Synopsis
Begin 2017-04-12 07:22:00
End 2017-04-12 09:05:00
Affected Users of virtual machines that run on vos or wolf

An unattended update of the virtual hosting software (libvirt) triggered a netwerk stack restart. This made the virtual machines unavailable on the network. After a reboot of the virtual hosts (vos and wolf), the virtual machines were available again. Action was not taken immediately, because the alarm mails didn’t reach us, because the mailservers could not be reached. To improve this situation, steps are taken to route the alarms also via alternative channels (sms, whatsapp, signal, etc). Also the network configuration of the virtual machines has been changed, this will hopefully make it less vulnerable to these updates.

Affected services:

  • Email: smtp1, smtp2, mx1 (mail servers)
  • Web services: est2, redmine2 (webserver, redmine server of blackholecam.org)
  • Windows users: ms3 (windows login server), bfacdc03, bfacdc04 (windows domain controllers)
  • GitLab users: gitlab.science.ru.nl, gitlab.pep.cs.ru.nl (GitLab servers)