retitle 743265 systemd drops into emergency mode if devices from /etc/fstat are missing thanks
Hi, Am 02.04.2014 05:59, schrieb Norbert Preining: > On Tue, 01 Apr 2014, Michael Biebl wrote: >>> Apr 01 15:49:04 wienerschnitzel systemd[1]: Failed to start Load Kernel >>> Modules. > > I confirm that adding the "noauto" option to the strange devices > allows booting with init=/bin/systemd. Another option is to use nofail [0]. > The above message interestingly remains, but that might have > different reasons (self compiled kernel). > > I guess it boils down to whether non-existing devices should break > the boot process. With sysv scripts this what not the case, > just a warning messages. I actually do think that systemd handles this correctly. How is it supposed to know if the mount points listed in /etc/fstab aren't vital for the system to function properly unless you tell it so? Say you do have devices in /etc/fstab which do not show up under sysvinit in time during boot. It will just happily continue to boot. If there is e.g. data on those devices which are critical for your server processes to work properly, you might get data loss, undefined behaviour etc. which can be hard to diagnose. Dropping into the rescue shell and letting the administrator examine the situation looks like a much better alternative to me. Can we do something to address this nonetheless? Should we add something to systemd's README.Debian? Do you consider it important enough to add this information to the release notes? Regards, Michael [0] man 5 fstab -- Why is it that all of the instruments seeking intelligent life in the universe are pointed away from Earth?
signature.asc
Description: OpenPGP digital signature