A dnf update with a new kernel, then the usual reboot. Except this time the machine doesn’t come back: no ping, and it no longer answers ARP on the local network. It came back half an hour later, after a manual power cycle.
First question once it’s up again: was it the OS (kernel, packages, filesystems) or the hardware that got stuck? The obvious first move, the journal, doesn’t help much here: journalctl only keeps the current boot on this machine. More in the next post.
This post is part of a series of sysadmin notes (tag “sysadmin notebook”).
Series “The server that never came back”, part 1/3. · Next post → · Version française