Exactly a week ago we witnessed a more unique than rare event: Google and Youtube have suffered generalized downtime and inefficiencies . All of Big G's services were found to be impossible to use for an hour, resulting in difficulties for all users and businesses.
7 days later Google reveals the reason behind the inefficiencies of 14 and 15 December, which also caused problems for distance learning. The Google Classrooom and Meet services, used extensively by schools, were unusable for almost an hour.
Google and YouTube down: the causes of inefficiencies
Google first reassured users by explaining that it was not a hacker attack . In the same days, in fact, there was a cyberattack on US government companies that also affected Microsoft. The Mountain View company, on the other hand, seems not to have been hit. Behind the inefficiencies of Google and YouTube there is instead the new filing system . The migration started in October, but an error occurred on Monday that caused Google services to down. In a report on the Google Cloud infrastructure, the company explains the reasons for the inefficiencies, which began at 12:46 and ended at 13:33 (Italian time).
The presence of a bug on the memory management system caused the access capacity to the storage space to decrease , bringing it to 0. This prevented the correct authentication of the users, who were not able to access the services. Google uses User ID Service to store unique account identifiers, managing OAuth tokens for authentication. If the access attempt has obsolete data, the service prevents the login. When the new rules for the service were activated and the bug then presented itself, the write access was blocked (due to the memory set to 0). Consequently, the data obsolescence check for read operations was positive for all accesses , preventing users from authenticating.
The next day there were also problems with the mail, preventing many users from sending emails. The malfunction occurred during the night, so it did not impact the work excessively, at least in Italy, and the duration was much shorter. In this case the error was always related to the migration of Google systems, but it hit the GMail SMTP server.
The impact of disservices
Google's services failed to verify the correct authentication of user requests, thus showing page loading errors and preventing you from logging in to the various features. The most affected services were all those of Google Workspace Services , namely Gmail, Calendar, Meet, Docs and Drive. The Google Cloud Platform also suffered major disruptions, which impacted the Cloud Console, Cloud Storage, BigQuery and Kubernetes Engine.
In the first case, the apps have completely stopped the service for all users. The resumption took place around 14:00, but some users had problems until 14:30, due to the large traffic recorded immediately after the services recovered. A little better, but not too much, for the services of the Cloud platform: the Cloud Storage recorded only 15% of failed requests, and the Kubernetes Engine only 4% . Worse for BigQuery services: 75% of requests failed, and Cloud Networking also experienced a high failure rate.
The problem was promptly resolved and Google is working to upgrade its systems and speed up the response to errors. The company has in fact confirmed that, after a careful investigation of the accident, all the necessary measures will be taken to prevent and stem any future problem .
The article Google and YouTube down: this is why the disservices come from Tech CuE .