These servers were part of S3, Amazon’s renowned service hosting, and when they came out of the line itself took major online sites. Quora, Trello, and IFTTT were among the affected sites. Servers came online a few hours later.
Now, Amazon has revealed the cause of this problem. Tuesday, team members maintain a system S3 is billing.
As part of this process, they needed to disclose a server offline. Unfortunately, there was an error of text commands and offline brought a greater share of the server. 2 from S3 subsystems found on these servers.
These were important subsystems. One to manage the meta data and position information across the region. Without this subsystem can not perform the services: the common operations.
After the incident different systems require a full reboot, which requires more time. Amazon said the S3 was designed to deal with the loss of some servers, but was renewed mass systems that caused problems.
As a result, Amazon is making changes to allow the system to recover more quickly in the event of another accident. Also, the S3 engineers now can not remove capacity from S3 if it would lead down a number of subsystems limit the capacity of the server. / PCWorld Albanian