Description of problem: A NFS mount declared as "bg" in /etc/fstab doesn't work as expected. It ends up blocking the bootup process. Version-Release number of selected component (if applicable): systemd-208-9.fc20.x86_64 How reproducible: 100% Steps to Reproduce: 1. Add a NFS mount of an unreachable NFS server in /etc/fstab, using "defaults,bg" as the mount options 2. Reboot Actual results: The reboot stalls at the mount job, waiting for a timeout. Expected results: The mount process should be backgrounded and the boot process continue immediately. Additional info: nfs(5) describes the bg mount option: If the bg option is specified, a timeout or failure causes the mount(8) command to fork a child which con‐ tinues to attempt to mount the export. The parent imme‐ diately returns with a zero exit code. This is known as a "background" mount. The fstab line that triggers the bug on my system is: 192.168.0.2:/fud /fud nfs defaults,bg 0 0 tore@sloth:~$ systemctl status fud.mount fud.mount - /fud Loaded: loaded (/etc/fstab) Active: failed (Result: timeout) since ti. 2013-12-24 12:35:17 CET; 21s ago Where: /fud What: 192.168.0.2:/fud Process: 1132 ExecMount=/bin/mount 192.168.0.2:/fud /fud -t nfs -o defaults,bg (code=killed, signal=TERM) des. 24 12:16:06 sloth.fud.no systemd[1]: fud.mount mounting timed out. Stopping. des. 24 12:16:06 sloth.fud.no systemd[1]: Mounted /fud. des. 24 12:35:17 sloth.fud.no systemd[1]: Unit fud.mount entered failed state. As you can see from the timestamps above, the start job stalled for 20 seconds. This stalled the entire bootup process, too.
Well, systemd watches the kernel for the mount to show up, and right-fully so as it shouldn't proceed with other mounts that are inside of this one before it actually was mouinted in the kernel. If you don't want to wait for the thing at boot, use "nofail".