I would note that the kernel watchdog timeouts here are always at 20 odd seconds. They are not increasing so whatever is occuring is progressing at least as far as the kernel is concerned. If we assume the systemd log is still working (and it was shortly before the event when it reported reaching shutdown state) then we would expect it to be in the process of attempting to deconstruct the system before calling reboot. Most of the deconstructors it calls are reported before calling. There is one, cg_trim(), which is not announced. Looking at the implementation of that it is doing a hierachical remove of the /sys/fs/cgroup hierachy. On my system this is some 15000 files in 1300 directories. If there was a performance issue in there we could easily spend hours in this call with nothing logged.
-- You received this bug notification because you are a member of Ubuntu Bugs, which is subscribed to Ubuntu. https://bugs.launchpad.net/bugs/1730717 Title: Some VMs fail to reboot with "watchdog: BUG: soft lockup - CPU#0 stuck for 22s! [systemd:1]" To manage notifications about this bug go to: https://bugs.launchpad.net/ubuntu/+source/linux/+bug/1730717/+subscriptions -- ubuntu-bugs mailing list [email protected] https://lists.ubuntu.com/mailman/listinfo/ubuntu-bugs
