On March 24th, 2023, we received reports that some customers were seeing an error message indicating that the site was down, when the site was in fact up. Our engineers determined that one of our servers was generating an increased number of error messages, indicating that the server was unable to process the incoming requests. The server was replaced and the issue was resolved.
A server was generating an increased number of error messages. This meant that when attempting to load a page, if the request was handled by the server having issues, customers would see the error message. Requests handled by other servers were not impacted.
We're upgrading our tooling to provide better metrics in order to reduce the time to diagnose and resolve similar issues faster. We are also updating our server configuration to help mitigate this type of issue moving forward.