Many large websites were suddenly no longer accessible on Tuesday – now it is clear what was behind them.
The total failure was triggered by a mistake at the cloud provider Fastly, based in San Francisco. The company announced that on Wednesday in one Blog entry famous.
“We experienced a global outage due to an undetected software bug that appeared on June 8th when it was triggered by a valid configuration change by a customer,” said the company.
Means: The admin of ONE company changed a setting – and thus triggered the error. Almost the whole internet crashed!
The White House was also affected
The websites of the British government, the “New York Times”, CNN, BBC, Amazon’s streaming platform Twitch and the online forum Reddit were affected.
In the UK were noisy “Daily Mail” also Amazon, PayPal, Spotify and Netflix can no longer be reached. The White House website was also affected.
According to Fastly, the outage was “broad and severe,” but the problem was quickly identified. Most of the network was operational again after 49 minutes.
According to the company, the error was contained in a software update that was rolled out in May. An investigation is now underway to find out why the error was not discovered during various tests.
Fastly self-critical: “Even if there were certain conditions that triggered this failure, we should have foreseen it.”
.