services of hundreds of companies were affected by the failure, including the financial services platform Venmo and Robinhod Markets, Apple’s music and television services, software companies such as Zoom Communications, Salesforce and Snowflake, the catering giant McDonald’s and the game company Epic Games. Alexa, the voice assistant under the Amazon, and Ring, the family security system, were not spared.
Corey Quinn, Chief Cloud Economist of Duckbil Group, Inc., stated that this could be the worst crash of AWS since the December 2021 crash. “The question is, is this a serious incident? Or is it because we are more interconnected and rely more on the Amazon?” he says.
AWS had earlier indicated that the digital catalogue of a key database service had malfunctioned, resulting in software relying on the database not being able to retrieve information, thereby triggering a chain failure. According to the company, this fundamental problem was located and repaired in the early morning of Monday New York time, and the failure mainly affected the area of AWS operations on the east coast of the United States, its largest data centre cluster.
However, in the course of the repairs, the engineers found that other subsystems were also affected by database failures, including the critical components that clients used to activate the new leased servers.
Most of the failures of large technological systems are usually quickly repaired. However, the high degree of interconnectivity of technology systems means that a company ‘ s problems may trigger a chain reaction globally. Last year, a software upgrade error by CrowdStrike, a network security company, led to the suspension of flights and the collapse of systems worldwide, resulting in billions of dollars in losses.
