SLES11.1 boot fails if failed disk mount

ron7000 · 10-20-2011, 01:22 PM

I have used Suse Linux Enterprise Server (SLES) versions 10.1 and 11.1.

Problem I have in SLES11.1 is I have 2 harddrives in the system, and they are mounted by disk ID. disk 1 the boot disk needs to be mounted, it has the root partition and without it the system won't work (obviously). But disk 2 mounts as /scratch and is not required for the system to boot.
In SLES10.1 if I have the mount line in /etc/fstab for /scratch and if that drive is not there during boot up it will spit out an error saying hey I didn't find that disk but the system will continue to boot and get to runlevel 5 and be run fine without /scratch mounted.

In SLES11.1 for the same situation, the system will drop to single user mode and want root password to log in via shell window then gives me a repair filesystem prompt. I don't want this. How do I make it still boot when other non-required mounts fail in /etc/fstab during boot?

AlucardZero · 10-20-2011, 03:19 PM

Change the sixth field (fsck pass number) in the /scratch mount line to 0

ron7000 · 10-24-2011, 10:15 AM

thanks, I'll try that.

however they were the same,
in both sles10 and sles11 fstab files they were
/dev/disk/by-id/scsi-3500blabla-part1 /scratch1 xfs defaults 1 2

does sles11 do something differently ?

AlucardZero · 10-24-2011, 11:58 AM

It seems that way. Changing the 2 to a 0 will disable fscking at boot, therefore it won't panic when it can't fsck it.