Continue to boot when mount point fails

I have Fedora (34) on a (mostly) NAS server. It has a mount point for a data store in /etc/fstab. If, for some reason, this volume fails to mount at boot, boot is halted. Obviously I do care if this volume fails to mount; however, there’s nothing on this volume that should be essential to the basic function of the system. It seems more sensible to me that boot should be allowed to proceed. (Then, at least, some attempt could be made to troubleshoot the problem remotely.) What is the correct way to do this?

I think adding nofail to the entry parameters is what you need.

1 Like

Use a systemd automount instead. That will make it much more resilient. I do that for all network and external drives.

This is what the /etc/fstab entry look like for a mount to a NAS using nfs. /mnt/backups nfs x-systemd.automount,x-systemd.device-timeout=10,timeo=14,x-systemd.idle-timeout=1min 0 0

See also: _netdev + nofail @ systemd.mount

Thanks; I think the systemd automount is the way to go.

I was considering that and even wrote the unit files; but I was behind the times a bit and didn’t realize that /etc/fstab is just used to generate those nowadays.

In my case, the volume in question is a btrfs RAID. A failure to mount most likely means there’s a problem with (some of) the drives.

BTW, I tend to use autofs for mounting nfs volumes, mainly because its interface for mounting user home directories is still way nicer than systemd's (last I checked, anyway).