Tuesday 23rd May 2017

Main web service Downtime last night

Tonight between 00.30 UTC and 6.05 UTC, pretix was not reachable due to a server failure. pretix is constructed in a way that a single server failure should never cause such a downtime. Therefore, this could only happen because of multiple failures happening at the same time:

  • The failed server did not try to restart automatically, as it should
  • The backup server did not properly take over
  • The monitoring system detected the failure but did not assign the correct urgency level that would be required to wake us up

We'll be working on each of those errors over the coming days to ensure that this will never be the cause of such a downtime again. We are terribly sorry for any caused inconvenience.