1-touch linux restore generating fstab file--feature request.

  • 29 March 2023
  • 1 reply

Badge +4

Not sure how old this behavior is but im using the latest Linux 1-touch 11.28 x64 iso. I restored a linux server to original hardware. When booting the server after restore, we noticed the nfs volumes were not mounted. they were usually mounted in the fstab file. When i looked at the fstab file it was not the original we used but a new one created automatically by the restore. It looks like it saved the original fstab in a backup file. Its easy to fix, either manually edit the fstab file or restore the old one but its strange we have to do this.

There is an option to ‘Retain only selected mount points in /etc/fstab’ (checked by default) I tried to test it with this checked and unchecked and either way the old fstab was not restored and nfs mounts were not mounted. I dont know if there are other ways in which this might cause a problem but i suspect there might be. 

It’s not a huge deal but its very annoying when it seems restoring the original fstab file should be the default option. From experience these restores can happen in stressful time sensitive situations in the middle of the night, where people are not thinking straight or only those who have less experience are working and these little things can cause big delays in getting servers running again. Not a linux expert but id dont see a huge issue if there is an fstab entry for a volume that you haven't restored, it just won’t mount right?


Best answer by Damian Andre 30 March 2023, 01:55

View original

1 reply

Userlevel 7
Badge +23

Hey @gibby101,

Speaking to one of our engineers, I don't think this is expected behavior. It should be retained upon restore. Side note I submitted a documentation modification as the “Retain only selected mount points” option isn't explained anywhere - but I would assume selecting that option would limit the entries in fstab.

I know we have to make changes to fstab especially for restoring to dissimilar hardware, and we are conservative there to ensure that we can boot the system (non-boot is the worst scenario) - but a case might be the best way to investigate this and see if its expected, or if there is something about your NFS mounts that made the software exclude them.

In either case, feature requests have to have a support case number assigned to them generally, so it may result in a feature request which is a good outcome anyway!