On Thu, Dec 2, 2021 at 2:06 PM Gordon Messmer gordon.messmer@gmail.com wrote:
On 12/1/21 18:03, murph nj wrote:
lsof showed nothing that I could see.
Nothing, or nothing interesting? I'd expect you to at least see the CWD of the debug shell.
The error messages were continuously overwriting everything in the shell. If it was a single line, I may not have been able to read it.
The stop job rotates between four mounted volumes, and the lvm2-monitor.service.
In the past, I've seen lvm2-monitor behave badly in reaction to logical volumes (block devices) used for VMs that used LVM internally. Do you have any logical volumes that aren't used for those four mounted volumes?
No LVM on this system. (I don't recall why, I'm not averse to LVM, but this one doesn't have it.)
systemctl status doesn't seem to display much, but the stop job (and cylon eye) keep overwriting everything, making it really hard to see the output.
Several suggestions: Chris suggested that you check "systemctl list-jobs". The contents of /proc/mounts might be interesting. I'd run "lsof /path" where path is the longest mount point path of the volumes still listed in /proc/mounts.
I'll give it a shot next time it hangs up on shutdown. Any ideas on how to make the error messages abate (in the debug shell) so that I can read these things more easily?
Thanks, --murph