Package updates left you with too many processes using old files as shown by zypper ps that it’s easier to reboot instead of restarting the processes, but you are nostalgic about your long uptime record? Here’s how to preserve your uptime.
Log out of the desktop. Go to a virtual console and become root, then:
This changes the runlevel to single user. It will prompt you for the root password to login, then:
which will bring it back up to multiuser with GUI. Use
if you run it in CLI mode, e.g. a server.
Is there a possibility to preserve uptime when the updates bring in a new kernel?
No, the uptime relates to the running kernel and of course if you replace it … I suppose there might be a trick to preset the uptime of a new started kernel, but I think that’s carrying vanity too far.
I thought that ksplice would allow you to do exactly that, maybe I misunderstood. In any case, if I have made a significant change, I’d rather reboot and check that I haven’t done something that stops it from booting.
Yes, ksplice allows you to keep running the same kernel with live patches. But you would be doing that for the sake of continuous service and not the frivolity of seeing your uptime counter preserved.